The Trader's Guide To Key Economic Indicators Bloomberg.the Traders Indicators.2004.isb

User Manual:

Open the PDF directly: View PDF .
Page Count: 284

Download
Open PDF In Browser	View PDF

The Trader’s Guide to

Source:

Key Economic Indicators

Also available from
Bloomberg Press
New Thinking in Technical Analysis: Trading Models from the Masters
Edited by Rick Bensignor
Tom Dorsey’s Trading Tips: A Playbook for Stock Market Success
by Thomas J. Dorsey and the DWA Analysts
Profit in the Futures Markets!:
Insights and Strategies for Futures and Futures Options Trading
by Jake Bernstein
Wall Street Secrets for Tax-Efficient Investing:
From Tax Pain to Investment Gain
by Robert N. Gordon with Jan M. Rosen
Investing in Hedge Funds: Strategies for the New Marketplace
by Joseph G. Nicholas
Investing Under Fire: Winning Strategies from the Masters
for Bulls, Bears, and the Bewildered
Edited by Alan R. Ackerman
The Economist Guide to Economic Indicators (5th Edition):
Making Sense of Economics

A complete list of our titles is available at
www.bloomberg.com/books

This book is available for bulk purchase at special discount. Special editions
or chapter reprints can also be customized to specifications. For information, please e-mail Bloomberg Press, press@bloomberg.com, Attention:
Director of Special Sales, or phone 609-750-5070.

Source:

Attention Corporations

BLOOMBERG PROFESSIONAL LIBRARY

The Trader’s Guide to
Key Economic Indicators

Source:

R ICHARD YAMARONE

p r i n c e t o n

© 2004 by Richard Yamarone. All rights reserved. Protected under the Berne Convention.
Printed in the United States of America. No part of this book may be reproduced, stored in a
retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the publisher except
in the case of brief quotations embodied in critical articles and reviews. For information, please
write: Permissions Department, Bloomberg Press, 100 Business Park Drive, P.O. Box 888,
Princeton, NJ 08542-0888 U.S.A.

BLOOMBERG, BLOOMBERG NEWS, BLOOMBERG FINANCIAL MARKETS,
OPEN BLOOMBERG, THE BLOOMBERG FORUM, COMPANY CONNECTION, COMPANY CONNEX, BLOOMBERG PRESS, BLOOMBERG PROFESSIONAL LIBRARY, BLOOMBERG PERSONAL BOOKSHELF, and BLOOMBERG
SMALL BUSINESS are trademarks and service marks of Bloomberg L.P. All rights
reserved.
This publication contains the author’s opinions and is designed to provide accurate and
authoritative information. It is sold with the understanding that the author, publisher, and
Bloomberg L.P. are not engaged in rendering legal, accounting, investment planning, business
management, or other professional advice. The reader should seek the services of a qualified
professional for such advice; the author, publisher, and Bloomberg L.P. cannot be held responsible for any loss incurred as a result of specific investments or planning decisions made by
the reader.
First edition published 2004
1 3 5 7 9 10 8 6 4 2
Library of Congress Cataloging-in-Publication Data
Yamarone, Richard.
The trader’s guide to key economic indicators / Richard Yamarone.
p. cm.
Includes bibliographical references and index.
ISBN 1-57660-139-0 (alk. paper)
1. Economic indicators--United States. 2. Investments--United States. I. Title.
HC106.83.Y35 2004
330’.01’5195--dc22

2003027990

Source:

Acquired by Kathleen A. Peterson
Edited by Elizabeth Ungar

Source:

To Suzie,
Milton, Oskar, and Nash—felinus economicus

This page intentionally left blank

Contents

Acknowledgments

Introduction

The Business Cycle
Indicators and the Markets
How to Use This Book
Who Can Benefit from This Book?

3
5
6
8

Gross Domestic Product

Evolution of an Indicator
Digging for the Data

13
14

Some Definitions

GDP Versus GNP

Calculating GDP: The Aggregate Expenditure Approach

Nominal and Real Numbers

Deflators

National Income

Employee Compensation

Other Income Categories

GNP, GDP, and National Income

What Does It All Mean?

Source:

xiii

GDP Growth

32
32

Deflators

Consumption Expenditures

Investment Spending

Government Spending

Net Exports

Final Sales

Corporate Profits

How to Use What You See
Tricks From the Trenches

46
46

Evolution of an Indicator
Digging for the Data

53
53

Coincident Index

Leading Economic Index

Lagging Index

What Does It All Mean?

Coincident Index

Leading Economic Index

Lagging Index

How to Use What You See
Tricks From the Trenches

66
67

The Employment Situation

Evolution of an Indicator
Digging for the Data

71
72

Household Survey (A Tables)

Establishment Survey (B Tables)

What Does It All Mean?

Employment, Unemployment, and the Business Cycle

Inflation Indicators

Sentiment and Unemployment

Average Hours Worked and Temporary Workers

How to Use What You See
Tricks From the Trenches

Industrial Production and Capacity Utilization
Evolution of an Indicator
Digging for the Data

85
86

89
90
91

Industrial Production

Capacity Utilization

What Does It All Mean?
Industrial Production

95
95

Capacity Utilization

100

How to Use What You See
Trick From the Trenches

104
104

Source:

Indices of Leading, Lagging, and Coincident Indicators

Institute for Supply Management Indices
Evolution of an Indicator
Digging for the Data
What Does It All Mean?

114

ISM Employment Index

117

ISM Price Index

119

ISM Supplier Deliveries Index

121

ISM Non-Manufacturing Indices

123

Manufacturers’ Shipments, Inventories, and Orders

131
132
134

Durable Goods Report

134

Factory Orders Report

138

What Does It All Mean?

139

Durable Goods Report

139

Factory Orders Report

142

Manufacturing and Trade Inventories and Sales
Evolution of an Indicator
Digging for the Data
What Does It All Mean?
Inventories and the Business Cycle
Inventories-to-Sales Ratios

How to Use What You See
Trick From the Trenches

Source:

125
126

Evolution of an Indicator
Digging for the Data

How to Use What You See
Tricks From the Trenches

108
109
113

PMI

How to Use What You See
Tricks From the Trenches

107

143
144

147
149
149
151
153
154

155
156

Evolution of an Indicator
Digging for the Data
What Does It All Mean?

164

Regional Differences

166

Housing and the Business Cycle

167

Single-Family Housing Starts

Conference Board Consumer Confidence and
University of Michigan Consumer Sentiment Indices
Evolution of an Indicator
Digging for the Data
What Does It All Mean?
The Expectation Indices
Confidence and Durables Spending

How to Use What You See

169

171
172

175
177
177
179
182
184

185

Employment and Sentiment

186

Noneconomic Influences on Sentiment

186

Tricks From the Trenches

160
161
163

Influences on Residential Construction

How to Use What You See
Tricks From the Trenches

159

Advance Monthly Sales for Retail Trade and
Food Services
Evolution of an Indicator
Digging for the Data
Surging Subcategories: Superstores and E-Commerce

What Does It All Mean?

189

193
194
195
198

200

Total Retail and Food Service Sales, Nominal and Real Figures 200
Total Sales Excluding Motor Vehicles and Parts

201

GAFO

202

How to Use What You See
Same-Store Sales
Seasonality

Tricks From the Trenches

204
205
206

207

Source:

New Residential Construction

Personal Income and Outlays
Evolution of an Indicator
Digging for the Data

210
211

Personal Income

213

Personal Consumption Expenditures

213

Personal Savings

215

What Does It All Mean?

215

Personal Income

216

Consumer Spending

219

Personal Savings Rate

220

How to Use What You See
Tricks From the Trenches

222
223

Consumer and Producer Price Indices
Evolution of an Indicator
Producer Price Index
Consumer Price Index

Digging for the Data

229
231
231
232

233

Consumer Price Index Data Sources

233

Producer Price Index Data Sources

235

Calculating the Inflation Rate

What Does It All Mean?

237

238

Price Trends

238

Price Indices and the Markets

240

Price Indices and the Business Cycle

How to Use What You See
Tricks From the Trenches

Source:

209

241

242
243

References

245

Index

259

This page intentionally left blank

Acknowledgments

his project could never have been completed without
the support and assistance of my wife, Suzie. She has helped
in countless and immeasurable ways. Families play an integral
role in dreams, aspirations, and accomplishments. My family has
paved a clear path for any and every ambition that I could possibly
have. They are indubitably responsible for all of my successes. My
mother and father have always instilled the importance of education, hard work, free thinking, and discipline. For that I could never
thank them enough. A special thanks to my brother Robert and to
my in-laws, Richard and Nancy McCabe.
Educators can have a profound influence on one’s life. My life, as
well as this project, was no exception in the way of encouragement
and wisdom disseminated by those in academia. Professors and
mentors have helped, not only in the understanding of some of the
roles that economic indicators and statistics assume, but as counselors to a not-so-quick-to-learn student. They include, and are not
limited to, David W. Ring, William O’Dea, Robert Carson, and
Thomas Gergel of the State University of New York at Oneonta.
Prof. Ring taught me to work hard, Prof. Carson to look at each
situation from alternative perspectives, and Prof. Gergel to have a
passion about whatever task I might take on. I practice these three
lessons every day of my life.
The list of professional associates who have helped me with this
project could easily take up the entire book, but some individual
recognition is essential. In my eighteen-plus years of work experience on Wall Street I have never been associated with a more
professional outfit than Bloomberg News. To the scores of friends
and acquaintances at the Bloomberg offices all around the world,
thanks for all your help and input into this project. You are unaware
xiii

xiv



Acknowledgments

of how helpful you have been. I especially wish to thank Vinny Del
Giudice, Vince Golle, Yvette Fernandez, Monee Fields-White,
Jackie Jozefek, and Al Yoon.
Undoubtedly the greatest gratitude with respect to the creation
of this work has to go to my editors at Bloomberg Press, namely
Kathleen Peterson, Chris Miles, Tracy Tait, and mostly, Betsy
Ungar. Betsy’s extraordinary talents have transformed my muddled
manuscript into a more readable and effective publication.
Many thanks to the analysts at Argus Research Corp., which
help me each day with insight into their respective industries,
particularly, Wendy Abramowitz, Robert Becker, Kevin Calabrese,
Marie Driscoll, Jeffrey Gildersleeve, Gary F. Hovis, Jim Kelleher,
David Kerans, and David Ritter. I must tip my hat in appreciation
to John Eade, Argus Research’s CEO and Director of Research, as
well as the Dorsey family for the opportunity to work at the most
prestigious research institution on Wall Street.
Acknowledgment wouldn’t be complete without special thanks
to Charles Gilbert and Michele Johnson (Board of Governors
of the Federal Reserve System); Lynn Franco (The Conference
Board); Richard Deitz (Federal Reserve Bank of New York); Guhan
Venkatu (Federal Reserve Bank of Cleveland); Scott Scheleur (Service Sector Statistics Division, U.S. Census Bureau); Kristen Kioa
(Institute for Supply Management); Jeannine Aversa and Marty
Crutsinger (Associated Press); Garrett Bekker (Kim & Co.); Steve
Berman (U.S. Department of Commerce, Bureau of Census);
George Hager, Barbara Hagenbaugh, Sue Kirchhoff (USA Today);
Jason Hecht (Ramapo College); Robert Bricken (Prebon, Yamane
(USA) Inc.); David Jozefek (Morgan Stanley); Thomas Feeney
(Shippensburg University); Jeffrey J. Junior (Aries Appraisal Group,
Inc.); and Joe Pregiato (Arbor & Ivy).
Any errors or oversights that may exist in this book were not intentional and are not the fault of any of those individuals named above.

Introduction

nvesting without understanding the economy is like taking a trip without knowing anything about the climate of your
destination. Inclement weather can wreak havoc with a vacation,
especially if it involves outdoor activities. Just so, putting hardearned money into the stock or bond market when economic conditions are unfavorable can destroy financial plans for a comfortable
retirement, a new house, or a child’s college education.
No one understands this better than Wall Street investment
banks, brokers, and research institutions. All of these have adopted
a top-down approach to securities analysis that begins with a forecast of the general economic climate, including interest rate projections, currency forecasts, and estimates of domestic and foreign
economic growth. In this, they are following one of the precepts
laid down by Benjamin Graham and David Dodd in their 1940
investors’ bible, Security Analysis: “Economic forecasts provide
essential underpinning for stock and bond market, industry, and
company projections.”
You don’t need to manage millions or billions of dollars, however, to study economic conditions and plan your investment strategy accordingly. You can get much of the same information that
Wall Street professionals use in their analyses from the business
sections of the nation’s newspapers, magazines, and evening news
programs. Furthermore, you don’t need a degree in economics or
mathematics to interpret this information. In fact, many graduates
of such programs at the nation’s top universities find themselves entirely unprepared for the real world of finance. This book attempts
to bridge the wide gap between the sometimes mind-numbing
theories of textbook economics—the principles that are taught on
college campuses across the country—and the everyday world of
1



The Trader’s Guide to Key Economic Indicators

Wall Street. It does so by focusing on a dozen economic indicators
that are among the most important of any analyst’s or economist’s
tools. Understanding these indicators will make the study of economics more palatable and exciting.
Over the past century, thousands of economic indicators have
emerged, predicting everything from the demand for gasoline to
the size of harvests. Some are more fun than functional, such as
those claiming links between stock performance for the year and
which conference, the NFC or the AFC, wins the Super Bowl,
or whether women’s hemlines rise to midthigh or fall to midcalf.
Others indicators are more serious, solidly based in economic
observations. These range from the arcane—such as the indicator
connecting the production level of titanium dioxide, an ingredient
of pigments used in paints and plastics, with the demand for building materials—to the commonsensical. The price of copper, used
in wiring and many other construction elements, for instance, has a
clear relationship with the pace of housing activity. The same could
be said of economic growth and railroad car loadings, shipping container production, wooden pallet shipments, and the manufacture
of corrugated boxboard and packaging, all of which are connected
with transporting freight or manufactured goods.
Over time, economists have weeded out the least successful
indicators, based on the most dubious relationships, to arrive at a
core of about fifty consistently reliable ones. This book presents
the dozen that are must-haves in any analytical toolbox. Virtually all
Wall Street economists use these indicators in the analyses and their
writings. Federal Reserve officials conduct monetary policy with
respect to the trends that these indicators project. They are also
considered “must haves” in the sense that they are among the most
accurate at depicting economic relationships as well as attendant
market-movability. That is, each of these indicators at one time or
another typically figures among the top-tier factors to engender big
swings in the financial markets.
Some of the dozen indicators discussed are constructed by U.S.
government agencies such as the U.S. Department of Commerce’s
Census Bureau, the U.S. Department of Labor, and the Board of

Source:

Introduction



Governors of the Federal Reserve. Others are the products of private organizations such as the Institute for Supply Management,
the Conference Board, and the University of Michigan. Some have
excellent predictive powers. Others reflect principally the current
state of the economy, and still others highlight industries that might
outperform and so help identify the likely path of economic activity. All have one thing in common, however: In one way or another,
they all relate to the business cycle.

Source:

THE BUSINESS CYCLE
The business cycle is one of the central concepts in modern economics. It was defined by celebrated economists Arthur Burns and
Wesley Mitchell in their pioneering 1946 study, Measuring Business Cycles, written for the National Bureau of Economic Research
(NBER), which today is the official arbiter of the U.S. business
cycle. According to Burns and Mitchell, the business cycle is “a
type of fluctuation found in the aggregate economic activity of
nations that organize their work mainly in business enterprises: a
cycle consists of expansions occurring at about the same time in
many economic activities, followed by similarly general recessions,
contractions, and revivals, which merge into the expansion phase of
the next cycle.”
No two business cycles are the same. As illustrated in FIGURE I-1,
during the relatively short time that people have been measuring the
U.S. economy, the length of expansions, from economic trough to
peak, and of contractions, from peak to trough, have varied widely—
although the former, especially recently, have generally been longer and steadier than the latter. Expansions have ranged from 120
months (April 1991 to April 2001) to 10 (March 1919 to January
1920), and downturns from 43 months (September 1929 to March
1933) to 6 (February 1980 to July 1980). The amplitude of the peaks
and troughs has also differed significantly from cycle to cycle.
One way to think of the business cycle is as a graphical representation of the total economic activity of a country. Because the
accepted benchmark for economic activity in the United States



The Trader’s Guide to Key Economic Indicators

Figure I-1

U.S. Business Cycle Durations

Business Cycle Reference Dates
Peak

Duration in Months

Trough

Quarterly dates
are in parentheses

Expansion

Peak
to trough

Previous trough
to this peak

Cycle

—
18
8
32
18
65

—
30
22
46
18
34

—
48
30
78
36
99

—
—
40
54
50
52

Trough from
Peak from
previous trough previous peak

June 1857 (II)
October 1860 (III)
April 1865 (I)
June 1869 (II)
October 1873 (III)

December 1854
December 1858
June 1861 (III)
December 1867
December 1870
March 1879 (I)

March 1882 (I)
March 1887 (II)
July 1890 (III)
January 1893 (I)
December 1895 (IV)

May 1885 (II)
April 1888 (I)
May 1891 (II)
June 1894 (II)
June 1897 (II)

38
13
10
17
18

36
22
27
20
18

74
35
37
37
36

101
60
40
30
35

June 1899 (III)
September 1902 (IV)
May 1907 (II)
January 1910 (I)
January 1913 (I)

December 1900 (IV)
August 1904 (III)
June 1908 (II)
January 1912 (IV)
December 1914 (IV)

18
23
13
24
23

24
21
33
19
12

42
44
46
43
35

42
39
56
32
36

August 1918 (III)
January 1920 (I)
May 1923 (II)
October 1926 (III)
August 1929 (III)

March 1919 (I)
July 1921 (III)
July 1924 (III)
November 1927 (IV)
March 1933 (I)

7
18
14
13
43

44
10
22
27
21

51
28
36
40
64

67
17
40
41
34

May 1937 (II)
February 1945 (I)
November 1948 (IV)
July 1953 (II)
August 1957 (III)

June 1938 (II)
October 1945 (IV)
October 1949 (IV)
May 1954 (II)
April 1958 (II)

13
8
11
10
8

50
80
37
45
39

63
88
48
55
47

93
93
45
56
49

April 1960 (II)
December 1969 (IV)
November 1973 (IV)
January 1980 (I)
July 1981 (III)

February 1961 (I)
November 1970 (IV)
March 1975 (I)
July 1980 (III)
November 1982 (IV)

10
11
16
6
16

24
106
36
58
12

34
117
52
64
28

32
116
47
74
18

July 1990 (III)
March 2001 (I)

March 1991(I)
November 2001 (IV)

8
8

92
120

100
128

108
128

17
22
18
10

38
27
35
57

55
48
53
67

56*
49†
53
67

18
22
20
10

33
24
26
52

51
46
46
63

52‡
47§
45
63

Average, all cycles:
1854–1991 (32 cycles)
1854–1919 (16 cycles)
1919–1945 (6 cycles)
1945–1991 (10 cycles)
Average, peacetime cycles:
1854–1991 (27 cycles)
1854–1919 (14 cycles)
1919–1945 (5 cycles)
1945–1991 (8 cycles)

(IV)
(IV)

Contraction

(I)
(IV)

*31 cycles, †5 cycles, ‡26 cycles, §13 cycles

Figures printed in bold italic are the wartime expansions (Civil War, World Wars I and II, Korean War, and
Vietnam War), the wartime contractions, and the full cycles that include the wartime expansions.

Source: NBER

Introduction



Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; NBER

Figure I-2 GDP and Highlighted Recessions
YOY %

10
Shaded areas = Recession

8
6
4
2
0
–2
–4
1953

1963

1973

1983

1993

2003

is currently gross domestic product (GDP), economists generally identify the business cycle with the alternating increases and
declines in GDP. Rising GDP marks economic expansion; falling
GDP, a contraction (see FIGURE I-2). That said, the business cycle, as
defined by Burns and Mitchell, can’t be fully captured by one indicator, even the GDP. Rather, it is a compendium of indicators that
reflects various aspects of the economy.
Economic indicators are classified according to how they relate
to the business cycle. Those that reflect the current state of the
economy are coincident; those that predict future conditions are
leading; and those that confirm that a turning occurred are lagging.

Source:

INDICATORS AND THE MARKETS
The organization responsible for an indicator generally distributes
its report about an hour before the official release time to financial
news outlets such as Bloomberg News, Dow Jones Newswires,
Reuters, CNBC, and CNNfN. The reporters, who are literally
locked in a room and not permitted to have contact with anyone
outside, ask questions of agency officials and prepare headlines and
analyses of the report contents. These stories are embargoed until



The Trader’s Guide to Key Economic Indicators

the official release, at which time they are transmitted over the
newswires to be dissected by the Wall Street community. Most Wall
Street firms employ economists to provide live broadcasts of the
numbers as they run across the newswires, together with interpretation and commentary regarding the likely market reaction. This
task, known as the “hoot-and-holler” or tape reading, is among the
most stressful performed by an economist. One slip-up can cost a
trader or entire trading floor millions of dollars.
The more an indicator deviates from Street expectations, the
greater its effect on the financial markets. A 0.1 percent decline
in retail sales, for example, might not move the markets much if
economists were looking for a flat reading or a 0.1 percent rise.
But if the consensus were for an increase of 0.7 percent, and
instead the 0.1 percent decline hit the tape, the markets might well
be rocked. That said, it is always prudent for traders and other market participants to keep apprised of what the Street expectations are
for key economic indicators such as those covered here.

You’ve no doubt read in a paper or heard on TV or the radio forecasts of economic expansion or recession. You also probably realize that the one is desirable and the other is not. But you may not
know how the economists quoted came up with their predictions.
Without this knowledge, how can you judge how well considered or
rash they are—and whether to trust them in creating your investment strategy? This book seeks to help you form your own opinions
about the possible direction of the economy and the markets and to
decide how to act based on those opinions.
Each chapter corresponds to an indicator, beginning with
the most comprehensive—the GDP and indices of leading, lagging, and coincident indicators—and continuing with those tied
to particular aspects or segments of the overall economy, such as
consumer prices, manufacturing, housing, and retail sales. Every
chapter contains four principal sections: an introduction sketching
out the major attributes of the indicator and its effect on the mar-

Source:

HOW TO USE THIS BOOK

Source:

Introduction



kets; a discussion of its origins and development; a description of
how the relevant data are obtained, analyzed, and presented; and an
explanation of how to incorporate these data into your investment
process. The last section also contains at least one “trick”—involving either a little-known subcomponent of the indicator or a combination of subcomponents—that Wall Street economists use to get
a clearer or more timely picture of business activity. At the end of
the book is a listing of additional reading and resources, organized
by chapter, pointing those interested to references that discuss the
relevant indicator in greater detail.
In putting what you learn from this book into practice, you might
take some pointers from Wall Street. Just about every investment firm
has a pre-market-opening meeting in which the day’s events and potential trading strategies are presented. This always includes a discussion
of the economic indicators scheduled for release that day. No trader
wants to be caught off guard by an unexpected market-moving release. For the same reason, many traders have on their desks calendars
showing which economic release is scheduled for a particular day and
indicating both the value or percentage change of the previous report
and the Street’s estimates—highest, lowest, and consensus—for the
upcoming one. That way, when the actual figure is released, they will
know how it compares with expectations and can react accordingly.
Of course, no single economic indicator will tell you all you
need to know about the current or future economic climate. Each
has drawbacks and may send false signals because of unforeseen
shocks, faulty measurements, or suspect collection processes. Piecing together the information from all twelve indicators discussed in
this book like tiles in a mosaic will give you a dynamic representation of the economy. But if you are truly serious about understanding the macroeconomic climate and individual industry conditions,
you should also take advantage of the Securities and Exchange
Commission’s Regulation Fair Disclosure of 2000, which mandates
for individual investors the same access to companies’ quarterly
earnings conference calls that professional analysts have.
These calls provide a great deal of insight into corporate spending plans, manufacturing and production activity, international



The Trader’s Guide to Key Economic Indicators

conditions, pricing, and the general business climate. Especially
informative are the announcements of industrial behemoths such
as Alcoa, Boeing, Caterpillar, Cummins, Emerson Electric, Ford
Motor Company, General Electric, Illinois Tool Works, Johnson
Controls, and United Technologies. Many companies also offer slide
presentations, handouts, and supplemental data with these quarterly
presentations, which often provide even greater detail on their buying intentions, prospective employment changes, and any threats to
performance that they foresee. There’s no cheaper and easier way
to gather anecdotal evidence about business conditions. If you can’t
listen in, the presentations are almost always archived on company
websites, from which they may be readily retrieved 24/7.

This book was written primarily for those traders and investors
lacking a formal introduction to the most popular economic indicators on Wall Street. Just because an individual is entrusted with investing millions of dollars does not guarantee a practical command
of economic indicators and their meaning for investment. When
newly minted MBAs arrive on the trading floors of financial firms,
for example, few are equipped with a complete appreciation of these
indicators—no matter from which institution that degree has come.
My years of experience on a few of the largest trading floors in the
world has suggested the need to fill what can be viewed as a surprisingly expansive void regarding indicators, statistics, the economic
meaning of the associated figures, and the market’s likely reaction.
Those new to the field of investing and economics, including
students of the subject, also should benefit from the fundamental,
application-oriented nature of this book. As most academics know,
if students cannot see the results or directly test theories with practical data, the knowledge they hold tends to remain more theoretical
than real-world and they eventually may lose interest in the field. It
is here that many future economists are lost. As exercises within an
imperfect “science,” experiments conducted in the social discipline
of economics are predominantly theorized or hypothesized and

Source:

WHO CAN BENEFIT FROM THIS BOOK?

Introduction



Source:

seldom tested with tangible data. In this sense, economists are not
as fortunate as physicists or natural scientists, who conduct experiments in a controlled environment such as a laboratory, riverbed, or
ocean. The economic indicators contained in these chapters serve
as concrete guideposts within the discipline of economics, and as
such make experimentation, testing, and study for investments not
only possible but understandable.

This page intentionally left blank

Gross Domestic Product

conomics has received a bad rap. In the mid-nineteenth
century, the great Scottish historian Thomas Carlyle dubbed
this discipline “the dismal science,” and jokes about economists
being more boring than accountants abound on the Street. But
truth be told, there is nothing more exciting than watching the
newswire on a trading floor of a money-center bank minutes ahead
of the release of a major market-moving economic report. One of
the top excitement generators is the report on gross domestic product (GDP)—an indicator that is a combination of economics and
accounting.
Economists, policymakers, and politicians revere GDP above all
other economic statistics because it is the broadest, most comprehensive barometer available of a country’s overall economic condition. GDP is the sum of the market values of all final goods and
services produced in a country (that is, domestically) during a specific period using that country’s resources, regardless of the ownership of the resources. For example, all the automobiles made in the
United States are included in the GDP—even those manufactured
in U.S. plants owned by Germany’s DaimlerChrysler and Japan’s
Lexus. In contrast, gross national product (GNP) is the sum of the
market values of all final goods and services produced by a country’s
permanent residents and firms regardless of their location—that is,
whether the production occurs domestically or abroad—during a
given period. Baked goods produced in Canada by U.S. conglomerate Sarah Lee, for example, are included in the United States’ GNP,
but not its GDP.
11



The Trader’s Guide to Key Economic Indicators

GDP is a more relevant measure of U.S. economic conditions
than GNP, because the resources that are utilized in the production process are predominantly domestic. There are strong parallels
between the GDP data and other U.S. economic indicators, such as
industrial production and the Conference Board’s Index of Coincidental Indicators, which will be explored in later chapters.
The GDP is calculated and reported on a quarterly basis as
part of the National Income and Product Accounts. The NIPAs,
which were developed and are maintained today by the Commerce
Department’s Bureau of Economic Analysis (BEA), are the most
comprehensive set of data available regarding U.S. national output,
production, and the distribution of income. Each GDP report contains data on the following:
 personal income and consumption expenditures
 corporate profits
 national income
 inflation

Source:

These data tell the story of how the economy performed—
whether it expanded or contracted—during a specific period,
usually the preceding quarter. By looking at changes in the
GDP’s components and subcomponents and comparing these
with changes that have occurred in the past, economists can draw
inferences about the direction the economy might take in the
future.
Of all the tasks market economists perform, generating a forecast for overall economic performance as measured by the GDP
data is the one to which they dedicate the most time. In fact, the
latest report on GDP is within arm’s reach of most Wall Street
economists. Because several departments in a trading institution
rely on the economist’s forecasts, this indicator has emerged as the
foundation for all research and trading activity and usually sets the
tone of all of Wall Street’s financial prognostications.

Gross Domestic Product



Source:

EVOLUTION OF AN INDICATOR
Measuring a nation’s output and performance is known formally
as national income accounting. This process was largely pioneered
by Simon Kuznets, an economist hired by the U.S. Department
of Commerce in the 1930s—with additional funding from the
National Bureau of Economic Research—to create an accurate
representation of how much the U.S. economy was producing. Up
to that time, there was no government agency calculating this most
critical of economic statistics.
The initial national income estimates produced by Kuznets in
1934 were representations of income produced, measures of the national economy’s net product, and the national income “paid out,”
or the total compensation for the work performed in the production
of net product. At that time, no in-depth breakdown of components
yet existed. In fact, Kuznets didn’t even have a detailed representation of national consumption expenditures. This was the first step
of several in the creation of a formal method of national income
accounting, and yet was still a far cry from today’s highly detailed
representation of the macroeconomy.
The result was the National Income and Product Accounts.
In addition to this immense task, Kuznets reconstructed the national income accounts of the United States back to 1869. (He
was awarded a Nobel Prize in Economics in 1971 in part for this
accomplishment.) Kuznets’s first research report, presented to
Congress in 1937, covered national income and output from 1929
through 1935.
In 1947, the first formal presentation of the national income
accounts appeared as a supplement to the July issue of the Survey
of Current Business. This supplement contained annual data from
1929 to 1946 disseminated in thirty-seven tables. These data were
separated into six accounts:
1. national income and product account
2. income and product account for the business sector
3. government receipt/expenditure account
4. foreign account



The Trader’s Guide to Key Economic Indicators

5. personal income/expenditure account
6. gross savings and investment account
Before the creation of the NIPAs, households, investors, government policymakers, corporations, and economists had little or
no information about the complete macroeconomic picture. There
were indices regarding production of raw materials and commodities. There were statistics on prices and government spending. But
a comprehensive representation of total economic activity wasn’t
available. In fact, the term macroeconomy didn’t appear in print until
1939. Policymaking without knowing the past performance of the
economy, how it operated under different conditions and scenarios,
or which sectors were weak and which were strong was a daunting
task. This may have been the reason for many of the economicpolicy failures of the early twentieth century.
Many economists have laid the blame for the Great Depression
of the 1930s on the Federal Reserve’s failure to respond to the ebullient activity during the Roaring Twenties (sound familiar?). The
Fed may bear much of the responsibility; but very few, if any, have
defended the Federal Reserve’s failures on the grounds of insufficient information. The Great Depression forced the government
to develop some sort of national accounting method. World War II
furthered the government’s need to understand the nation’s capacity, the composition of its output, and the general economic state of
affairs. How could the government possibly plan for war without an
accurate appreciation of its resources? The NIPAs permit policymakers to formulate reasonable objectives such as higher economic
growth rates or lower inflation rates as well as to formulate policies
to attain these objectives and steer the economy around any roadblocks that might impede the attainment of these goals.

Tracking the developments in an economy as large and dynamic as
that of the United States is not easy. But through constant revision
and upgrading, a relatively small group of dedicated economists at

Source:

DIGGING FOR THE DATA

Gross Domestic Product



the BEA accomplishes this huge task every quarter. Each quarterly
report of economic activity goes through three versions, all available
on the BEA website, www.bea.gov. The first, the advance report,
comes one month after the end of the quarter covered, hitting the
newswires at 8:30 a.m. ET. So, the GDP report pertaining to the
first three months of the year is released sometime during the last
week of April, the second quarter’s advance report during the last
week of July, the third quarter’s in October, and the fourth quarter’s
during the last week of January of the following year. Because not
all the data are available during this initial release, the BEA must
estimate some series, particularly those involving inventories and
foreign trade.
As new data become available, the BEA makes the necessary
refinements, deriving a more accurate estimate for GDP. The
second release, called the preliminary report, comes two months
after the quarter covered, one month after the advance report, and
reflects the refinements made to date. The last revision to the data
is contained in the final report, which is released three months after
the relevant quarter and a month after the preliminary report. The
release dates for 2003 are shown in FIGURE 1-1.
Annual revisions are calculated during July of every year, based
on data that become available to the BEA only on an annual basis,
such as state and local government consumption expenditures. The
BEA estimates these data on a quarterly basis via a judgmental trend
based on annual surveys of state and local governments. Judgmental
trends are quarterly interpolations of source data that are only available on an annual basis. Because the surveys are available on an annual basis, estimates can only be made during the annual revision.
Figure 1-1

2003 Release Schedule for GDP Reports

Advance report
Preliminary report
(1st revision)
Final report
(2nd revision)

2002: QIV

2003: QI

2003: QIII

January 30
February 28

April 25
May 29

July 31
August 28

October 30
November 25

March 27

June 26

September 26

December 23

Source: U.S. Department of Commerce, Bureau of Economic Analysis



The Trader’s Guide to Key Economic Indicators

As source data for the components of the accounts are continuously updated and revised, the components of the NIPAs must be
updated to reflect these revisions. That’s the primary function of
the annual revision. Each of the three years (twelve quarters) worth
of data is subject to revision during this annual updating. Every
five years the BEA issues a so-called benchmark revision of all of
the data in the NIPAs. This typically has resulted in considerable
changes to the five years of quarterly figures.
Benchmark revisions are different from annual revisions in that
they generally contain major overhauls to the structure of the report, usually definitional, re-classifications, and new presentations
of data. New tables need to be created to account for products that
are developed. As the economy evolves, new goods and services
come to market and therefore need to be accounted for. Obviously,
there were times when CDs, microwave ovens, MP3 players, and
DVDs didn’t exist. Because the U.S. economy develops and produces these goods, there needs to be a place for this production
to be recorded. All of the data—quarterly and annual—are revised
during benchmark revisions.

As noted previously, the GDP is the sum of the market values of all
final goods and services produced by the resources (labor and property) of a country residing in that country. This definition contains
two particularly important terms: final and produced. When economists refer to final goods, they mean those goods produced for their
final intended use, that is, as end products, not as component or
intermediate parts in another stage of manufacture. As an example,
consider that each year, Goodyear Tire & Rubber produces some
hundred million tires. Most are produced for use on new vehicles.
But there are still quite a number created for distribution in retail
and wholesale stores as replacements and spares. Those tires produced and delivered to automakers intended for use on new automobiles are not counted as production because we do not calculate
the value of automobiles in the national accounts by summing the

Source:

SOME DEFINITIONS

Gross Domestic Product



value of its components. In other words, we don’t add the cost of
the radio, the seats, the heating elements, the spark plugs, and so
on. We only count the value of the final product, the automobile.
Obviously, the economists at the BEA would make a serious miscalculation if they counted all the tires sold by the manufacturer to
Wal-Mart and Sears, as well as those sold by the automakers as part
of their automobiles. The same holds true for the production of
wool. BEA economists only count the wool purchased for final use.
Because countless final uses exist for wool—sweaters, hats, blankets,
and so on—the BEA would make the same double-counting error
by adding the production of raw wool as well as the wool used in
sweaters, blankets, and the like.
Let’s consider the other important term, produced. Resales are
not included in the accounts. Rightfully so, the BEA has determined
that because the pace of reselling is not indicative of the current
pace of production, it shouldn’t be included in the output figures.
Another segment of the economy that the BEA excludes from
the GDP release is the activity that goes on “off the books.” This
seems an obvious exclusion, but it’s a big one. Believe it or not, some
of the most conservative studies have set the size of the U.S. underground economy at around 10 percent of the official U.S. GDP, or
what was roughly $1 trillion in the first quarter of 2003. The BEA
doesn’t count or make any adjustments for non-state-sanctioned
gambling, prostitution, trade in illegal drugs, fraud, the production
and sale of counterfeit merchandise, and the like because, officially,
they don’t exist—wink, wink, nudge, nudge. These activities aren’t
reported, so how can they be measured? Clandestine activity like
this can understandably alter the estimate of several economic indicators, but none more than the GDP.

Source:

GDP VERSUS GNP
The NIPAs contain figures for both gross domestic product and
gross national product. Before 1991, GNP was the benchmark for
all economic activity in commentaries, reports, articles, and texts.
The GDP became the official barometer when the BEA decided



The Trader’s Guide to Key Economic Indicators

Figure 1-2

GNP Derived From GDP (QIV 2002 Report)

U.S. GDP
plus receipts from the rest of the world
minus payments to the rest of the world
equals U.S. GNP

$10.588 trillion
+ $284.200 billion
– $293.400 billion
= $10.579 trillion

that the measure was a better fit with the United Nations System of
National Accounts used by other nations, and so made international
comparisons of economic growth easier.
GDP differs from GNP in what economists call “net factor
income from foreign sources”: the difference between the value
of receipts from foreign sources and the payments made to foreign
sources. The table in FIGURE 1-2, using data from the final GDP
report of the fourth quarter of 2002, illustrates how the BEA quantifies this relationship in its GDP report.
The difference between the value of GDP and GNP is typically
minuscule, usually less than 0.5 percent. In Figure 1-2, for example,
GDP is approximately $10.588 trillion and GNP $10.579 trillion, a
difference of under $10 billion, or 0.09 percent.
CALCULATING GDP: THE AGGREGATE EXPENDITURE APPROACH
Every transaction in an economy involves two parties, a buyer and
a seller. To calculate total economic activity, economists can focus
either on the buyers’ actions, adding together all the expenditures
on goods and services, or on the sellers’ actions, tallying the total
income received by those employed in the production process.
These two approaches correspond to the two methods of calculating the GDP: the aggregate expenditure method, which is the more
popular and the one used on most Wall Street trading floors, and
the income approach. The totals reached by both measures should
theoretically be the same. In practice, however, there are small differences.
To calculate GDP, the BEA uses the aggregate expenditure
equation:
GDP = C + I + G + (X – M )

Source: U.S. Department of Commerce, Bureau of Economic Analysis

Gross Domestic Product



where C is personal consumption expenditures, I is gross private
domestic investment, G is government consumption expenditures
and gross investment, and (X – M ) is the net export value of goods
and services (exports minus imports). The identity expressed in
this equation is probably the most widely cited of all economic
relationships and appears in virtually all introductory macroeconomic texts.
Because the U.S. economy is extremely dynamic and susceptible
to sudden and unforeseen influences like inclement weather and
war, the percentage of GDP contributed by each of the equation’s
components varies over time, even from quarter to quarter. For the
most part, though, the proportions don’t deviate significantly from
those represented in FIGURE 1-3, which depicts the composition of
first quarter 2003 GDP.
Personal consumption expenditures (also referred to as consumer spending or simply spending) are the largest component
of GDP, accounting for roughly two-thirds of total economic output. During the first quarter of 2003, consumer spending climbed
to approximately 70 percent of GDP ($7.503 trillion divided by
$10.698 trillion).
Consumer spending is the total market value of household

Source: U.S. Department of Commerce, Bureau of Economic
Analysis

Figure 1-3 Composition of GDP
Net exports –4.54%

Government 19.20%

Investment 15.21%

Consumption 70.14%



The Trader’s Guide to Key Economic Indicators

purchases during the accounting term, including items such as
beer, telephone service, golf clubs, CDs, gasoline, musical instruments, and taxicab rides. As shown in the table in FIGURE 1-4, these
items fall into three categories: durable goods, nondurable goods,
and services. Durable goods are those with shelf lives of three or
more years. Examples include automobiles, refrigerators, washing
machines, televisions, and other big-ticket items, such as jewelry,
sporting equipment, and guns. Nondurable goods are food, clothing and shoes, energy products such as gasoline and fuel oil, and
other items such as tobacco, cosmetics, prescription drugs, magazines, and sundries. Services include housing, household operation,
transportation, medical care, and recreation, as well as hair styling,
dry cleaning, funeral services, legal services, and education.
Services constitute by far the largest category of consumer
purchases. They account today for roughly 59 percent of all consumer spending, up from a mere third in 1950. No wonder the
Consumer Spending Breakdown
2003: QI
($ in billions)

Consumer Spending

Percent of
Total Spending

$7,503

100.00%

$863
$366
$317
$180

11.50%
4.88%
4.23%
2.40%

Nondurable Goods
Food
Clothing and shoes
Gasoline, fuel oil, and other energy goods
Other

$2,197
$1,059
$327
$210
$601

29.28%
14.11%
4.35%
2.80%
8.01%

Services
Housing
Household operation
Transportation
Medical care
Recreation
Other

$4,443
$1,102
$419
$278
$1,193
$293
$1,159

59.21%
14.69%
5.58%
3.71%
15.89%
3.90%
15.44%

Durable Goods
Motor vehicles and parts
Furniture and household equipment
Other

Source: U.S. Department of Commerce, Bureau of Economic Analysis

Figure 1-4

Source:

Gross Domestic Product



United States is said to have a service-based economy. Spending on
nondurable goods is the second-largest category of expenditures,
representing about 29 percent of the total. Durable goods expenditures, the most volatile component, account for the remaining
11 percent.
A more detailed summary of personal consumption expenditures
is available on a monthly basis in the BEA’s report on Personal
Income and Outlays, which is the direct source of data for this
component of the GDP report. Personal income and outlays are
discussed in Chapter 11.
Gross private domestic investment encompasses spending by
businesses—on equipment such as computers, on the construction
of factories and production plants, and in mining operations—
expenditures on residential housing and apartments; and inventories. Inventories, which consist of the goods businesses produce
during a period that remain unsold, are valued by the BEA at the
prevailing market price. This value fluctuates greatly from quarter
to quarter, making the level of gross private domestic investment
quite volatile. Accordingly, economists often look at fixed investment—gross private domestic investment minus inventories. This,
in turn, has two major components, residential and nonresidential.
The latter, which is also referred to as capital spending, includes
expenditures on computers and peripheral equipment, industrial
equipment, software, and nonresidential buildings such as plants
and factories. The former comprises spending on the construction
of new houses and apartment buildings and on related equipment.
Even without the volatile influence of inventories, investment
spending is prone to extreme movements, because most of this
activity is linked to the ever-changing interest rate environment.
Gross private domestic investment usually accounts for 15 percent
of GDP. During the first quarter of 2003, it represented 15.2 percent ($1.627 trillion divided by $10.698 trillion) of GDP.
Government consumption expenditures and gross investment covers all the money laid out by federal, state, and local
governments for goods (both durable and nondurable) and services,
for both military and nonmilitary purposes. The category includes



The Trader’s Guide to Key Economic Indicators

spending on building and maintaining toll bridges, libraries, parks,
highways, and federal office buildings; on compensation for government employees; on research and development, spare parts,
food, clothing, ammunition; and on travel, rents, and utilities.
Government expenditures and investment usually account for 20
percent of total GDP. During the first quarter of 2003, government
consumption expenditures and gross investment accounted for
19.2 percent of total economic activity ($2.054 trillion divided by
$10.698 trillion).
Net exports of goods and services, the last component in the
equation, is simply the difference between the dollar value of the
goods and services the United States sends abroad (exports) and the
dollar value of those it takes in across its borders (imports). Because
the country generally imports more than it exports, this figure is usually negative, thus acting as a drag on economic growth. During the
first quarter of 2003, net exports subtracted 4.5 percent from total
economic activity (–$485.7 billion divided by $10.698 trillion).
NOMINAL AND REAL NUMBERS
The data reported in the GDP release are presented in two forms,
nominal and real. Nominal, also known as current dollar, GDP is
the total value, at current prices, of all final goods and services produced during the reporting period. Real, or constant dollar, GDP
is the value of these goods and services using the prices in effect in a
specified base year. Economists tend to prefer the real to the nominal measure. To understand why, consider a country that produces
only two goods—pencils and vodka, a very interesting economy. If
during Year One, it sells two thousand pencils at $0.10 each and one
thousand bottles of vodka at $5.00 a bottle (cheap vodka), its nominal GDP will be $5,200:
2,000 x $0.10
1,000 x $5.00

=
=

$ 200
$5,000
$5,200
Source:

Pencils
Vodka
Nominal GDP

Gross Domestic Product



Next year, the same country produces only a thousand pencils
and five hundred bottles of vodka but doubles its selling prices,
to $0.20 a pencil and $10.00 a bottle. Its nominal GDP is again
$5,200:
Pencils
Vodka
Nominal GDP

1,000 x $ 0.20
500 x $10.00

=
=

$ 200
$5,000
$5,200

Is the economy larger during the second year? Did it produce
the same amount? The difficulty in answering these questions illustrates the problem with nominal values. Economists have no way
of telling whether it was the price or the quantity produced that
increased, or by what magnitude. As more goods and services are
considered, the problem gets bigger.
Real GDP is a more accurate indicator of changes in production. Referring to a base year eliminates the uncertainty of whether
an increase in the value of the goods and services produced was the
result of increased prices or of higher production. FIGURE 1-5 shows
how real GDP would be calculated in another country with two
products—in this case, telescopes and hockey sticks.
To calculate Year 1 GDP, the quantities of the goods produced
that year are multiplied by the prices at which they were sold and
the results summed, to yield $6,000. For Year 2, instead of multiplying the quantities of goods produced by that year’s prices—which
would yield the nominal value—they are multiplied by their prices
in the base year, Year 1. This yields a real, or inflation-adjusted,
GDP of $7,650. According to this calculation, Year 2 GDP rose a
real $1,650 over Year 1.
Until 1996, the BEA used 1982 as the base year for calculating
Figure 1-5 Real GDP Calculation, Using Year 1 as the Base Year

Source:

Product

Telescopes
Hockey sticks

Quantity
Year 1
Year 2

10
200

14
250

Price

GDP

Year 1

Year 2

Year 1

Year 2

$100
$25

$125
$27

$1,000
$5,000
$6,000

$1,400
$6,250
$7,650



The Trader’s Guide to Key Economic Indicators

Source:

all real GDP estimates. Settling on one base year in this manner
has the effect of imposing that year’s price structure on subsequent
periods and fixing the relative weights given the goods associated
with these prices in the GDP calculation. The BEA found, however, that this fixed-weight approach introduced distortions: The
farther away a period under study was from the chosen base year,
the more inflated its real GDP growth rate tended to be. For example, Karl Whelan, an economist at the Federal Reserve’s Board
of Governors, has observed in a working paper that the growth rate
of fixed-weight real GDP in 1998 was 4.5 percent when calculated
using a base year of 1995, 6.5 percent using 1990 prices, 18.8 percent using 1980 prices, and an incredible 37.4 percent when 1970 is
the base year.
The BEA constantly refines its measures. (That’s part of the
reason the economic statistics in the United States are better and
more accurate than those in any other developed nation.) In the
mid-1990s, the Bureau decided it was time to refine its weighting
method and, in late 1995, adopted chain weighting. The chainweighting process is far too complex for this introduction, but in
essence, rather than holding constant a basket of goods and services, as in the fixed-weight system, it holds the “utility” of the basket constant, allowing substitution of cheaper for more expensive
items. Moreover, the base year is moved forward as the estimate
progresses through time. The result is a series of links, or a “chain”
of estimates that minimizes deviations.
The primary drawback of using chain-weighted (chained) data
is the loss of additivity. In the fixed-weight calculation, total real
GDP measured in 1996 dollars was equal to the sum of its components valued in 1996 dollars, and the value of each component was
equal to the sum of the values of its subcomponents. As illustrated
in FIGURE 1-6, this is not the case when chain weighting is used. Note
that when the real chained components are summed, they do not
add up to the actual real chained dollar total consumption figure of
$6,637.9 billion.

Gross Domestic Product



Figure 1-6 First Quarter 2003 Consumption Expenditures ($ in billions)
Nominal (Current Dollar)

Nondurable goods
Durable goods
Services
Total consumption

$2,150.0
873.9
4,401.5
= $7,425.4

Real Chained Dollar

$1,950.0
1,010.6
3,707.0
≠ $6,637.9

Source: U.S. Department of Commerce, Bureau of Economic Analysis

DEFLATORS
The difference between nominal GDP and real GDP is essentially
inflation. It is thus possible to compute an economy’s inflation rate
from this difference. The result of the computation is called an implicit price deflator.
Every GDP report contains implicit price deflators for the headline GDP number and also for many of its subcomponents, such as
consumption expenditures, government spending, and gross private
domestic investment. Economists at the BEA calculate the GDP
implicit price deflator using the formula:
(Nominal Value)/(Real Value) x 100 = Implicit deflator,

For example, using data from the 2003 first quarter GDP report,
the GDP deflator for that period would be:
$10.698 x 100 = 111.947, or approximately 111.95
$ 9.556

An annualized inflation rate for a period can be derived using the
formula:
[(current-period deflator / previous-period deflator )4 – 1] x 100
= annualized inflation.

To compute the annualized inflation rate for first quarter
2003, for example, the first quarter 2003 GDP deflator computed above and the fourth quarter 2002 deflator of 111.25



The Trader’s Guide to Key Economic Indicators

would be plugged into the formula, to give
[(111.95 / 111.25)4 – 1] x 100 = [(1.00629)4 – 1] x 100
= (1.025398 – 1) x 100 = 2.539%, or approximately 2.54%.

A similar formula is used to calculate the annualized quarterly
growth rate of GDP as a whole as well as each of its components
and subcomponents:
[(current quarter / previous quarter)4 – 1] x 100
= quarterly annualized growth rate.

For example, to compute the fourth quarter 2002 growth rate,
the third and fourth quarter 2002 GDP figures would be plugged
into the formula, giving:

=
=
=
=

[(QIII 2002 GDP) / (QIV 2002 GDP)4 – 1] x 100
[(9,518.2 / 9,485.6)4 – 1] x 100
[(1.00344)4 – 1] x 100 = [(1.013831165) – 1] x 100
0.013831165 x 100
1.383%, or approximately 1.4%

As noted earlier, economic activity has two sides—expenditures and
income—which correspond to two different ways of calculating
GDP. The discussion so far has involved expenditures. The income
side of GDP calculation is less sexy than the expenditure approach
because it doesn’t identify the industries or products that are being
created. Traders tend to pay less attention to the factors involved
in national income, but it is equally important. Investors, particularly equity traders, like to see the quarterly performance of their
respective investment industries. For example, those traders heavily
invested in software stocks want to know how software investment
fared during the particular quarter. The income-determined approach of GDP calculations does not provide this perspective.

Source:

NATIONAL INCOME

Gross Domestic Product



The sum of the incomes generated in the course of production
is termed national income. Its components fall into the following
five categories:
1. compensation of employees (wages and salaries, plus
supplements)
2. net interest
3. proprietors’ income
4. rental income of persons
5. corporate profits
FIGURE 1-7, from the BEA’s fourth quarter 2002 report on GDP,
identifies these components together with the percentage each contributes to total national income. Unlike expenditure-based GDP
and its components, the income data are reported only in nominal
terms—that is, they are valued only in current prices. They are also
subjected to valuation adjustments.

EMPLOYEE COMPENSATION
Employee compensation accounts for roughly 70 percent of
national income. It comprises two parts. The largest is composed
of wages and salaries, including commissions, tips, bonuses, and
employee contributions to deferred compensation plans such as

Source: U.S. Department of Commerce, Bureau of
Economic Analysis

Figure 1-7

National Income ($ in billions)
2002: QIV

Compensation of employees
Wage and salary accruals
Supplements to wages and salaries
Proprietors’ income with inventory valuation
and capital consumption adjustments
Rental income of persons with capital
consumption adjustment
Corporate profits with inventory valuation and
capital consumption adjustments
Net interest
National Income

$6,048.8
$5,052.4
$996.4

Percent of
Total Income

71.62%
59.82%
11.80%

$771.6

9.14%

$130.6

1.55%

$796.1
$698.3
$8,445.4

9.43%
8.27%
100.00%



The Trader’s Guide to Key Economic Indicators

401(k)s. For the most part, the BEA estimates this component by
multiplying employment in the Bureau of Labor Statistics’ monthly
Employment Situation report (described in Chapter 3) by earnings
and the number of hours worked. The second component of compensation, accounting for approximately 16 percent of the total, is
composed of “supplements,” such as employer contributions for
social and unemployment insurance.
Net interest is the interest that businesses, foreign corporations operating in the United States, life insurance companies, and
several other related interest-disseminating sources pay out as part
of the expense of operating, less the interest they receive. Interest payments on mortgages and on home improvement and equity
loans are considered business costs because the NIPAs treat home
ownership as a business. The BEA gathers most of the data for the
net interest calculation from Internal Revenue Service (IRS) tax returns, the Federal Reserve Board, regulatory agency annual reports,
and the Department of Agriculture.

The last three categories—proprietors’ income, rental income, and
corporate profits—are usually tweaked, through the application of
an inventory valuation adjustment (IVA) and a capital consumption
adjustment (CCAdj). The IVA adjusts for the data discrepancies
that occur because some businesses value their inventories at initialacquisition, or historical, cost rather than at current-replacement
cost, which is the BEA’s method. The CCAdj deals with the fact that
businesses account differently from national income accountants
(i.e., the BEA) for depreciation (referred to by economists as capital
consumption)—that is, the reduction in value throughout the measurement period of income, profits, inventories, and goods. Because
businesses have several methods of inventory accounting, including
the BEA’s CCAdj, the BEA has adopted the CCAdj as a more consistent and uniform inventory and capital consumption adjustment
system. The IVA and CCAdj are two reasons that the income and
the expenditure computations of GDP aren’t the same.

Source:

OTHER INCOME CATEGORIES

Gross Domestic Product



Proprietors’ income comprises the earnings of nonincorporated businesses (sole proprietorships and partnerships). The dollar
amount of this income is calculated using IRS business tax returns,
with inventory valuation and capital consumption adjustments. The
category accounts for about 9 percent of total income.
Rental income is composed of the rents earned from residential and nonresidential property by people not primarily engaged
in the real estate business, plus royalties received from copyrights
and patents.
The GDP report refers to several types of corporate profits.
Pretax profits, also known as book profits, are what companies
earn before paying taxes and distributing dividends to shareholders.
Applying the IVA and CCAdj to this total results in profits from
current production, termed operating profits in the business
community. This is the corporate profits figure used in computing
national income. Subtracting companies’ tax liabilities from book
profit gives after-tax profits. FIGURE 1-8, taken from the final GDP
report of fourth quarter 2002, illustrates how the various corporate
profit measures are related.
The corporate profits data are obtained from IRS tabulations, as
well as from the Census Bureau’s quarterly survey of corporate profits
and publicly available corporate financial statements. Corporate profits account for approximately 10 percent of total national income.
Figure 1-8

Corporate Profits ($ in billions)

Source: U.S. Department of Commerce, Bureau of
Economic Analysis

2002-I

Corporate profits with inventory valuation
and capital consumption adjustments
Corporate profits with inventory valuation
adjustment
Profits before tax
Profits tax liability
Profits after tax
Dividends
Undistributed profits
Inventory valuation adjustment
Capital consumption adjustment
Net interest

$797.6
641.3
639.4
202.4
437
424.2
12.8
1.9
156.3
672.8

2002-II

$785
652.2
657.9
213.7
444.3
430.8
13.5
-5.7
132.8
678.1

2002-III

$771
653.4
668.5
214.7
453.8
437.7
16.1
-15.1
117.6
687.6

2002-IV

$796.1
686.4
694.9
222.4
472.5
444.3
28.2
-8.5
109.7
698.3



The Trader’s Guide to Key Economic Indicators

Not every GDP report depicts corporate profits in the same
detail as the table shown in Figure 1-8. Because corporate earnings
reports are scattered throughout the quarter and IRS processing
of corporate tax returns is rather lengthy, accurate tallies are only
possible months after the end of the quarter. The most complete
presentation of corporate profits is usually provided in a year’s final
report of GDP.

National income, as noted above, is the sum of all the incomes
generated by the factors involved in production. This total does
not equal the expenditure-determined GDP. To reach equality,
several adjustments must be made. These adjustments are shown
in FIGURE 1-9, which reproduces Table 8 of the fourth quarter 2002
GDP report.
The largest adjustment concerns the consumption of fixed
capital. This is essentially the depreciation charge taken by private and government owners of fixed capital located in the United
States to account for the assets used up in the course of production. The amount of the charge is estimated by the BEA from IRS
business tax returns and studies of resale prices of used equipment
and structures.
The next category of adjustments involves indirect business
taxes (such as sales, excise, and property taxes, and customs duties)
and nontax liabilities, which include fines. The BEA estimates most
of these using a so-called judgmental trend based on the Census
Bureau’s quinquennial censuses and annual and quarterly surveys.
Again, because data are not always available on a quarterly basis,
the BEA must extrapolate quarterly data from annual surveys. This
process of extrapolation and interpolation is referred to by the BEA
as a judgment trend.
The third category consists of transfer payments—distributions
that private (i.e., nongovernmental) businesses make to individuals
without any product changing hands or services being rendered.
Two examples are charitable donations and liability payments made

Source:

GNP, GDP, AND NATIONAL INCOME

Gross Domestic Product
Figure 1-9



Relation of GDP, GNP, and National Income ($ in billions)

Source: U.S. Department of Commerce, Bureau of Economic
Analysis

QIV: 2002

Gross Domestic Product
Plus: Income receipts from the rest of the world
Less: Income payments to the rest of the world
Equals: Gross National Product

$10,588.8
$284.2
$293.4
$10,579.6

Less: Consumption of fixed capital
Less: Indirect business tax and nontax liability
Less: Business transfer payments
Less: Statistical discrepancy
Plus: Subsidies less current surplus of government enterprises

$1,415.4
$813.3
$44.3
$109.6
$29.0

Equals: National Income

$8,445.2

for personal injury. The BEA gets these data from IRS business tax
refunds, government agency reports, and other trade sources.
All these figures—for depreciation, taxes, and transfers—are
added to national income. In contrast, the fourth category of adjustments, subsidies less the current surplus of government enterprises,
is subtracted. The subsidies referred to are the distributions that
government agencies make to private businesses as well as to other
levels of government, such as to the U.S. Post Office.
After all these adjustments are made, gross domestic income
should equal the gross domestic product. However, a difference,
termed the statistical discrepancy, still remains—fourth quarter
2002 GDP, for instance, was $10.579.6 billion and gross domestic
income $10,689.2 billion by the income-based calculation, a difference of $109.6 billion. This discrepancy reflects differences in the
sources for data used in the two calculations. Those used in deriving
national income are less directly observable, and so less reliable. As
mentioned earlier, moreover, illicit expenditures are not reported or
estimated.
This section of the chapter has described the multitude of
figures included in the GDP report, how they are related to one
another, and how they are derived. Next comes the nuts and bolts:



The Trader’s Guide to Key Economic Indicators

how economists and traders use the report numbers in determining
both big-picture issues, such as the future course of the business
cycle, and smaller ones, such as when to put their money where.

WHAT DOES IT ALL MEAN?
The GDP report is a mother lode of information about the nation’s
economy. Each of its components tells a different story about a particular group, sector, industry, or activity. Not surprisingly, then,
different market participants look at different sections and draw
different inferences. Retail analysts, for instance, focus mostly on
consumer spending. Those covering housing, construction, or real
estate investment trusts (REITs) concentrate on the residential activity in investment spending. Military and defense industry analysts
focus on the national defense spending component of government
consumption expenditures and gross investment. Fixed-income
analysts and investors, ever wary of the eroding effects of inflation,
concern themselves with the GDP deflators and GDP growth rate.
Traders, who are always on the lookout for possible market movers,
watch for numbers that contradict expectations, which they track
carefully, often jotting them down in notebooks kept at their desks,
for quick reference when the real figures are announced.

The quarterly annualized growth rate of real GDP is the headline
number of the GDP report. As with most economic figures, strong
positive postings are generally good news for the economy, corporate
profits, and stock valuations. Not so for bonds, however. Inflation
erodes the value of fixed-income securities, and more torrid economic growth is usually associated with higher rates of inflation.
Market reactions—both positive and negative—are more pronounced when the announced numbers differ from the expected
ones. The larger the difference, the greater the market move. Say
the Street consensus for the third quarter was for an annualized
GDP growth rate of 4.2 percent. On the one hand, a weak post-

Source:

GDP GROWTH

Gross Domestic Product



ing of between 1.0 and 2.0 percent would probably spark a sell-off
in the stock market and boost the price of fixed-income securities,
lowering yields. Stronger-than-expected growth of 5.5 to 6.5 percent, on the other hand, would be well received by equity traders
and frowned upon by fixed-income dealers.
Although the quarterly annualized figure is important, many
economists prefer to look at the year-over-year change in GDP.
The longer perspective makes it is easier to spot turning points in
the economy, such as an approaching recession or an acceleration of
activity. FIGURE 1-10 illustrates this predictive effect.
As the chart shows, in the past twenty years the U.S. economy has
experienced two recessions—in 1990–91 and 2001—both of which
were preceded by significant declines in the growth rates of real and
nominal GDP. Note that although the real GDP growth rate falls
below zero, the nominal rate declines but stays out of negative territory. This is because the nominal figure incorporates the effects of inflation, which is almost always rising. For the growth rate of nominal
GDP to become negative, the inflation rate would have to be falling—
a condition known as deflation—at the same time that the economy
was contracting. Deflation is extremely rare in the United States and
indeed has been recorded only a couple of times anywhere.

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; NBER

Figure 1-10

Year-Over-Year Percent Change in Real and Nominal GDP

Percent

14
Shaded areas = Recession

Nominal
Real

10
8
6
4
2
0
–2
1983

1988

1993

1998

2003



The Trader’s Guide to Key Economic Indicators

On average, the year-over-year growth rate in GDP starts declining four to five quarters before a recession. Not all slowdowns,
however, result in recession. By the time the warning signals appear,
government policymakers have usually put in place measures to
avert an economic downturn. Still, watching changes in year-overyear GDP growth can be useful for short-term forecasts: Very rarely
do trends reverse immediately. It takes a great deal to knock a $10
trillion economy like that of the United States off kilter. Luckily for
those in the financial markets, several “leading” indicators usually
send alerts when the behemoth is running out of energy.
In order to form a clearer picture, economists like to look at several indicators at once. This helps reduce the transmission of false
signals. There are times when some individual indicators trend lower,
suggesting a potential decline in activity. If several indicators are observed, and a majority point to positive activity, then it is possible to
dismiss the weaker performing indicators as outliers, and draw the
conclusion that the economy isn’t on course to fall off track.

If GDP growth is the most important number in the release, the
GDP deflators run a close second. As indicators of inflation, these
deflators are preferred to the Consumer Price Index (CPI), the
Producer Price Index (PPI), and other commodity price gauges by
many traders and economists, including those at the Federal Reserve. Special favorites are the deflators for government consumption expenditures and gross investment, for personal consumption
expenditures, and for personal consumption expenditures less food
and energy, also known as PCEDXF&E. The Street has adopted
the last deflator as an unofficial benchmark for the core rate of inflation. Bond traders in particular watch the deflators, knowing that
greater-than-expected increases in these numbers usually depress
fixed-income prices.
Why have the deflators superseded the other inflation measures?
For starters, policymakers, traders, and investors in general, want to
see overarching economic trends, not smaller, more targeted ones.

Source:

DEFLATORS

Gross Domestic Product



GDP deflators reflect price activity in the broader economy. The
CPI, in contrast, is merely a “basket” of a few hundred goods and
services, chosen by the Bureau of Labor Statistics. (For a fuller explanation of price activity and the core rate of inflation, see Chapter
12, Consumer and Producer Price Indices.)
Traders focus on movements in the personal consumption expenditure deflator excluding food and energy, commonly referred
to as the Core PCED. This inflation measure is preferred to most
of the others as it measures the core, ex-food and energy, rate of
inflation that consumers face. Because prices of food and energy
can fluctuate greatly during the month, economists like to view
price trends without these noisy readings. Also, because private
individuals are doing the overwhelming majority of the economy’s
consumption, and this indicator contains all of the goods and services consumed, as opposed to a couple of hundred as in the case
with the CPI, the Core PCED has risen to the top of the list of
most watched inflation gauges.

Source:

CONSUMPTION EXPENDITURES
As the consumer goes, so goes the U.S. economy. And this old
saw may be more truthful than ever before. It is believed that the
consumer’s utter resilience to recent disruptions such as war, attacks against the United States on U.S. soil, widespread corporate
malfeasance, eight of the top ten corporate bankruptcies in U.S.
history, and presidential impeachment proceedings is the reason for
the underlying strength of the economy. In previous decades, any
single one of these disruptions likely would have upended the U.S.
economy. Now it seems as though the consumer is capable of keeping the economy humming. It is the consumer that has prolonged
expansion, and made recessions shorter and milder.
Generally, a drop in the growth rate of consumer spending is a
surefire sign that the economy is on the verge of petering out. When
people are feeling uneasy about the economic climate—perhaps
unemployment is on the rise, or inflation is eroding the dollar’s
purchasing power, or individuals are just feeling tapped out—



The Trader’s Guide to Key Economic Indicators

it shows in their spending habits. As the chart in FIGURE 1-11 shows,
pronounced declines in the year-over-year growth in consumer expenditures have preceded each of the six recessions in the United
States since 1963. Traditionally, the first retrenchment occurs in
purchases of big-ticket items, such as durable goods. So it is in that
portion of consumer spending where you’ll find early warnings of
economic downturns.
(For a more detailed discussion of consumer spending and its
trends, see Chapter 11, Personal Income and Outlays.)
INVESTMENT SPENDING
Capital equipment comprises all the industrial and technological items used to produce other goods and services for sale. The
amount of money companies invest in this equipment is thus a good
predictor of future economic activity. It indicates whether corporate
profitability is accelerating or decelerating, how managers view future economic conditions, and how strong or weak the economy is.
As explained earlier, the Street tends to focus on fixed investment—gross domestic investment minus inventories. Of the two
categories of fixed investment, residential and nonresidential (or
Personal Consumption Expenditures

YOY%

Shaded areas = Recession

12
10
8
6
4
2
0
1963

1973

1983

1993

2003

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; NBER

Figure 1-11

Source:

Gross Domestic Product



capital spending), the former is by far the smaller, accounting for
just 25 to 30 percent of the total. One shouldn’t underestimate the
influence of residential business investment, however. It represents
roughly 4.5 percent of total economic output, and housing construction has a tremendous multiplier effect on the economy: Once
a house or apartment building has been erected, personal consumption expenditures usually receive a big boost as owners head out to
paint, decorate, and furnish their homes.
That said, analysts and economists tend to pay more attention
to nonresidential investment. In part, this is because of the component’s size—it accounts for almost three quarters of total fixed
investment. It also provides a great deal of insight into how the
corporate sector views economic conditions. Finally, many equity
traders, especially those active in the Nasdaq and on the lookout for
the next Microsoft or Intel, are particularly interested in technology
investment, which falls into the nonresidential category.
A certain amount of nonresidential fixed investment always
needs to be performed through the year regardless of the overall
state of the economy. Equipment and machinery, for example,
constantly need to be refurbished, updated, and repaired. Every
year the auto industry shuts down its plants for about two weeks to
allow engineers to retool machinery for upcoming new car models.
Weather, overuse, and just plain wear and tear cause capital equipment to break down. During booming periods of technological advances, some capital equipment becomes obsolete. Upgrades often
help a business raise its level of productivity, which in turn helps the
company’s bottom line.
Rising capital spending is generally associated with periods of
solid corporate profitability and economic prosperity. For businesses to invest in new capital equipment, they need sufficient profit
growth. After all, they can’t spend what they don’t have. (Actually
businesses can spend or invest by borrowing via issuance of bonds.
But if the company doesn’t have respected profit growth, then the
ability to obtain the financing is hampered. With a poor financial
history, companies are saddled with low credit ratings and are
forced to pay higher returns for borrowing those needed funds.)



The Trader’s Guide to Key Economic Indicators

Management also needs to be positive about the economic outlook. If conditions are soft and consumer demand unpromising,
they will be less inclined to purchase new machinery and equipment. If, however, the economy is expanding at a respectable pace,
economic fundamentals are conducive to continuing growth (low
interest rates, low inflation, firm labor market growth), and consumers are spending, then businesses will be more likely to pick up
the pace of their investment.
Capital equipment is generally very costly—think of the specialized machinery on automakers’ assembly lines, the ovens and packaging systems in food-processing plants, the industrial-size kilns of
cement manufacturers. Companies thus usually need to borrow to
purchase it. So the amount of business investment is closely related
to the level of interest rates: Lower rates ease spending; higher rates
make it more difficult. Accordingly, the Federal Reserve can influence capital spending by altering its target for the Federal funds
rate, the rate banks charge each other for overnight loans used to
meet reserve requirements. If the Fed wants to spark capital spending, it lowers the overnight rate. Over time, yields on the entire
maturity spectrum, from three-month Treasury bills to the ten-year
Treasury note, decline as well, making it less expensive for businesses to finance costly investments such as new plants, factories,
and equipment.
When investors realize that interest rates may be headed lower,
whether as a result of slower inflation rates or by the Federal
Reserve’s influence, they know that businesses are likely to pick
up the pace of investment, because the financing of those products
and services is going to be cheaper. In order to capitalize on such
developments, traders might bid-up the prices of those stocks that
have their primary business in investment-related concerns like
technology, machinery, tools, or capital equipment. Some of the
more common companies that are involved in capital equipment include: Cummins Inc., Deere & Co., Paccar Inc., Briggs & Stratton,
Danaher Corp., Dover Corp., Eaton Corp., Illinois Tool Works,
Ingersoll-Rand, Parker-Hannifin Corp., Timken Co., and Wolverine Tube Inc.

Source:

Gross Domestic Product



Source:

GOVERNMENT SPENDING
Wall Street doesn’t generally pay much attention to government
consumption expenditures and gross investment. One reason is that
number’s stability. Since 1947, government spending and investment has accounted for about 15 percent of total economic output.
Only during periods of profound economic weakness or military
conflict does the percentage rise, as the government picks up the
pace of spending to boost economic growth or to support the war
effort. In the post–World War II era, a peak of 24 percent was registered in 1953, at the end of the Korean War.
Within the government data, however, is one item to which
some economists do pay attention, especially in recent times. That
item is national defense spending. The long-term trend in national
defense as a percentage of total government spending since the end
of World War II has been consistently downward. Still, increases
(in some instances, slight) have occurred when the government has
ramped up purchases for military conflicts such as the Korean War,
in the early 1950s; Vietnam, in the mid-1960s to early 1970s; Desert Storm, in 1990; and, most recently, the war against terrorism
in Afghanistan and Iraq. Keep in mind that government spending
on national defense isn’t limited to the increased output of aircraft,
electronic tracking devices, and missiles. Greater defense spending raises the level of employment—everything from engineers to
manufacturing positions. And due to security reasons, those jobs
tend to stay here in the United States and are not shipped abroad as
so many of the manufacturing positions have been in recent years.
Stock analysts responsible for the defense contractors and
aerospace companies, such as Northrop Grumman, Raytheon,
Lockheed Martin, General Dynamics, Curtiss-Wright, and Boeing, find the detail on national defense expenditures in the report
a treasure trove. The category is broken down into spending on
aircraft, missiles, ships, vehicles, electronics and software, ammunition, petroleum, and compensation. If the government bought it,
it’ll be recorded here.



The Trader’s Guide to Key Economic Indicators

When the United States imports more than it exports—as has been
the case for the better part of the past three decades—the net export balance is said to be in deficit. This reduces the level of GDP
produced in a given period. Conversely, when exports outweigh
imports, the trade balance is said to be in surplus. This results in an
addition to economic activity. Such an outcome stands to reason,
as U.S. export goods are produced by plants located in the United
States whereas imports have been produced by foreigners and sent
to the United States. FIGURE 1-12 represents the value of net exports
as a percent of GDP. This percentage has been negative for a majority of the last thirty years, implying that the pace of imports is
greater than that of exports, which reduces the level of domestic
economic activity.
Imports needn’t have a negative connotation, however. A number of resources are not as abundant in the United States as they are
outside its borders. One obvious example is crude oil. The United
States has domestic sources of oil but not enough to fuel its consumption. For that reason, it has to import about half its crude oil
from foreign countries. Should we consider these imports disapprovingly? Absolutely not. The mere fact that the United States
consumes so much crude is testament to its economic vitality. Its
plants and factories need a great deal of oil to produce what is the
largest output in the world, employing millions of people and creating an economic climate that permits its citizens to prosper like no
others on Earth. Spending on imports to heat our homes, run our
transportation system, and conduct business should not be considered a drag on prosperity but an enhancement.
As with government expenditures, the trading community has
little reason to get excited about the net export balance. It’s true that
the business community frowns on widening trade deficits, because
increasing imports slow U.S. GDP growth. But rising imports also
mean that U.S. businesses and households are consuming more
goods and services that they deem attractive. Nobody forces consumers to purchase Italian wine, Japanese cars, or Canadian lumber.

Source:

NET EXPORTS

Gross Domestic Product

Source: U.S. Department of Commerce, Bureau of Economic
Analysis

Figure 1-12



Net Exports as a Percentage of GDP

Percent

5
4
3
2
1
0
–1
–2
–3
–4
–5

1947

1954

1961

1968

1975

1982

1989

1996

2003

U.S. businesses and households purchase foreign-made goods for
any number of reasons including, price, quality, size, and taste. The
primary force behind demand for foreign-produced goods is simply
desirability.
Furthermore, several foreign produced goods tend to be cheaper.
Because many countries in the world, particularly China, India, and
several Asian-Pacific nations, have practically free labor, they are
capable of producing goods at little cost. These low-priced products
are usually sent to the United States, which influences the prices of
similar U.S.-produced good. This globalization has led to a lower inflation rate here in the United States—especially since the mid 1990s.
Perhaps the major reason investors ignore the trade data is the
data’s minor influence on total economic activity. Over the past
fifty-five years, the net export position has averaged a mere half a
percentage point of total economic output.
FINAL SALES
Included in the addenda to Table 1 in the GDP report are three
measures little noted by the financial media but closely scrutinized
by the trading community because of the insights they provide into



The Trader’s Guide to Key Economic Indicators

the underlying spending patterns in the GDP numbers. These
three indicators are the final sales of domestic product, gross domestic purchases, and final sales to domestic purchasers.
Final sales of domestic product is a measure of the dollar
value of goods produced in the United States in a particular period
that are actually sold, rather than put into inventory. To calculate
this figure, the BEA first computes “the change in private inventories,” or CPI, by comparing the current level of inventories with
that of the previous period. This indicates how many goods have
been added to businesses storage and thus how much of current
production has remained unsold. CPI is then subtracted from
GDP to give final sales. This is an important number, because it
paints a more accurate picture than GDP of the current pace of
spending in the economy. Economists say current pace because
the quarterly figure excludes inventories that have been produced in previous quarters. Many times economists will compare
the growth rates of GDP with those of final sales to determine
whether economic growth is being driven by new production or
by the consumption of goods that were previously produced and
stored as inventories.
Gross domestic purchases measures all the goods U.S. residents have bought, no matter where the goods were produced. This
figure is obtained by subtracting net exports from GDP. There is
indeed a difference between GDP and gross domestic purchases.
GDP is a measure of domestically-produced goods and services,
while gross domestic purchases is a measure of all the goods domestically purchased. Strong quarterly increases in gross domestic
purchases generally imply solid demand by U.S. consumers as only
those purchases of domestic goods are calculated.
Final sales to domestic purchasers is the level of gross domestic purchases less the change in private inventories. It depicts the
desire of Americans, both households and businesses, to spend, no
matter where the goods or services are produced. Some economists
consider it a good indicator of overall economic well-being. Slumping final sales to domestic purchasers suggests that U.S. consumers
are tapped out.

Source:

Gross Domestic Product



Economists keep track of the year-over-year percentage change
in final sales to domestic purchasers because of this measure’s excellent record of foretelling periods of softer economic growth. As the
chart in FIGURE 1-13 illustrates, each of the four recessions since 1980
was preceded by about a three quarter long decline in the year-overyear growth rate of final sales to domestic purchasers.
CORPORATE PROFITS
Market participants don’t generally pay as much attention to the
income side as to the expenditure side of GDP. That isn’t to say
the trends in wages and salaries aren’t important to economists
or to analysts who cover retail issues. What could be more telling
about the future pace of spending, after all, than the amount of
income earned by would-be consumers? It’s just that the trends of
the expenditure side are accepted as being more accurate, because
they aren’t subject to inventory and capital-consumption value
adjustments, as the income-determined data are. Still, some income-side components can give valuable insights into economic
trends. Among the most important of these are the measures of
corporate profits.

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; NBER

Figure 1-13

Final Sales to Domestic Purchasers

YOY%

Shaded areas = Recession

12
10
8
6
4
2
0

1980

1986

1992

1998



The Trader’s Guide to Key Economic Indicators

As with most of the other measures discussed, a rise in corporate profits indicates a healthy business climate. The economy’s
growth cycle really starts with a lift in corporate profits. When
businesses are successful, their incomes exceed costs, and they
make profits. This permits them to invest in new capital equipment or employees.
Even more significant than pretax earnings are after-tax profits. From this figure, economists and analysts can judge how much
money companies actually have to spend on new equipment or additional staff. As the chart in FIGURE 1-14 shows, businesses generally
shed workers when corporate profit growth contracted (below zero
in the chart). The same holds true for business investment. Aftertax corporate profits decline approximately three quarters prior to
periods of slowing economic growth or recessions.
The best measure of the funds that companies have available for
spending and hiring, however, is the level of undistributed profits.
These are a company’s earnings after tax payments and dividend
distributions. One striking feature of the chart in FIGURE 1-15, which
shows the amount of undistributed profits in the last third of the
twentieth century, is the paltry level of undistributed profits during
the early 1970s, 1987, and 2002. All three periods were associated
After-Tax Corporate Profits and Payroll Growth

Payrolls %

Profits %

Payrolls
After-tax corp profit

–5

–1

–10

–2

–15

–3

–20
1980

1984

1988

1992

1996

2000

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; U.S. Department of Labor, Bureau of Labor Stattistics

Figure 1-14

Gross Domestic Product



with tumbling stock prices, high unemployment rates, and lackluster
business investment.
The economic signals associated with corporate profits might
not be as telling as they once were. As was noted earlier, in recent
years the U.S. economy has become practically impervious to a
whole host of negative influences that, if they had occurred in
previous periods, would have resulted in recession and, in some
instances, quite possibly depression. Beginning in early 2001, the
stock market bubble of the late 1990s burst, wiping out trillions of
dollars in personal wealth. Widespread accounting scandals and
egregious corporate impropriety also hammered investors’ confidence, stalling the financial markets. For the first time in more
than fifty years, the United States was attacked on its own soil,
virtually paralyzing the economy. Hundreds of thousands of businesses closed for weeks, and the borders were sealed. Fear of anthrax attacks was widespread. As if all of this weren’t enough, U.S.
armed forces engaged in military conflicts in Afghanistan and Iraq.
Yet despite all these profoundly negative influences in a relatively
short period, the economy managed to avoid a deep or prolonged
recession. Perhaps the ultimate sign of resiliency is that consumer
spending never fell.

Source: U.S. Department of Commerce, Bureau of Economic
Analysis

Figure 1-15

Undistributed Profits

$ billions

250
200
150
100
50
0
1970

1980

1990

2000



The Trader’s Guide to Key Economic Indicators

HOW TO USE WHAT YOU SEE
There aren’t as many tricks associated with the National Income
and Product Accounts as with other economic series. One reason
may be that these accounts are the benchmark of economic activity,
and traders use other indicators to anticipate movements in GDP.
In other words, the level of GDP is usually the variable that other
indicators attempt to forecast or emulate. GDP is also released on
a quarterly basis, and the economic associations and relationships
it points to aren’t as predictive as those expressed on a monthly or
weekly basis. That said, Wall Street economists and policymakers
do have one particularly useful strategy that employs data from the
GDP report: calculating the output gap.
THE

TRENCHES

The output gap is the difference between the economy’s actual
and potential levels of production. This difference yields insight
into important economic conditions, such as employment and
inflation.
The economy’s potential output is the amount of goods and
services it would produce if it utilized all its resources. To determine this figure—the trend level—economists estimate the
rate at which the economy can expand without sparking a rise in
inflation. It is not an easy calculation, and it yields as many different answers as there are economists with different definitions
for the maximum level of output, productivity, hours worked, and
so on. Luckily, a widely accepted estimate of potential output
is reported relatively frequently, about once a quarter, by the
Congressional Budget Office. The CBO’s website, www.cbo.gov,
contains information about its methodology and underlying
assumptions in computing the trend level, as well as a detailed
historical data set.
A negative output gap exists when actual GDP growth is below
its estimated potential. This suggests that the economy isn’t utilizing
all its labor and capital resources. Such periods of “undercapacitiza-

Source:

TRICKS FROM

Gross Domestic Product



tion” are usually characterized by high unemployment and low inflation, with plants and factories closing down, workers furloughed,
and machinery idled. The chart in FIGURE 1-16 shows that in 1990–91
and 2001–02, periods of profound economic weakness, the actual
growth rate of real GDP was considerably below its potential.
When GDP growth exceeds its calculated potential, creating a
positive gap, the economy is pushed to its limit. All plants and factories are running at capacity, the labor force is fully employed, and
economic output is sky-rocketing. The chart in FIGURE 1-17 illustrates
the relationship between a positive gap and falling unemployment.
In periods of overcapacitization, such as 1997–2001, strains on the
system develop, usually sparking inflation.
Economists sometimes express the output gap in the form of a
ratio, derived by dividing actual output by potential output. When
this ratio falls below zero, conditions are recessional; when it rises
above zero, conditions are expansionary.
Because the output gap provides such telling economic insight into a whole host of economic relationships, it is a favorite of policymakers. The Federal Reserve, for example,
considers it in determining where to set the Fed funds rate. If the
gap is negative, indicating that the economy is growing below its

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; Congressional Budget Office

Figure 1-16

GDP Versus Potential GDP

$ billions

10,000

9,000

Potential
Actual

8,000

7,000

6,000
1987

1989

1991

1993

1995

1997

1999

2001

2003



The Trader’s Guide to Key Economic Indicators

Figure 1-17

Output Gap and Unemployment (inverted scale)

Unemployment reverse scale

Gap

Unemployment rate

Output gap

2
1

0
6

–1
–2

–3

–4
1990

1992

1994

1996

1998

2000

2002

Sources: U.S. Department of Commerce, Bureau of Economic Analysis; Congressional Budget
Office; U.S. Department of Labor, Bureau of Labor Statistics

Figure 1-18

Output Gap and Fed Funds Rate

Gap

FF%

–1

4
3

–2

Output gap
Fed funds

–3

1
0

–4
1988

1990

1992

1994

1996

1998

2000

2002

Sources: U.S. Department of Commerce, Bureau of Economic Analysis; Congressional Budget
Office; Board of Governors of the Federal Reserve System

Gross Domestic Product



Source:

potential, the Fed may try to spark activity by lowering the overnight rate. This results in a decline all along the maturity spectrum,
making it easier for companies to fund capital projects. It also spurs
individuals’ spending by rendering loans to purchase items such as
automobiles and homes more affordable. Conversely, when the gap
is positive, indicating that the economic party is getting a bit out
of hand, the Fed may take away the punch bowl by increasing its
overnight target rate, thus discouraging consumers and businesses
from spending and investing. The chart in FIGURE 1-18 illustrates the
tendency of the Fed funds rate to follow the output gap.

This page intentionally left blank

Indices of Leading,
Lagging, and Coincident
Indicators

f a market economist were given one wish, it might be for a
single indicator that would consistently predict both the direction
and the pace of the economy. Unfortunately, none has so far been
discovered that fits the entire bill. Some indicators are wonderful at
pinpointing levels of activity but fail to depict trends. Others excel
at identifying particular areas of economic strength and weakness
but can’t measure broad-based performance.
Unable to find a single omnipotent indicator, economists have
taken an assortment of those showing the most predictive accuracy
and combined them into the index of leading economic indicators
(LEI). The LEI is one of three composite indices—along with the
indices of lagging and coincident indicators—that the Conference
Board compiles and publishes in its monthly Business Cycle Indicators report. This report is usually released to the public at 10:00
a.m. ET four to five weeks after the end of the record month. It
is available, together with historical data and explanations of the
methodology behind the indices, on the Conference Board’s website, www.conference-board.org or www.globalindicators.org, and
by subscription for a small annual fee.
Wall Streeters often refer to the entire Business Cycle Indicators
report as the index of leading indicators, because that’s the part to
which they pay the most attention. In actuality, the charts, commentary, and data provided on all three indices are extremely useful
in identifying and explaining the different phases of the business
cycle. Whereas the leading economic index points to future trends
and turning points, the coincident index identifies those that are in
51



The Trader’s Guide to Key Economic Indicators

Figure 2-1

Timing of the Composite Indices Relative to Cyclical Turning Points
Composite

Composite

Leading

Coincident

Lagging

Index

Leads (-) or lags (+) at business cycle peaks (months)
Apr
Dec
Nov
Jan
Jul
Jul

1960
1969
1973
1980
1981
1990

…….
…….
…….
…….
…….
…….

-11
-8
-11*
-9
-15
-3
-8*
-6 ** -18*

0
-2
0
0
+1
-1

-3*

+3
+3
+13
+3
+2
-9

+3*
-8*

Leads (-) or lags (+) at business cycle troughs (months)
1961
1970
1975
1980
1982
1991

…….
…….
…….
…….
…….
…….

-3
-7
-1
-3
-8
-2

-2*
-1*
-2*
-10*

0
0
0
0
+1
0

+9
+15
+18
+3
+6
+21

+6*
+21*
+7*
+36*

*Timing of current (1997) version when different from the Department of Commerce
peak/tough designations
**-25 for absolute peak in cycle

the process of developing, and the lagging index confirms that these
events have indeed occurred. The table in FIGURE 2-1 shows the number of months by which the three composite indicators led or lagged
behind business cycle peaks or troughs, as defined by the National
Bureau of Economic Research (NBER), from 1960 through 1991.
Because the indices’ components are all released earlier than the
indices themselves, the markets generally don’t react strongly to
the indicator report. Market participants, however, can still glean
a great deal of information from the movements of the indices and
their components, not to mention the commentary and interpretation that the Conference Board’s staff economists supply each
month.

Source: The Conference Board

Feb
Nov
Mar
Jul
Nov
Mar

Indices of Leading, Lagging, and Coincident Indicators



EVOLUTION OF AN INDICATOR
The concept of composite economic indicators is not new, nor did
it originate with the Conference Board. In the early 1930s, economists Arthur Burns and Wesley Mitchell at the NBER were already
combining economic data series to identify trends and turning
points in the economy. NBER first published the results of these
efforts in 1938. By the 1960s, the U.S. Department of Commerce
was releasing monthly reports containing the NBER’s leading,
lagging, and coincident indicators. The Commerce DepartmentNBER collaboration lasted until 1995, when the Conference
Board—a private, nonprofit, nonadvocacy research and businessmembership group—assumed the responsibilities of calculating,
reporting, and maintaining the composite indices.
The leading, lagging, and coincident indices have all undergone
considerable revision in the course of their history. As the structure
of the U.S. economy has changed, newer indicators have periodically replaced older ones that no longer accurately reflected the
business cycle or simply weren’t calculated any more. As recently as
November 1996, the Conference Board dropped two indicators—
the price of sensitive materials and the volume of unfilled orders for
manufactured durable goods—from the leading index and added a
new one: the yield spread. This fine-tuning has kept the report an
accurate tool for Wall Street economists and market participants.

Source:

DIGGING FOR THE DATA
The three indices in the Conference Board’s report are composed
from series of cyclical indicators, most of which are seasonally adjusted. When the indices are constructed, some of the components
must be estimated; all are subject to later revision. The indices
themselves thus also need to be revised. The monthly release contains both initial values for the record month and revisions for the
previous six months.
In constructing an index, each component’s month-over-month
percentage change is calculated and then “standardized”—that is,



The Trader’s Guide to Key Economic Indicators

it is adjusted for volatility so that indicators with more dramatic
month-to-month movements won’t dominate the index. The standardized percentage changes for all the index’s components are then
added together. The sums derived for the leading and lagging indices are adjusted again, so that their standard deviations equal that
of the coincident index. Finally, the results for all three indices are
translated into levels representing changes from a base date, currently 1996, whose level is set at 100.
Because the composition of the three indices is modified as the
structure of the economy evolves, to remain in or be added to one
of the indices, a component must demonstrate consistency as a leading, coincident, or lagging indicator. It must also be the end product
of a reliable data-collection process; adhere to a timely publication
schedule; and be subject to only minor, or no, revisions.

The four components of the Conference Board’s coincident index
are the number of employees on nonagricultural (i.e., nonfarm)
payrolls (thousands), personal income less transfer payments (nominal rate in billions 1996$), industrial production index (1997=100),
and manufacturing and retail trade sales (in millions of 1996$).
The number of employees on nonagricultural payrolls, is
obtained from a survey of about 160,000 businesses, conducted by
the Bureau of Labor Statistics. The change in this number is one of
the headline figures in the BLS’s monthly Employment Situation
report (see Chapter 3). The path followed by nonfarm payrolls has,
in the main, paralleled that of growth in gross domestic product
(GDP).
Personal income less transfer payments is derived from the
Personal Income and Outlays report, produced by the Bureau of
Economic Analysis (BEA) (see Chapter 11). The largest income
source is wages and salaries, which account for about 55 percent
of the total; transfer payments—government disbursements such
as Social Security payments, veteran’s benefits, and food stamps—
usually constitute about 15 percent. Transfer payments are gener-

Source:

COINCIDENT INDEX

Indices of Leading, Lagging, and Coincident Indicators



ally spent immediately on basic necessities, such as food or rent,
not on durable goods and services. They thus have relatively little
influence on macroeconomic activity. So income less transfer
payments is generally considered a stronger, more representative
economic indicator.
The total industrial production index is the headliner of the
monthly Industrial Production and Capacity Utilization report
published by the Federal Reserve (see Chapter 4). It is constructed
of 295 components—representing the manufacturing, mining, and
utilities industries—that are weighted according to the value they
add during the production process. The index mirrors the general
economy so closely that it is often used as a more timely proxy for
the quarterly GDP report.
Manufacturing and retail trade sales data are collected as part
of the National Income and Product Accounts calculations. These
data may be found in the Manufacturing and Trade Inventories and
Sales (MTIS) report published by the Department of Commerce.
(For additional information, see the discussion of the sales portion
of the MTIS report examined in Chapter 7.)

Source:

LEADING ECONOMIC INDEX
The ten components of the leading index are the following:
1. Average weekly hours worked in manufacturing
2. Average weekly initial claims for unemployment insurance
3. Manufacturers’ new orders for consumer goods and
materials
4. The slower deliveries diffusion index of vendor
performance
5. Manufacturers’ new orders for nondefense capital goods
6. Monthly building permits for new private housing
7. Stock prices, 500 common stocks
8. The M2 money supply (in 1996 dollars)
9. The interest rate spread between the 10-year Treasury
bond and the federal funds rate
10. The Index of Consumer Expectations



The Trader’s Guide to Key Economic Indicators

The rationale behind including some of these indicators is clear
from their names: new “orders” and “expectations,” for instance,
are obviously forward looking. The inclusion of others is less selfevident. All the components, however, were chosen because of their
potency as predictors of economic activity.
The number of average weekly hours worked in manufacturing is derived from the same survey as the nonagricultural payroll
figure described above and is also published in the BLS’s Employment Situation report. Average weekly manufacturing hours constitute a good measure of future production levels and economic
strength. Assuming workers maintain the same level of productivity, the more hours they put in on the job, the greater their output. When manufacturers foresee a softening in demand for their
products, they tend to reduce the number of workers’ hours before
cutting staff, which is more time-consuming and expensive to implement. It is also easier and cheaper to extend hours, should business
seem poised to pick up, than to hire new employees. Substantial
changes in the average hours worked thus reflect companies’ pessimism or optimism about future economic conditions.
The average number of weekly initial claims for unemployment reflects the condition of the labor market (see Chapter 3).
A rise in claims, as businesses lay off more and more employees,
usually occurs in the early stages of economic downturns and can
thus point to a coming recession. The correlation between jobless
claims and the economy is not precise, however, in part because
unemployment statistics are distorted by the differing eligibility
requirements imposed by different states.
Manufacturers’ new orders for consumer goods and materials and for nondefense capital goods are excellent signs of how
businesses regard the coming economic climate (see Chapter 6).
Given the expenses involved in financing large purchases and in
carrying inventory, wholesalers and retailers don’t place orders for
consumer goods unless they foresee a demand for these products.
Similarly, companies don’t invest in costly capital goods unless they
believe they’ll need the additional production capacity or efficiency
created by such investments. Capital goods orders constitute a par-

Source:

Indices of Leading, Lagging, and Coincident Indicators



ticularly powerful leading indicator, because business investment
makes up approximately 15 percent of total GDP.
The vendor performance diffusion index is one of the five
seasonally adjusted diffusion indices that the Institute for Supply
Management uses to construct the Purchasing Managers’ Index
(PMI), the headline index of its monthly Manufacturing ISM Report on Business (see Chapter 5). The vendor index—which the
ISM creates from responses to its survey of approximately 400
purchasing managers across the United States—measures how long
it takes suppliers to deliver parts and materials that are integral to
the production process. Readings above 50 percent indicate slowing
deliveries, usually a sign of increased demand and robust economic
activity; readings below 50 percent indicate faster deliveries and
economic stagnation.
Statistics concerning building permits for new private housing are contained in the New Residential Construction report,
released jointly by the U.S. Department of Housing and Urban
Development and the U.S. Department of Commerce’s Census
Bureau (see Chapter 8). These data present insights into an element of the U.S. economy that both is crucial to its growth and
signals its general well-being. Although new housing construction
accounts directly for only a small percentage of GDP, it drives other
activity, such as purchases of paint, home furnishings, and countless
other consumer durables. Moreover, because buying a house is a
huge undertaking for most individuals, it implies confidence in the
stability of employment and earnings, as well as sound economic
fundamentals.
The stock price component of the LEI is the monthly average for the S&P 500 Index, published in the S&P publication The
Outlook. Inclusion of this has been questioned by some economists,
who argue that stock prices are determined by speculation rather
than by economic fundamentals and so should not be considered
accurate gauges of future economic activity. The rationale for including stock prices is that they reflect the informed expectations
of sophisticated traders and investors. To get ahead of the curve,
these knowledgeable market participants make their trades before



The Trader’s Guide to Key Economic Indicators

earnings are actually announced. Rising equity prices thus indicate
expectations of greater corporate profitability, which in turn implies
an expanding economy: When businesses are more profitable, they
are better able to invest in new projects, plants, and factories, and
to hire additional workers. Falling prices, conversely, indicate that
investors expect lower profitability, which in turn means a slower
economy.
Money supply is simply the amount of money in the economy.
The Federal Reserve recognizes three types of monetary aggregates,
which it labels M1, M2, and M3. The leading economic index uses
M2, which in addition to currency in circulation and deposits in
savings and checking accounts includes money market fund shares
and other liquid assets, such as overnight repurchase agreements issued by commercial banks.
Economists commonly refer to money as the oil in the engine of
economic activity. So it makes sense that the growth rate of the money
supply is related to the growth rate of the economy. The relationship
that associates money with economic activity is called the quantity
theory of money, which may be summed in the following expression: M x V = GDP, where M is the money supply; V is the velocity
of money—how often a dollar changes hands in a given period—
and GDP is the nominal gross domestic product (see Chapter 1).
Economists have assumed that velocity changes slowly, if at all,
over time. Given this assumption, any increase in the money supply would be mirrored by an increase in nominal GDP. Conversely,
a contraction in money supply would be reflected in a contracting
economy. This relationship is illustrated in the chart in FIGURE 2-2.
Economists have discovered, however, that the velocity of money
has not been constant. Without V as a constant, the equation of exchange breaks down. Historically, velocity levels have varied for a
number of reasons, including new regulations and innovations in
banking such as the advent of ATM machines, direct-deposit banking, and e-banking. Still, the relationship between money and economic activity enjoys a long, successful association, as evidenced in
the associated chart, and is therefore included in the index of leading economic indicators.

Source:

Indices of Leading, Lagging, and Coincident Indicators
Figure 2-2

Changes in Money Supply and in GDP

GDP YOY%
Sources: Board of Governors of the Federal Reserve System;
The Conference Board



M2 YOY%

10
8

GDP
M2-SA

–2

–4

0
1960

1970

1980

1990

2000

The interest rate spread component of the leading index is the
difference between the 10-year Treasury note yield and the Federal
funds rate—the rate banks charge one another on overnight loans
needed to meet reserve requirements set by the Federal Reserve.
For instance, if the Fed funds rate is 3.25 percent and the 10-year
Treasury is yielding 5.35 percent, the spread is 2.10 percent, or 210
basis points (a basis point is one-hundredth of a percent). The interest rate spread is included among the leading indicators because the
shape of the yield curve embodies fixed-income traders’ expectations about the economy, and interest rate spreads determine the
shape of the yield curve.
The yield curve plots yields of U.S. Treasury securities against
their maturities. Longer-term rates are usually higher than shorterterm ones, because more things can affect the value of the bond in
ten years than in two, and lenders require greater rewards for undertaking these greater risks. Thus, under “normal,” economically
favorable conditions, interest rate spreads are positive, and the shape
of the yield curve is gently convex—rising somewhat more steeply at
the short end and leveling off a bit at the longer maturities.
Steep curves—large spreads—may temporarily be the result of
current economic weakness. The Federal Reserve seeks to counter



The Trader’s Guide to Key Economic Indicators

such weakness by pushing down the overnight rate, thus lowering
borrowing costs and encouraging business investment and consumer spending on interest rate–sensitive goods and services like
housing and automobiles. This move stimulates the economy but
can spark inflationary fears among the fixed-income community.
Inflation erodes the value of future interest and principal payments.
In anticipation, fixed-income investors sell off longer-term (more
inflation-sensitive) bonds, depressing their prices and raising their
yields. This, combined with the Fed’s lowering of the short-term
rate, steepens the yield curve. Conversely, when the economy seems
to be running too hot, the Fed may seek to forestall a rise in inflation by raising its overnight target rate, discouraging spending and
so slowing growth. The result is a flatter curve (smaller spreads).
The curve may also invert, with short-term rates rising above
long-term ones and spreads falling below zero. This situation is
generally associated with economic downturns, even recessions,
as illustrated in FIGURE 2-3: The coincident index—which, because
it reflects current economic conditions, may serve as a proxy for
the business cycle—declines every time the spread between the
Fed funds rate and the 10-year Treasury becomes negative. This
close correlation is one reason the Conference Board decided to
The Interest Rate Spread and the Coincident Index

Ind

Spread

400

120120

200
0

9090

–200
6060

–400
Spread
Coincident

–600

3030
1960

–800
1970

1980

1990

2000

Sources: Board of Governors of the Federal Reserve System;
The Conference Board

Figure 2-3

Indices of Leading, Lagging, and Coincident Indicators



include this gauge in its index of leading economic indicators.
Why does an inverted yield curve predict recessions? There is
no definitive answer. Actually, one answer sometimes put forward
is that an inverted curve may result from the Fed overdoing it—
raising rates so high they not only cool but stifle growth (as well
as any fears of inflation). What is clear is that expectations of weak
economic conditions may encourage expectations of lower interest
rates. This in turn leads to more purchases of longer-term bonds,
pushing up their prices and lowering their yields. The result is an
inverted curve.
The Index of Consumer Expectations is compiled monthly,
along with the indices of Consumer Sentiment and Current Economic Conditions, by the University of Michigan’s Survey Research
Center, using responses to the university’s Survey of Consumers.
The survey asks consumers about their personal financial situations,
overall economic conditions, and their buying attitudes, as well as
various current issues and concerns. The Index of Consumer Expectations summarizes the economic trends the respondents foresee.
Including this index among the leading indicators is a no-brainer.
Consumer expectations about the economy are mainly shaped by
their experiences in the workplace. High confidence springs from
expanding employment, increased production schedules, and rising
wages and thus points to a positive economic climate. It also helps
foster that climate by encouraging spending, one of the major contributors to GDP. On the other hand, consumers are among the first
to sense worsening economic conditions, reflected in slowdowns at
their workplaces, and to retrench. This generally depresses the
economy further. Not surprisingly, then, the Index of Consumer
Expectations has a good record of predicting turning points in both
consumer spending and total economic activity.

Source:

LAGGING INDEX
The lagging index has the following seven components:
1. Average duration of unemployment
2. Ratio of manufacturing and trade inventories to sales



3.
4.
5.
6.
7.

The Trader’s Guide to Key Economic Indicators

Manufacturing labor cost per unit of output
Average prime rate
Commercial and industrial loans outstanding
Ratio of consumer installment credit to personal income
Change in the Consumer Price Index for services

The average duration of unemployment is the average number of weeks that people are out of work (see also Chapter 3). As
this number rises, so does consumer frustration, which depresses
spending and holds back economic growth. Decreases in the length
of unemployment traditionally occur after a recovery is already
under way. This is generally a function of business’s reluctance to
take on new workers until they are absolutely assured of recovery.
Similarly, the steepest rises generally take place after a downturn
has begun. That is why this is a lagging indicator.
The ratio of manufacturing and trade inventories to sales
is calculated by the U.S. Department of Commerce’s Census Bureau, using data from its Manufacturers’ Shipments, Inventories,
and Orders, or M3, survey (see Chapter 6) and its Wholesale Trade
Survey. The results are published in the Commerce Department’s
monthly Manufacturing and Trade Inventories and Sales (MTIS)
report (see Chapter 7). The ratio indicates how many months, given
the current pace of sales, it will take for inventories to be entirely
liquidated. A rising ratio means that businesses are unable to effect
a steady reduction in their back stock, either because sales are too
weak or their inventories are accumulating too fast. In either case,
this is a sign of economic weakness. A falling ratio, conversely, indicates that companies’ shelves are emptying and that manufacturers
may soon have to ramp up production to replenish their disappearing stocks—a bullish economic signal. Although economists watch
the ratio for insight into future production activity, it is a lagging
economic indicator. That’s because, historically, inventories rise
long after sales growth has halted. So the ratio reaches its peak in
the middle of a recession.
The percentage change in manufacturing labor cost per unit
of output is measured by an index constructed by the Conference

Source:

Indices of Leading, Lagging, and Coincident Indicators



Board from sources including the BEA’s seasonally adjusted data
on manufacturing employees’ compensation and by the Federal
Reserve Board’s data on manufacturing production. The index rises
when manufacturers’ labor costs increase faster than their output.
Because monthly index movements are erratic, the percentage
change used is measured over a six-month period. Peaks in the sixmonth rate of change are typically reached during recessions.
Data on the average monthly prime rate are compiled by the
Fed. As the interest rate that banks charge their most creditworthy
customers, such as blue-chip companies, the prime rate serves as
a benchmark for loans to lesser credits. For instance, a smaller,
younger company might have to pay two percentage points over
prime. Because the prime rate moves with respect to changes in the
federal funds overnight rate, periods of rising prime rates are usually
the result of rate hikes instituted by the Federal Reserve, foreseeing
a potential overheating in the economy and possible mounting inflationary pressures. Falling prime rates are usually the result of Fed
rate reductions in their overnight target rate, which are engendered
to stimulate economic activity. Banks tend to change the prime rate
only after movements occur in the general economy.
The value of outstanding commercial and industrial loans is
computed by the Fed and adjusted for inflation by the Conference
Board. High commercial and industrial loan levels are an indication
that businesses have a favorable economic outlook, and are willing
to build and expand their operations, and finance these plans via
loaned monies. Conversely, when the outlook is less encouraging,
and businesses skeptical, loan growth is weaker. It tends to reach a
peak after an expansion reaches its high-water mark and to bottom
out more than a year after the end of a recession.
The ratio of consumer installment credit to personal income
is computed using data from the Fed’s monthly release detailing the
amount of currently outstanding consumer credit as well as from
the BEA’s monthly Personal Income and Outlays report (see Chapter 11). Consumer credit is not included in the BEA income figure,
but for many Americans it is a critical income supplement. In times
of financial insecurity, such as those that occur during downturns



The Trader’s Guide to Key Economic Indicators

and recessions, people tend to reduce their personal borrowing and
don’t pick up the pace again until a trend of increasing income is
firmly established. Accordingly, this ratio generally reaches its nadir
a year or more after the end of a recession.
Change in the Consumer Price Index for services measures the
movement in the services component of the CPI, which is composed
monthly by the BLS (see Chapter 12). The month-to-month change
in the CPI is the most popular measure of inflation. Service-sector
inflation tends to increase after a recession has already begun and decrease even after it has ended. These tendencies result from what has
been termed recognition lags and other such rigidities in the market.

WHAT DOES IT ALL MEAN?
The Conference Board’s function in creating, refining, and maintaining the leading, lagging, and coincident indices, presented
monthly in its Business Cycle Indicators report, shouldn’t be confused with what the National Bureau of Economic Research does.
The NBER is the official arbiter of peaks and troughs in the business cycle. In pinpointing the dates of these crucial turning points,
the bureau’s economists consider many factors and consult several
indicators, including, but not limited to, components of the coincident index. The Conference Board’s Business Cycle Indicators
report does not determine the official peaks and troughs of the U.S.
economy. However, the turning points these indicators signal are
remarkably similar to those the NBER designates.

The coincident index is rarely mentioned in the business press. Still,
it is very useful for assessing the current pace of economic activity.
As the table in Figure 2-1 demonstrates, the coincident index tracks
closely the turning points in the business cycles. It can thus serve as
a benchmark in assessing the relationship of any economic statistic
to the business cycle. One of the most commonly used representatives of this cycle is the GDP. A simple linear regression between

Source:

COINCIDENT INDEX

Indices of Leading, Lagging, and Coincident Indicators

Sources: The Conference Board; U.S. Department of Commerce,
Bureau of Economic Analysis

Figure 2-4



Index of Coincident Economic Indicators and GDP

GDP YOY%

Coincident YOY%

10
GDP

Coincident

–2

–4

–6
1960

1970

1980

1990

2000

the growth rates of real GDP and the coincident index yields an impressive correlation of 86 percent. This close correlation, illustrated
in FIGURE 2-4, makes the coincident index a useful, more timely proxy
for the quarterly GDP.
LEADING ECONOMIC INDEX
The individual indicators composing the LEI differ considerably
in their abilities to predict economic turning points. Some are
very far-seeing, others relatively nearsighted. The composite index
combines these components in such away that the whole outperforms any of its parts. The predictive accuracy of the composite is
illustrated in FIGURE 2-5, which charts the quarterly year-over-year
percent change in the LEI against real GDP.
The chart clearly shows that hikes and dips in the LEI precede
those in the economy by significant periods. According to the latest research, the index’s average lead time is nine months. The
individual periods composing this average, however, vary considerable. This is in part because of the revisions that the index’s
components undergo, necessitating commensurate revisions in the
composite. It also reflects the fact that every recession and every



The Trader’s Guide to Key Economic Indicators

Figure 2-5

Index of Leading Economic Indicators and GDP
LEI YOY%

GDP YOY%

1010
GDP

LEI

GDP
LEI

–5

00
–2-2

Shaded areas = Recession

–4-4
1960

–10
–15

1970

1980

1990

2000

recovery is caused by different sets of circumstances. The LEI’s
ability to foresee these turning points therefore also varies.
LAGGING INDEX
The lagging index follows downturns in the business cycle (as
represented by the coincident index) by about three months and
expansions by about fifteen. At first blush, this may seem to be
pretty useless information—like driving a car by looking through
the rearview mirror. Economists, however, argue that you can’t
know where you’re going if you don’t know where you’ve been.
The index of lagging economic indicators confirms that turning
points in economic activity that were identified by the leading and
coincident indices actually have occurred. It thus helps prevent the
transmission of false signals.

HOW TO USE WHAT YOU SEE
Market participants don’t generally pay a great deal of attention to
the Conference Board’s Business Cycle Indicators report because
they’ve already had a chance to view and process for themselves

Sources: The Conference Board; U.S. Department of Commerce,
Bureau of Economic Analysis; NBER

Indices of Leading, Lagging, and Coincident Indicators



the underlying data. Nevertheless, economists and businesses have
traditionally looked for longer-term trends in the leading index to
predict turning points in the economy.
The old rule of thumb was that three consecutive monthly declines in the index signaled a recession within a year, whereas three
consecutive increases signaled a recovery. This rule was roughly
accurate. It did predict several recessions that failed to materialize, however, and in the case of some correct calls, the lead times
were negative—that is, the predictions came after the recession was
already established. A reason for false recession predictions could
be that although the index contains components representing the
manufacturing, consumer, financial, employment, and business investment sectors, it has none that reflect demand for or investment
and employment in the services industries that now dominate the
economy. Moreover, the financial sectors that are represented often
move in ways that don’t parallel movement in the broader economy,
generating both volatility and some of the false signals mentioned.
The LEI’s record of calling, as a popular quip has it, “seven of
the last five recessions” has led some cynics to term it the index
of misleading indicators. That’s not really fair. Still, to improve its
predictive accuracy, economists often consider the LEI’s moves in
three dimensions—duration, depth, and diffusion—instead of just
one, duration, as the three-month rule did. That is, in addition
to requiring that changes extend over three months, the refined
method looks at how large changes are and how many components
are involved. For example, if nine of the index’s ten components
show increases, but one—say the level of average weekly hours
worked in manufacturing—falls, an expansion is more certain than
if only four components increase, three decrease, and three are
unchanged.

Source:

TRICKS FROM

THE

TRENCHES

Wall Streeters, being innovators, have sought ways to improve even
on the three-dimensional analysis. Their trick, with respect to the
Business Cycle Indicators report, is to compute the ratio of the co-



The Trader’s Guide to Key Economic Indicators

Figure 2-6

Coincident-to-Lagging Ratio and Recessions*

120
110

90
80
70
Shaded areas = Recession

60
50
1970

1980

1990

2000

* as designated by the NBER

incident index to the lagging index. The theory behind this ratio,
informally referred to as the coincident-to-lagging index, is this:
In the early stages of a recovery, coincident indicators are rising
while lagging ones, reflecting the conditions of earlier months,
remain unchanged, resulting in a rising ratio. When an expansion
is peaking, both sets of indicators will be rising, but the rate of
increase for the coincident ones will be slower, so the ratio will fall.
Similarly, near the nadir of a recession, all the component indicators
will again be moving in the same direction—this time, down—but
the coincident ones will fall more slowly, so the ratio will rise.
As you can see from the chart in FIGURE 2-6, the coincident-tolagging ratio, like the LEI, has declined before every recession
since 1959. But it has transmitted fewer false signals. One explanation for this relative success is that the coincident and lagging
indices do a better job of representing, respectively, current and
past economic performance than the leading index does of assessing future activity.

Sources: The Conference Board; NBER

100

The Employment Situation

he most important economic indicator by far is the monthly
Employment Situation, published by the Bureau of Labor
Statistics (BLS). No economic release can move stocks and bonds
like employment, and no indicator is more revealing of general
economic conditions than labor market data. This is why the first
Friday of every month, when the Employment Situation is released,
is the most important trading session of the month.
The fixed-income market often moves violently, in a matter of
seconds, after the employment report is released. The Dow Jones
Industrial Average, which begins trading an hour after the report’s
release, has on occasion opened up or down from the previous day’s
close by a couple hundred points. The employment report is so
crucial to financial market participants that dealers, brokers, and
economists plan their vacations around its release. Many traders can
“make their month” (i.e., earn a month’s salary in a single trading
session) on the day the report is released. People have actually been
fired for missing the 8:30 a.m. EST release.
Employment data are important because they reveal how firms,
corporations, and others responsible for hiring decisions view the
current and upcoming economic environment. Companies will not
shoulder the expenses involved in adding to their payrolls if they believe they won’t need the extra workers in the near future. Similarly,
they will be reluctant to dismiss workers if they foresee increasing
demand for their wares. From the household perspective, nothing is
more important than employment status. It is said that consumers
can be expected to cut spending when faced with higher prices or
69



The Trader’s Guide to Key Economic Indicators

declining wealth. Truth be told, consumers will indeed continue to
spend despite higher prices and lower portfolio values. But nothing
erodes consumer attitudes and subsequently stops a consumer from
spending like the loss of a job. Because consumer expenditures account for about 70 percent of economic activity, every economist,
trader, and investor should know the current condition of the labor
market.
The monthly employment report is based on two separate surveys: the Current Population Survey (CPS), known as the household survey, and the Current Employment Statistics survey (CES),
referred to as the establishment, or payrolls, survey. The household
data are aggregated and disseminated in the “A” tables found in the
first half of the report; the establishment survey information is presented in the “B” tables. The Street tends to pay more attention to
the B than to the A tables.
Supplemental to each release, the commissioner of the Bureau of
Labor Statistics provides a statement to the Joint Economic Committee of the U.S. Congress. The statement, generally three pages
long, highlights significant strengths and weaknesses in the month’s
employment statistics.
The employment report contains several headliners, but top billing is generally shared by two figures: the unemployment rate and
the monthly change in nonfarm payrolls. Average hourly earnings,
hours worked, overtime hours worked, and the monthly change in
manufacturing jobs also command a great deal of Wall Street’s attention. Unlike many other economic releases, the Employment
Situation takes a great deal of time to digest.
When unexpected increases in the unemployment rate occur,
the equity markets generally sell off. The same occurs when nonfarm payrolls decline by a particularly large amount—usually in
the vicinity of 150,000 or more. Because employment determines
income and spending, and consumer spending accounts for the
largest portion of economic activity, traders like to see solid employment growth. When the unemployment rate declines and jobs
are being created, stock prices tend to rise.
Things are different in the fixed-income market, which is sensi-

Source:

The Employment Situation



tive to inflation threats. Increasing nonfarm payrolls and a falling
unemployment rate spark inflation fears, which can cause a sell-off
in bonds, depressing prices and raising yields.

Source:

EVOLUTION OF AN INDICATOR
Like many of the most respected economic indicators, the Employment Situation report was born in the 1930s, during the
Great Depression. The BLS had conducted the first monthly
studies of employment and payrolls in 1915, but these covered
only four manufacturing industries. By 1932, ninety-one manufacturing and fifteen nonmanufacturing industries were participating in the surveys. The deepening economic crisis of the early
1930s led the Hoover administration to expand the BLS program
to include working hours and earnings series. Statistics on average weekly hours and hourly earnings were published for the first
time in 1933. At the same time that this program was proceeding nationally, another was being rolled out on the state level.
In 1915, New York and Wisconsin entered into agreements with
the BLS to provide the agency with state employment data. This
pact grew to embrace all the states in the union plus the District
of Columbia and, today, Puerto Rico and the Virgin Islands. The
state and national efforts evolved into the Current Employment
Statistics survey, the source for Table B data in the employment
report.
The Current Population Survey, the source for Table A, started
as a program of the Work Projects Administration, or WPA,
which in 1940 initiated a national survey of households called the
Monthly Report of Unemployment. Responsibility for the survey
was transferred to the Census Bureau in late 1942, and a few years
later its name was changed to the Current Population Survey. In
1959, the Bureau of Labor Statistics, within the Census Bureau,
took over the job.
Both surveys have undergone refinements in sampling and
reporting techniques, incorporating advances in computer-aided
data-gathering and voice-recognition technologies. The result is



The Trader’s Guide to Key Economic Indicators

that today we have a timely, accurate, and comprehensive indicator
of labor market conditions, reported from both the employees’ and
the employers’ perspective.

In its employment surveys, the BLS includes only persons older
than sixteen. That seems logical, because most U.S. states have
compulsory education for youths through sixteen years of age, and
several states prohibit the employment of minors in many jobs. Also
excluded from surveys are people in mental or penal institutions
and members of the armed forces.
The monthly employment report includes figures for onemonth, three-month, six-month, and twelve-month periods. To understand the significance of these figures, you need to know what is
denoted by terms such as employed and unemployed. The meanings
may seem obvious, but BLS uses these and related words in quite
precise senses, developed through years of debate and experience.
People qualify as employed in two ways. First are those who,
during a given period, have worked as paid employees in someone
else’s company or in their own businesses or on their own farms or
have done fifteen hours or more of unpaid labor in a family-operated
enterprise. Second are those with jobs or in businesses from which
they have taken temporary leave, paid or unpaid, because of illness,
bad weather, vacation, child-care problems, labor disputes, maternity
or paternity leave, or other family or personal obligations.
Unemployed people are those not working during the period in
question, whether because they voluntarily terminated their employment, in which case they are classified as job leavers, or because
they were involuntarily laid off, making them job losers. Although
the report doesn’t make this distinction, economists identify several types of unemployment: Seasonal unemployment results from
short-term cyclical changes in the labor market; examples include
the January layoffs of retail staff who were added to take care of the
Christmas shopping push, and the winter furloughs of construction
and landscaping workers in regions where harsh weather make such

Source:

DIGGING FOR THE DATA

The Employment Situation



activity virtually impossible. Frictional unemployment refers to the
situation of workers in the process of changing occupations who
are temporarily between jobs. Structural unemployment is the result
of economic restructuring caused by new technologies or other innovations, as when the invention of the automobile put buggy-whip
makers out of a job. Finally, cyclical unemployment, the most relevant
type for Wall Street economists, occurs when jobs are eliminated as
part of the business cycle, because of declining demand and the
consequent drop in production.
To be included among the unemployed, a person must have
made an effort to find work. Those who have given up looking, believing their skills, qualifications, or geographic area preclude finding a job, are regarded as discouraged workers. Increasing numbers
of discouraged workers usually signal a weak economy.
Discouraged workers and others who don’t fit into either the
employed or unemployed groups are classified as “not in the labor
force.” The percentage of the employable population that is in the
labor force is known as the labor force participation rate. This
rate is generally in the mid-60 percent range. The employmentpopulation ratio is the percentage of employed persons in the total
population. It is usually lower than the participation rate.

Source:

HOUSEHOLD SURVEY (A TABLES)
Officially called the Current Population Survey, the household survey contains the responses of a sample of about 60,000 households
to questions about work and job searches. It is generally conducted
during the week containing the nineteenth day of the month. This
is known as the survey week. It addresses employment conditions
during the week containing the twelfth of the month, which is
known as the reference week. The statistics gathered are compiled
and presented in the following tables:
 Table A. Major indicators of labor market activity, seasonally
adjusted
 Table A-1. Employment status of the civilian population by
sex and age
















The Trader’s Guide to Key Economic Indicators

Table A-2. Employment status of the civilian population by
race, sex, and age
Table A-3. Employment status of the Hispanic or Latino
population by sex and age
Table A-4. Employment status of the civilian population 25
years and over by educational attainment
Table A-5. Employed persons by class of worker and parttime status
Table A-6. Selected employment indicators
Table A-7. Selected unemployment indicators, seasonally
adjusted
Table A-8. Unemployed persons by reason for unemployment
Table A-9. Unemployed persons by duration of unemployment
Table A-10. Employed and unemployed persons by occupation, not seasonally adjusted
Table A-11. Unemployed persons by industry, not seasonally adjusted
Table A-12. Alternative measures of labor underutilization
Table A-13. Persons not in the labor force and multiple jobholders by sex, not seasonally adjusted

The nation’s civilian unemployment rate is calculated by dividing the number of unemployed workers by the civilian labor force,
the figures of which are listed in the household survey. In June
2003, for example, the unemployment rate was computed to be 6.4
percent: 9.358 million unemployed divided by the 147.096 million
person labor force.

The establishment survey is based on a sample of about 160,000
businesses comprising some 400,000 individual work sites. Like
the household survey, it is conducted with respect to a reference
week, in this case the pay period containing the twelfth day of

Source:

ESTABLISHMENT SURVEY (B TABLES)

The Employment Situation



Source:

the month. The data are organized in the following tables:
 Table B-1. Employees on nonfarm payrolls by industry sector and selected industry detail
 Table B-2. Average weekly hours of production or nonsupervisory workers on nonfarm payrolls by industry sector
and selected industry detail
 Table B-3. Average hourly and weekly earnings of production or nonsupervisory workers on nonfarm payrolls by industry sector and selected industry detail
 Table B-4. Average hourly earnings of production or nonsupervisory workers on nonfarm payrolls by industry sector
and selected industry detail, seasonally adjusted
 Table B-5. Indices of aggregate weekly hours of production
or nonsupervisory workers on nonfarm payrolls by industry
sector and selected industry detail
 Table B-6. Indices of aggregate weekly payrolls of production or nonsupervisory workers on nonfarm payrolls by industry sector and selected industry detail
 Table B-7. Diffusion indices of employment change, seasonally adjusted
Nonfarm payrolls fall into two categories: goods-producing and
goods-providing. The goods-producing category includes manufacturing jobs, which account for 66 percent of the category total;
construction jobs, accounting for 30 percent; and jobs in natural
resources and mining, 4 percent. The majority of manufacturing positions are in the production of transportation equipment,
mostly motor vehicles. Other big manufacturing sectors are food
manufacturing, fabricated metal products, computer and electronic
products, machinery, and chemicals. The majority of construction
jobs are with specialty trade contractors, such as tradesmen engaged
in practices such as drywall and insulation, framing, roofing, siding,
electrical, masonry, and painting.
Over the past six decades, the U.S. economy has changed from
one based on manufacturing, with a heavily unionized labor force,
to one dominated by service industries. Service jobs, which fall in



The Trader’s Guide to Key Economic Indicators

Economists pay particular attention to the growth rate of total
private payrolls, that is, the number of employees on nonfarm and
nongovernmental payrolls. During periods of sub-par economic
growth, economists ideally wish to see widespread growth in payrolls across several industries. If job creation is limited to the government sector, it may be a signal that private industry is not very
confident with the economic environment and may not be willing
to hire new workers.
As shown by the chart in FIGURE 3-1, the level of employment
deduced from the household survey is different from, and generally higher than, that gleaned from the establishment survey. This
is largely because of the differences between their methodologies,
pools of respondents, sample sizes, and reference periods. That
said, the trends in employment revealed by the two surveys are
largely the same.
The data obtained in the establishment survey are used in
constructing the personal income report (aggregate earnings),
industrial production and capacity utilization report (aggregate
hours in manufacturing, mining, and public utilities), the Conference Board’s indices of leading and coincident economic indicators
(average weekly hours in manufacturing and employment, respectively), and the quarterly productivity measures (aggregate hours).
This survey attracts considerable attention from the investment

Source:

the goods-providing category, currently make up about 81 percent
of total nonfarm payrolls, compared with 56 percent during World
War II. Service payrolls are grouped into the following categories:
 government
 education and health services
 professional and business services
 retail trade
 leisure and hospitality
 finance, insurance, and real estate (FIRE)
 transportation and warehousing
 information
 other

The Employment Situation

Source: U.S. Department of Labor, Bureau of Labor Statistics

Figure 3-1



Employment Levels per Household Survey Versus Establishment Survey

In thousands

140,000
Household
Establishment

120,000
100,000
80,000
60,000
40,000
1948

1958

1968

1978

1988

1998

community and the business media. Because the data come directly
from corporations and firms, economists tend to trust these figures
more than the household statistics that are gathered via subjective,
and less scientific telephone interviews.

WHAT DOES IT ALL MEAN?

Source:

As has been noted, no indicator is as telling about the economic
state of affairs, on as timely a basis, as the BLS’s monthly Employment Situation. It is the first comprehensive report of the month
that presents both business and household perspectives on the most
important of all economic measures: employment.
Strong relationships exist between the employment data and
virtually every other economic indicator. The growth rate of nonfarm payrolls, for instance, is strongly correlated with the growth
rate of GDP, industrial production and capacity utilization, consumer confidence, spending, income—even with Federal Reserve
activity. If it’s relevant to economic activity, it will have links with
the payrolls data.



The Trader’s Guide to Key Economic Indicators

Payrolls (in thousands)

GDP YOY%

1,200

800

3
400

1
0

–400

Payrolls
Real GDP

–1

–800

–2
1985

1987

1989

1991

1993

1995

1997

1999

2001

2003

Sources: U.S. Department of Labor, Bureau of Labor Statistics;
U.S. Department of Commerce, Bureau of Economic Analysis

Figure 3-2 Nonfarm Payroll Growth and GDP

The ties between employment and the business cycle are extremely
close. As FIGURE 3-2 illustrates, the quarterly change in nonfarm payrolls has, in the main, hewed closely to the path of quarterly GDP
growth. This association is very useful for those monitoring economic growth. The GDP report is released quarterly, with a onemonth delay. The employment report is monthly. So those needing
a timely read on the economy can infer its growth rate from the
payrolls data.
One corollary of employment’s intimate relationship with
GDP growth is the coincidence of declining payrolls and recession. FIGURE 3-3 shows that, since 1960, there has never been an
instance when three consecutive monthly reductions in nonfarm
payrolls haven’t been accompanied by an economic downturn.
Conversely, each of the last ten post–World War II recessions was
characterized by at least three consecutive months—and most by
three consecutive quarters—of falling nonfarm payrolls.
In recoveries, unemployment is a lagging indicator—that is, it
continues to rise for several months after a definitive bottom has
been reached. The lag has become more marked in recent years. A

Source:

EMPLOYMENT, UNEMPLOYMENT, AND THE BUSINESS CYCLE

The Employment Situation



Figure 3-3 Growth Rate in Nonfarm Payrolls and Recessions
YOY%

6
Shaded areas = Recession

5
4
3
2
1
0
–1
–2
–3
1960

1970

1980

1990

2000

Sources: U.S. Department of Labor, Bureau of Labor Statistics; NBER

Figure 3-4 Unemployment Rate and Recessions
YOY%

12
Shaded areas = Recession

10
8
6
4
2
0
1960

1970

1980

1990

Source:

Sources: U.S. Department of Labor, Bureau of Labor Statistics; NBER

2000



The Trader’s Guide to Key Economic Indicators

look at FIGURES 3-3 AND 3-4 reveal that the last two U.S. recessions
(1990–91 and 2001) were both followed by protracted periods of
joblessness.

Average hourly earnings are considered a proxy for inflation and, as
such, are closely watched by fixed-income traders. When monthly
earnings are increasing at a torrential rate, the bond market usually
sells off, because inflation erodes the value of fixed-income holdings.
Inflation is inversely related to unemployment. That insight
grew out of a 1958 study by New Zealand–born economist A. W.
Phillips. The study demonstrated that when annual wage-growth
and unemployment rates were plotted against each other, the result
was a shallow convex curve. Known as the Phillips Curve, it served
to represent the fact that, during the period studied, years of low
unemployment coincided with rapid wage increases, whereas years
with high rates of unemployment saw inflation slow or even reverse.
This suggested a trade-off between inflation and unemployment
that could be exploited by economic policymakers: The central
bank could keep inflation low by accepting higher unemployment
levels or vice versa. This relationship soon became one of the most
disputed in economics.
The Phillips Curve and its policy implications led to the notion
of a non-accelerating inflation rate of unemployment. NAIRU, as it
is called, is the lowest level to which unemployment may fall without increasing the inflation rate. For decades economists believed
that NAIRU, sometimes referred to as the natural unemployment
rate, was around 6.0 percent. As the chart in FIGURE 3-5 graphically illustrates, particularly in the 1960s and 1970s, higher unemployment
rates—above 6.0 percent—were soon followed by periods of falling
inflation. Conversely, especially in the late 1980s, as employment
fell below 6.0 percent, inflation gathered steam. The hypothesis
pretty much lost its luster in the late 1990s, when the unemployment rate tumbled to 3.9 percent and inflation didn’t move higher.
In fact, inflation actually fell during the period, sparking fears of

Source:

INFLATION INDICATORS

The Employment Situation



Sources: U.S. Department of Labor, Bureau of Labor Statistics;
NBER

Figure 3-5 Unemployment Rate and Inflation
Unemployment

Inflation

16
Unemployment
Inflation

14
12

10
8

4
2
0

2
0
1961

1971

1981

1991

2001

deflation at the Federal Reserve and in the investment community.
Economists seeking another link between unemployment and
inflation latched on to the notion of “available labor pool.” This
is the number of unemployed who are actually available for work,
calculated by adding the number of unemployed (the traditional
measure defined as in the labor force, unemployed, and still looking for employment) to the number of workers not in the labor
force who currently want a job. The reasoning is that when the
pool of available workers begins to evaporate, employers have to
jack up wages to attract the employees they want, often having to
outbid other employers, and that this sparks widespread inflation.
The theory hasn’t been borne out, however. The available labor
pool shrank in the mid- to late 1990s to one of the lowest levels in
history, yet harmful rates of inflation never developed.

Source:

SENTIMENT AND UNEMPLOYMENT
President Truman once remarked: “It’s a recession when your
neighbor loses his job. It’s a depression when you lose yours.” The
barometer that best measures the downbeat sentiment associated
with job loss is the median duration of unemployment—that is,



The Trader’s Guide to Key Economic Indicators

Figure 3-6 Median Weeks of Unemployment Versus University of Michigan’s

Week

Sentiment

120

12
11
100

10
9
8

7
Weeks unemployment
U Michigan Sentiment

5
1993

1996

1999

2002

Sources: U.S. Department of Labor, Bureau of Labor Statistics;
University of Michigan

Consumer Sentiment Index

Source:

the median number of weeks that people are out of work. FIGURE 3-6
suggests that as the number of median weeks of unemployment
rises, workers become more and more frustrated, evidenced by
declines in the University of Michigan’s Consumer Sentiment
Index. Notice how the two spikes in the duration of unemployment during 2002 and 2003 were quickly followed by some of the
lowest levels of consumer sentiment in the decade. Keep in mind
that this measure is the median—the center point—not the average
number of weeks.
During extended periods of unemployment, marriages may be
strained and families disrupted. Depression is common. As underlying labor market conditions worsen, consumers become increasingly aware of how difficult it is to find a job. They hear of friends,
neighbors, and immediate family members losing their jobs. Then
distant relatives call asking about any available positions. Daily
reminders in the newspapers and evening news only increase the
gloom. These conditions aren’t exactly conducive to greater consumer spending and positive economic growth.

The Employment Situation



AVERAGE HOURS WORKED AND TEMPORARY WORKERS
The BLS calculates the aggregate weekly hours worked index,
formed by dividing the current month’s estimates of aggregate
hours by the corresponding annual average levels. This index, as
well as several sub-indices, are found in Table B-5 of the monthly
Employment Situation.
Because economic activity is basically a function of the number of
people employed and the amount of time they are working, economists have discovered the ability to arrive at a “synthetic” forecast for
economic growth, by charting the year-over-year percentage change
in the quarterly average of the aggregate hours index. Clearly, there is
an extremely tight correlation between these two indicators.
When economic conditions begin to sour, employers reduce the
number of hours worked before they eliminate staff. That way, if
economic activity recovers, they can ramp up production quickly by
merely adding hours, rather than having to spend time and money
finding and training new hires. This makes the aggregate hours
worked index, in some instances, a leading indicator of economic
growth. FIGURE 3-7 charts the average hours worked index over the
past two decades.

Sources: U.S. Department of Labor, Bureau of Labor Statistics;
NBER

Figure 3-7

Average Hours Worked Index and Recessions

Hours/Weeks

35.5
Shaded areas = Recession

35.0

34.5

34.0

33.5
1980

1983

1986

1989

1992

1995

1998

2001

The Trader’s Guide to Key Economic Indicators



–10
Shaded areas = Recession

–20
1991

1993

1995

1997

1999

2001

2003

If economic conditions—and profits—continue to deteriorate,
management’s next step is to reduce the number of workers on the
payroll. Among the first people to get pink slips are temporary, or
contingent, workers. The tasks they perform are not critical to the
day-to-day performance of the company—otherwise, they would
be employed full time. Furthermore, temp workers usually aren’t
unionized, so they can be cut most easily and cheaply during downturns. Temporary workers aren’t usually entitled to severance or
unemployment benefit insurance. Conversely, in the initial stages of
recovery, companies are not sure of future demand, so rather than
go through the costly process of hiring full-time workers, they add
temporary ones.
FIGURE 3-8 shows that a decline in temporary-help payrolls preceded the 2001 economic recession by a considerable degree, approximately six months. In general, increases in the growth rate of
temporary workers precede the upturn in the business cycle.
Knowing that businesses operate in this fashion, traders like to
keep a close eye on the trends in employment at temporary help
service establishments. An alternative manner at appreciating the
goings on in the temporary services sector may be identified by
watching the quarterly earnings announcements of the temporary

Source:

YOY%

Sources: U.S. Department of Labor, Bureau of Labor Statistics;
NBER

Figure 3-8 Temporary Employment and Recessions

The Employment Situation



staffing and recruiting agencies. Many companies, such as Adecco,
Kelly Services, Manpower, and Robert Half, provide a great deal
of information regarding industry trends, statistics, and forecasts.
Trade organizations such as the American Staffing Association provide timely outlooks and related publications on temporary help
and flexible staffing.

Source:

HOW TO USE WHAT YOU SEE
As always, the main strategy is to find ways to predict what a potentially market-moving number will be before it is released. Because
the payrolls figure is such an accurate indicator of economic activity,
for instance, economists and traders try to get ahead of the curve by
forecasting it. Many keep a journal by their desks, recording events
they come across in their daily reading that could influence employment. Unexpected disruptions like labor strikes, mass layoffs, and
natural disasters like hurricanes, tornados, floods, and blizzards can
greatly alter the number of workers in a given month.
Economists also watch a number of alternative indicators for
evidence to support or refute the developments suggested by data
in the employment report. One of these alternative indicators is the
index of monthly layoff announcements made by companies. The
index, compiled by the employment consulting firm Challenger,
Gray and Christmas, measures intended dismissals rather than
actual firings. It is thus something of a leading indicator. It gives
economists an insight into industries that may be experiencing difficulties. Movements in the index are also helpful in gauging the
bigger picture contained in the BLS employment report. That is,
increases in the number of layoff announcements usually portend a
softer payroll picture, while a decline in the number of announced
layoffs generally results in stronger payroll growth.
Another resource is the Help-Wanted Advertising Index.
Created and maintained by the Conference Board, it tracks the
monthly volume of help-wanted advertisements in the top fifty-one
newspapers across the nation, thus identifying regional demand for
labor. Since the advent of the Internet, however, businesses have



The Trader’s Guide to Key Economic Indicators

had other ways to advertise available positions, so the popularity of
the index has faded. Still, it can be helpful in determining general
trends in demand for workers.
Probably the most helpful resource for predicting movements in
monthly payrolls is the weekly claims for unemployment benefit insurance. Rising jobless claims usually portend a deteriorating labor
market. Many economists argue that when the four-week moving
average of claims tops 400,000, job creation is stagnant. Of course,
the correlation between claims for jobless benefits and the employment data is not precise. Employment conditions can change at any
time, and short-lived changes are more likely to show up in the
weekly jobless-claims report than in the monthly BLS employment
report. Also, although unemployment-insurance benefits generally
last only thirteen weeks, bear in mind that people can be out of
work for months at a time. Finally, some unemployed workers are
not entitled to jobless benefits.
THE

TRENCHES

This chapter’s trick is simply to call your attention to a littleknown, but very useful, section of the employment report: the
diffusion indices. The BLS provides diffusion indices for one-,
three-, six-, and twelve-month periods, both for private nonfarm
payrolls, which comprise 278 industries, and for manufacturing payrolls, which represent 84 industries. Economists tend to
gravitate to the one-month indices, as they are not as noisy as the
others.
The diffusion indices are derived from establishments’ responses to questions about whether they intend to add or eliminate
workers or leave payrolls unchanged. To calculate the indices, the
percent of responses indicating an intention to add workers is
added to half the percentage of the unchanged responses. When
the indices are above 50, indicating that a greater percentage of
industries intend to add workers than to lay them off or remain
stable, employment conditions are strong. High readings are usually accompanied by economic expansions. When the indices fall

Source:

TRICKS FROM

The Employment Situation



Sources: U.S. Department of Labor, Bureau of Labor Statistics;
NBER

Figure 3-9 Employment Diffusion Index and Recessions
90
80
70
60
50
40
30
20
10

Shaded areas = Recession

0
1991

1993

1995

1997

1999

2001

2003

Source:

below 50, industries are leaning toward cutting their payrolls. That
situation is typical of recession. As shown by FIGURE 3-9, most of the
recessions designated by the NBER coincide with sub-50 postings
in the BLS diffusion indices.

This page intentionally left blank

Industrial Production and
Capacity Utilization

he Industrial Production and Capacity Utilization
report, assembled and released around the fifteenth of each
month by the Board of Governors of the Federal Reserve System,
presents data on the output of the nation’s manufacturing, mining,
and utility sectors. The Federal Reserve’s G17 report, as it is also
known, organizes these data into industrial production and capacity utilization indices. The former measure the physical volume
of the output of various industries and markets; the latter show
what portion of the nation’s production capacity was involved
in creating that output. The total industrial production index is
usually the headline grabber. Investors may react strongly to the
monthly percentage change in this index, especially when it deviates from the Street’s consensus estimates. But market participants
also value the report’s individual-industry, aggregate-industry, and
market indices.
Equity analysts scan the Industrial Production and Capacity Utilization report for telling details about the condition of the chemical, home electronics, paper, textile, fabricated metals, lumber, or
industrial machinery sectors. Economists scrutinize the data for
early insights into what the quarterly National Income and Product
Accounts, or NIPAs (see Chapter 1), will say about the health of the
economy. Policymakers search the numbers for inflationary trends.
And the National Bureau of Economic Research (NBER), the official arbiter of U.S. business cycles, incorporates the industrial
production index—which it finds a good marker of the start and end
of recessions—in its statement on the status of the economy. The
89



The Trader’s Guide to Key Economic Indicators

Industrial Production and Capacity Utilization release, together
with historical data, is available on the Federal Reserve’s website,
www.federalreserve.gov.

The index of industrial production is among the oldest measures
of U.S. manufacturing and macroeconomic activity, predating even
the venerable National Income and Product Accounts. It is as old,
in fact, as the Federal Reserve Board of Governors. Soon after the
central bank was created, with the passage of the Federal Reserve
Act on December 23, 1913, its officials realized that they needed
a pool of accurate economic data and measurements if they were
to steer the nation through the widespread bank failures, frequent
recessions, and bouts of joblessness that characterized the late nineteenth and early twentieth centuries. No formal measurement of
economic activity existed, however, in 1913. The National Bureau
of Economic Research, now the primary agency responsible for
maintaining the national accounts, didn’t begin estimating income
until 1920. So the Fed created its own Physical Volume of Trade
report. This consisted of unaggregated indices tracking the production of commodities such as coal, coke, crude oil, steel, textiles,
metals, and paper, measured in tons, feet, barrels, and other relevant
physical units. By 1919, these indices represented seven sectors of
the economy: agriculture, forestry, mining, manufacturing, trade
and transportation, banking and finance, and labor.
In 1922, the Fed introduced the monthly Indexes of Domestic
Business report, which combined fifty-five gauges of commodity
production into indices representing three facets of production:
agriculture, mining, and manufacturing. Later that year, it created
the more detailed Index of Production in Selected Basic Industries.
This was composed of twenty-two commodities weighted according to the level of employment in manufacturing process of each
respective commodity industry, which was obtained in the 1919
census, and the value added by each industry—that is, the portion of
a good’s final value that is contributed by the industry in the course

Source:

EVOLUTION OF AN INDICATOR

Industrial Production and Capacity Utilization



of production. The two largest components in the index were pig
iron and cotton, with weights of 18 and 15 percent, respectively. In
structure it was similar to the industrial production index of today.
The industrial production index was used to estimate industrial
capacity utilization for about two decades after World War II. In
the 1960s, the Federal Reserve developed a process for estimating industry capacity and capacity utilization rates, releasing these
estimates in a separate statistical report called Capacity Utilization:
Manufacturing, Mining, Utilities and Industrial Materials. This
monthly report, known as statistical release G.3, was released one
business day after the industrial production report. In addition to
surveys of industry businessmen, capacity indices were estimated
using data obtained from McGraw-Hill and the U.S. Bureau of the
Census surveys of plant capacity.
Most of the meaningful changes to the capacity utilization measures occurred in the 1970s, particularly in 1974 with the revision of
the materials measures, and then again in 1976 with the augmentation of total materials data in the industrial production index.
In December 2002, the Federal Reserve conducted a revision of
the industrial-production and capacity-utilization measures. This
consisted primarily of switching from the Standard Industrial Classification (SIC) to the new North American Industry Classification
System (NAICS). It also introduced more reliable methods of calculating manufacturing activity in the communication equipment,
semiconductor, light vehicle, and newspaper industries and regrouped the major market classes according to stage of processing,
very much like the system used by the producer price index (PPI).

Source:

DIGGING FOR THE DATA
The Industrial Production and Capacity Utilization report is an assemblage of fifteen tables arranged over nineteen or twenty pages.
The tables display the current month’s values for the various industrial-production and capacity-utilization indices, revisions to previous months’ values, month-to-month percentage changes in the
indices, and their quarterly and annual rates of growth. The front



The Trader’s Guide to Key Economic Indicators

page of the release contains a summary of the most important indicators—the monthly percentage changes in the total industrial production index and the capacity utilization rate—as well as revisions
to the three previous months. The Federal Reserve also provides a
number of detailed charts—a trader’s best friend—before the meat
of the report is presented.

The industrial production indices measure quantity of output (i.e., in
terms of production units like tons, cubic feet, or kilowatt hours), not
dollar volume, relative to a base year, currently 1997, whose value is
set at 100. An index value of 109, for instance, denotes that output for
that month was 9 percent higher than the average for 1997.
The Federal Reserve obtains the production data it uses to
construct these indices both directly and indirectly. Direct sources
include trade associations such as the American Forest and Paper
Association (for pulp, wood, and paper and paperboard output), the
U.S. Geological Survey (for copper, lead, zinc, gold, and silver ore
numbers), the Internal Revenue Service (for beer, wine, and brandy),
and the Tanner’s Council of America (for leather and belting figures). Actual production data, however, are available at different
times for different industries. When hard figures aren’t available,
the Federal Reserve estimates output based on the number of production-worker hours in the Bureau of Labor Statistics’ monthly
Employment Situation report or on electric power use by industry.
Only a few sectors—usually motor vehicles, steel and other metals,
lumber, and paper—have hard figures ready for the initial release.
By the third monthly revision, actual physical production accounts
for 46 percent of the data by value added, production-worker hours
for 31 percent, electric power use for 19 percent, and Federal Reserve judgments based on anecdotal evidence for 4 percent.
The total industrial production index is constructed from 295
components, or individual series such as copper, instruments, computers, or lumber, each of which is weighted according to the value
of the associated sector added during the production process in the

Source:

INDUSTRIAL PRODUCTION

Industrial Production and Capacity Utilization

Source: Board of Governors of the
Federal Reserve System

Figure 4-1



Industry Groups
2002 Proportion

Total Industrial Production
Manufacturing
Durable
Nondurable
Other manufacturing (non-NAICS)
Mining
Utilities

100.00
84.62
43.41
35.90
5.30
6.03
9.34

base year. Each individual series is expressed in its own respective
quantity (i.e., steel in tons, automobiles in units) so that monthto-month changes in production are measured without respect to
price movements. The report presents these components according
to two different classification schemes: by industry, representing
the supplier perspective; and by market, representing the demand
perspective.
The industry schema is based on the North American Industry
Classification System (NAICS). The three primary industry groups
are manufacturing, mining, and utilities. Manufacturing is subdivided into durable and nondurable goods. The table in FIGURE 4-1
shows the major groupings together with the percentage each contributed to production growth in 2002.
In the classification by market groups, shown in FIGURE 4-2,
the total index is divided into two major groups: final products/
Figure 4-2

Market Groups

Source: Board of Governors of the Federal
Reserve System

2002 Proportion

Total Industrial Production
Final product and nonindustrial supplies
Consumer goods
Durable
Nondurable
Business Equipment
Defense and space equipment
Construction supplies
Business supplies
Materials

100.00
59.81
30.46
7.52
22.93
9.65
2.20
6.84
0.30
40.19



The Trader’s Guide to Key Economic Indicators

nonindustrial supplies and materials. Final products/nonindustrial
supplies is itself divided into consumer goods (further subdivided
into durable and nondurable), business equipment, defense and
space equipment, construction supplies, and business supplies.

Capacity utilization is a measure of how close the nation’s manufacturing sector is to running at full capacity. Formally, it is the ratio
of the index of industrial production to an index of full capacity. But
what is meant by full capacity? The Fed defines it as sustainable
practical capacity, or “the greatest level of output that a plant can
maintain within the framework of a realistic work schedule, taking
account of normal downtime and assuming sufficient availability of
inputs to operate the machinery and equipment in place.”
The annual full capacity number is derived through a complex
process that involves both hard data, obtained from industry surveys
such as the U.S. Census Bureau’s Annual Survey of Plant Capacity,
which all businesses (with a class D SIC code classification, manufacturing) with five employees or more must complete, and by inference,
using the ratio given above. The Fed assumes that month-to-month
growth is smooth and so derives the monthly capacity figure by
straight-line interpolation from the annual number.
The monthly capacity utilization rate is derived by dividing the
monthly industrial production number by the monthly capacity figure. For example, during July 2003 the index of industrial production was 98.6 and the capacity index was 132.7. Dividing the former
by the latter (98.6 / 132.7) results in a capacity utilization rate of
74.3 percent. In layman’s terms, this suggests that factories were
running at 74.3 percent of full capacity.
The report contains capacity and capacity utilization rates for
eighty-five industries, including the following major categories:
 semiconductors and related electronic components
 motor vehicles and parts
 apparel and leather
 paper

Source:

CAPACITY UTILIZATION

Industrial Production and Capacity Utilization



chemicals
wood products
 electric utilities



The total capacity utilization rate is compiled from these components, weighted as shown in FIGURE 4-3.

Source: Board of Governors of the
Federal Reserve System

Figure 4-3 Capacity Utilization Percentage of Capacity, Seasonally Adjusted
2002 Proportion

Total Industry
Manufacturing
Durable
Nondurable
Other manufacturing
Mining
Utilities

100.00
86.69
47.13
34.66
4.89
5.20
8.12

WHAT DOES IT ALL MEAN?
Economists, analysts, and investors look to the Industrial Production and Capacity Utilization report for timely indications of
overall economic health as well as manufacturing and inflationary
trends. The two main sections of the report provide different types
of information and signals.
INDUSTRIAL PRODUCTION
The index of industrial production is procyclical—that is, it moves in
unison with the business cycle. As the chart in FIGURE 4-4 illustrates,
the correlation between the index and economic activity is quite
tight—so tight, in fact, that the monthly index is used as a more
timely proxy for the quarterly GDP report.
The National Bureau of Economic Research uses the index to
discern turning points in the business cycle. As the chart in FIGURE 4-5
shows, each of the NBER-designated recessions since 1950 has
coincided with a precipitous drop in the 12-month growth rate



The Trader’s Guide to Key Economic Indicators

GDP $ in billions

10,500
Real GDP
Production

115

9,000

105
95

7,500

85
75

6,000

Shaded areas = Recession

4,500

55
1980

1985

1990

1995

2000

of industrial production. The converse, however, is not necessarily true: The manufacturing sector can be in recession while the
broader economy continues to prosper.
At first blush, the close relationship between the industrial production index and the overall economy seems odd. Manufacturing,
after all, accounts for only 20 percent of total economic activity.
The United States has evolved from a smokestack to a servicesdominant economy, and mammoth service industries, such as health
care, software, telecommunications, travel and entertainment, pharmaceuticals, and banking and finance, are not directly represented
in the industrial production index. Nor does the index contain any
measure of construction-related activity, although it does represent
industries that manufacture construction machinery such as wheel
loaders and wheel tractors.
On closer examination, however, the close relationship between the index and the broad economy makes sense. For starters,
one-fifth of total activity isn’t altogether small. The U.S. manufacturing sector moves very much in line with aggregate demand,
just as retail sales, another small and equally cyclical statistic,
moves with consumer spending. Moreover, manufacturing and
production activity have a large “multiplier effect.” Manufactures,

Source:

125

Sources: Board of Governors of the Federal Reserve System; U.S.
Department of Commerce, Bureau of Economic Analysis; NBER

Figure 4-4 Industrial Production and Real GDP

Industrial Production and Capacity Utilization



Source:

Sources: Board of Governors of the Federal Reserve System; NBER

Figure 4-5 Industrial Production and Recessions
YOY%

30
Shaded areas = Recession

25
20
15
10
5
0
–5
–10
–15
1950

1955

1960

1965

1970

1975

1980

1985

1990

1995

2000

especially durable goods—those with expected shelf lives of three
or more years—generally come with service contracts that cover
the cost of repairs, damages, and maintenance. Obviously, the
value and level of such services are driven by the demand for the
products themselves.
Finally, and most crucially, many service industries are large
consumers of manufactures. Taxi companies require automobiles;
airlines need jets and electronic security systems; bars and restaurants use refrigerators, dishwashers, ovens, and foodstuffs. Financial
institutions and law firms are big customers for computers and peripherals, and the health care industry is one of the largest consumers of industrial products, including operating tables, beds, lamps,
imaging machines, and surgical supplies.
The industrial production index shows another strong correlation—with the Purchasing Managers’ Index (PMI), published by the
Institute for Supply Management (ISM). The relationship between
the two indices is illustrated in the scatter chart in FIGURE 4-6. The
points in the upper-right quadrant represent periods of extremely
strong manufacturing output, such as recoveries from recession or
the 1991–2001 economic expansion, the longest in U.S. history.
The points in the lower-left quadrant represent periods of severe



The Trader’s Guide to Key Economic Indicators

Figure 4-6 Correlation of Purchasing Managers’ Index and Industrial

70
65
60
55
50
45
40
35
30
–10

–5

Industrial Production (YOY%)

manufacturing weakness, generally recessions. A simple econometric analysis suggests that when the PMI is at 50, a level generally consistent with an expansion in manufacturing, the industrial
production index is advancing at an annual rate of approximately
2 percent.
At a more detailed level, the industrial production sub-indices
can depict, and in some instances explain, what is happening in
specific industries. Take the spectacular expansion and equally
spectacular collapse in the technology sector that occurred during
the late 1990s and early 2000s. The chart in FIGURE 4-7 shows that
while high tech was soaring, the production of nontechnology
goods was plodding along, posting average annual gains of around
2 percent.
It was during this period that Federal Reserve chairman Alan
Greenspan uttered his famous “irrational exuberance” comment
during an American Enterprise Institute Francis Boyer Lecture.
Although he was referring to asset prices, Chairman Greenspan
was doubtless aware that production of high-tech goods such as
computers, communications equipment, and semiconductors was

Source:

ISM's PMI

Sources: Board of Governors of the Federal Reserve System; ISM

Production Index

Industrial Production and Capacity Utilization



Figure 4-7 Industrial Production Index: High-Tech and Excluding High-Tech

Source: Board of Governors of the Federal Reserve System

YOY%

60
50

Ex-High Tech
High Tech

40
30
20
10
0
–10
1987

1989

1991

1993

1995

1997

1999

2001

2003

Source:

growing at more than 40 percent annually, dwarfing the 15 percent
or so registered a mere year and a half earlier.
That pace was not sustainable. Beginning in late 2000, the
chart shows, there was a precipitous decline in the high-tech
growth rate. Manufacturers, afraid to miss out on the hot market
for their goods, had overproduced. When demand, inevitably,
slackened, these companies—particularly telecommunicationsequipment providers—were left with record inventories. In
reaction, they slashed production and, consequently, staff until
consumers could draw down existing stock. Inventory depletion
took the better part of three years. This situation, exacerbated by
fraudulent accounting practices, misstated earnings, and countless corporate improprieties, forced Global Crossing, WorldCom, and Qwest, among other companies, into bankruptcy. Not
surprisingly, the stock market plunged between 2000 and 2002.
This entire story can be traced in the contrasting paths of the
high-tech and ex-tech indices shown in the figure.

100



The Trader’s Guide to Key Economic Indicators

CAPACITY UTILIZATION
Economists, particularly central bankers, look at the total capacity
utilization rate to discern trends in production, general economic
activity, manufacturing conditions, and inflation. In addition, the
rates for particular industries can pinpoint areas of overcapacitization (production that pushes capacity to its limit) that could become
manufacturing bottlenecks, constraining production farther down
the line and possibly pushing up prices. Such information is useful not only to economists but also to company managers trying to
forecast costs and plan production schedules.
Low levels of capacity utilization—78 percent or below—
indicate that the economy is headed to, or already in, recession.
In fact, as the chart in FIGURE 4-8 illustrates, each of the last six economic recessions was characterized by rates in that range. This
relationship is logical: Subpar economic conditions simply don’t
warrant strong production.
When demand and commerce are booming, on the other hand,
factories tend to ramp up and produce at rates closer to their capacity. The downside to this is that the higher production rates tend to
stoke inflation.

Percent

90
Shaded areas = Recession

70
1967

1970

1973

1976

1979

1982

1985

1988

1991

1994

1997

2000

2003

Sources: Board of Governors of the Federal Reserve System; NBER

Figure 4-8 Capacity Utilization and Recessions

Source:

Industrial Production and Capacity Utilization



101

When factories approach their maximum production potential, machinery and other goods-producing capital are strained.
As a result, electronic components may short, pumps overheat,
lubricants dry up, or core parts crack. If the overworked equipment cannot be repaired in a reasonable, profitable period of time,
mass lay-offs could ensue. Complicating matters is the fact that all
this occurs just when demand is greatest and increased production
most necessary. To understand how this cycle can spark inflation,
consider the case of a company that produces about 50,000 tons of
cement a week.
During normal conditions the company’s plants operate at 80
percent of capacity. At that rate, market conditions dictate a price
of $50 a ton for the cement. Over time, though, a housing boom
develops, and demand for cement surges. To capture more of the
increased market, the company begins operating its plants at 92
percent of their full capacity. At this rate, the rotating kiln breaks
down because it is not accustomed to so much pressure. The damage may take weeks, even months to repair. Beyond laying off the
workers that operated the broken machinery, what can the company
do to offset the lost revenue?
Answer: Raise prices. Rather than charge the normal market rate
of $50 a ton, the cement manufacturer tells wholesalers that it expects $55, $60, or even $65 a ton. Because demand has skyrocketed,
the wholesalers will gladly fork over the additional money, knowing
that they too will be able to pass along the increase, to contractors and retailers. These price hikes are transferred to new homes,
driveways, sidewalks, and highways and eventually into the general
economy.
As this example illustrates, during periods of high capacity utilization, inflationary pressures mount. The economic consequences
can be serious. Inflation erodes the purchasing power of bonds’
coupon and principal payments, depressing their prices and raising
yields. Higher interest rates, in turn, impede future investment.
Concerned over this vicious cycle, economists have sought to quantify the relationship between the capacity utilization rate and inflation, just as they did for the unemployment rate (see Chapter 3).

102



The Trader’s Guide to Key Economic Indicators

PPI YOY%

–4

74
72

–8

Mfg CU
PPI

–12

68
1975

–16
1978

1981

1984

1987

1990

1993

1996

1999

2002

And just as they identified a minimum level of unemployment that
could be sustained without sparking inflation—the non-accelerating inflation rate of unemployment, or NAIRU—they have identified a maximum non-accelerating inflationary rate of capacity
utilization, or NAICU.
The NAICU for the manufacturing sector as a whole has long
been accepted to be 84 percent. Individual businesses differ, however, in their susceptibility to technical innovations, legal barriers,
work stoppages, and cyclical abandonment rates, all of which can
drastically affect total capacity. Each industry thus has its own
NAICU. Some rates are higher than 84 percent, like the paper
industry’s 87 percent; some are lower, like the mining sector’s 80
percent. Of course, none of these is set in stone. Changing business
conditions can alter particular NAICUs. It is thus best to regard
individual industry numbers as shorthands for ranges, rather than
as precise rates.
Even with this looser definition, the applicability of NAICUs
in today’s economy has been called into question, much as the applicability of the NAIRU has been. The chart in FIGURE 4-9 shows
why. The expected relationship between the general manufacturing
NAICU and inflation appears to exist through the late 1980s, but

Source:

CU%

Sources: Board of Governors of the Federal Reserve System;
U.S. Department of Labor, Bureau of Labor Statistics

Figure 4-9 Manufacturing Capacity Utilization and PPI for Intermediate Goods

Source:

Industrial Production and Capacity Utilization



103

then disappears. A sustained period of capacity utilization above 84
percent that occurred in the late 1970s was indeed followed by an
acceleration in the growth of the Producer Price Index for intermediate materials; likewise, depressed capacity utilization rates during
the early 1980s did precede a period of disinflation—positive but
slowing price growth—and even deflation, or falling prices. In the
1990s and early 2000s, however, a disconnect occurred, with capacity utilization rising to fairly high rates without spawning inflation.
Does this mean that the theory no longer holds and that high
capacity utilization rates do not increase inflationary pressures?
Probably not. It is more likely that the breakdown in the relationship illustrated in the chart occurred because of the heavy business
investment in productivity-enhancing technologies that took place
in the latter half of the 1990s, increasing manufacturing capacity. The substitution of low-priced imports for domestic products
played an important role, as well, by keeping manufacturing costs
low and inflation in check. That said, it is also possible that the true
capacity utilization threshold may be a tick or two higher than 84
percent, say 85 or 86 percent.
One indication that the general reasoning behind NAICUs still
holds is the reaction of the Federal Reserve to high capacity utilization readings. The Fed is concerned with keeping a lid on inflation
and adjusts its monetary policy accordingly. At the first signs of an
overheating economy, it generally raises its target for the Fed funds
rate (the interest rate banks charge each other for overnight loans
used to meet reserve requirements; see Chapter 1). This increase
eventually extends throughout the maturity spectrum, discouraging
borrowing and so slowing the pace of investment and production.
If the Fed governors are still using NAICU as an inflation indicator, you’d expect hikes in the Fed funds rate to correspond to high
rates of capacity utilization. And that does seem to be the case. The
chart in FIGURE 4-10 shows that from 1989 through 2003, whenever
the capacity utilization rate rose into the high 80s, the Fed funds
rate rose, as well.
The conclusion: Although the relationship has loosened in recent years, the Federal Reserve clearly believes that capacity utiliza-

104



The Trader’s Guide to Key Economic Indicators

Figure 4-10 Capacity Utilization and Growth in Fed Funds Rate
CU%

120

86.0

100

84.0

80
60

82.0

80.0

20
0

78.0

–20
–40

76.0

Fed funds YOY% change
Capacity utilization

–60

74.0

–80
1989

1991

1993

1995

1997

1999

2001

2003

Sources: Board of Governors of the Federal Reserve System

FF YOY%

tion is still a powerful inflation marker and watches the reported
rate carefully. That’s a good enough reason for traders, especially
those in the fixed-income market, to keep a close eye on capacity
utilization, too.

HOW TO USE WHAT YOU SEE
As with most indicators, Wall Streeters want to get a preview of what’s
inside the Industrial Production and Capacity Utilization report: The
earlier they can approximate the industrial production index readings, the earlier they can capitalize on any anomalies in the numbers.
Useful tools in this project are the number of worker hours in the
Department of Labor’s monthly employment report. This chapter’s
Tricks describe how market participants use the Labor Department’s
data in forecasting the industrial production indexes.
THE

TRENCHES

As noted above, when the actual data are not available, the Federal
Reserve may estimate industrial production based on the number
of production-worker hours. Wall Street economists use the same

Source:

TRICK FROM

Industrial Production and Capacity Utilization



105

statistics in a simple back-of-the-envelope calculation to predict the
industrial production index reading up to two weeks before it is released. Here’s what the computation would look like using the data
in the Bureau of Labor Statistics’ March 2002 Employment Situation report, released on April 5 of that year. The BLS report stated
that in February, 16.869 million people were employed in manufacturing, working an average of 40.7 hours a week, whereas in March,
16.831 people worked 41.1 hours a week on average. The first step
in the computation is to calculate the total number of manpower
hours (m/h) worked in each month:
February: 16.869 million workers x 40.7 hours = 686.5683 m/h
March: 16.831 million workers x 41.1 hours = 691.7541 m/h

The next step is to determine how many more or fewer man
hours were worked in March:
691.7541 – 686.5683 = 5.1858

The last is to derive the percent month-over-month change:
(5.1858 ÷ 691.7541) x 100 = 0.7496 percent.

Source:

So the predicted change in the March industrial production
index, based on the number of manufacturing workers and the
hours they worked, is 0.75 percent. The actual index reading in
the report released on April 16 was 138.8, a 0.65 percent increase
over the revised February reading of 137.9. Of course, the result
of the calculation is not always that close to the actual reading. But
the method is certainly simple, and it is helpful in determining the
direction, if not always the magnitude, of change in the index—
an extremely important statistic.

This page intentionally left blank

Institute for Supply
Management Indices

he Purchasing Managers’ Index (PMI) garners more attention than any other economic release except the monthly
Employment Situation report (see Chapter 3). Markets move considerably on its readings, and rumor has it that the index is Federal
Reserve chairman Alan Greenspan’s Desert Island Statistic—that is,
if he were stranded on an island and needed to conduct policy with
respect to only one economic indicator, this would be it.
The PMI is the headline index of the Manufacturing ISM Report
on Business. This report is created by the Tempe, Arizona–based
Institute for Supply Management (ISM), a not-for-profit professional association, and is made available on ISM’s website (at
www.ism.ws) on the first business day of every month, after 10 a.m.
ET. In addition to providing a comprehensive introduction of the
various indices contained in the report, the website houses a complete historical data set subject to timely revisions and updates.
The ISM’s Report on Business describes and discusses the current
readings of ten seasonally adjusted diffusion indices constructed by the
ISM from the responses to a survey of approximately 400 purchasing
managers across the United States. Conducted about the middle of
the previous month, it polls participants about their opinions on prices
of materials paid in the production process, production level, new orders, order backlog, the speed of supplier deliveries, inventories, customer inventories, employment, new export orders, and imports. The
PMI is a weighted composite of the following five of these indices:
 new orders
 production
107

108



The Trader’s Guide to Key Economic Indicators

employment
vendor performance
 inventories



Economists look at the PMI primarily to determine the health of
the manufacturing sector, but the index provides an accurate picture
of the broader economy as well. Its movements closely parallel those
of the index of leading economic indicators (LEI, see Chapter 2), and
it has shown an uncanny ability to predict recessions months before
the National Bureau of Economic Research (NBER) declares them.
This may not be so surprising: Consider, what better way is there
to find out about manufacturing activity and associated spending
than from those business people who are responsible for making the
purchasing decisions for the nation’s manufacturers?
According to research conducted by the Federal Reserve Bank
of New York, the financial markets, in particular the fixed-income
market, react very strongly to the monthly postings of the ISM’s
PMI. On several occasions, the PMI has been identified as the biggest market mover in the monthly reporting cycle of indicators.

The Manufacturing Report on Business has its origins in the 1920s,
when the ISM’s predecessor association (first called the National
Association of Purchasing Agents, then, the National Association
of Purchasing Management) began polling members sporadically—
at first, only about commodity availability across the country, but
soon about other types of information, as well. In 1930, the association was part of a committee formed by the Chamber of Commerce
(CoC) under President Herbert Hoover in response to the stock market collapse and the Great Depression. The committee was charged
with collecting business data from CoC members. Although the
committee was disbanded in 1931, the association decided to carry on
the project and, with encouragement from the government, started
conducting surveys and publishing the results on a regular basis.
This project has continued ever since except for a four-year

Source:

EVOLUTION OF AN INDICATOR

Institute for Supply Management Indices



109

hiatus during World War II, when the spotty availability of data
disrupted publication. The PMI—also known as the ISM index
or, among old-timers, the NAPM index (in reference to the association’s name until May 2001)—has been part of the association’s
monthly report since 1948.

Source:

DIGGING FOR THE DATA
As noted above, the ten diffusion indices contained in the Report on
Business reflect the survey responses of purchasing managers from
various regions of the United States. The participants are drawn
from twenty manufacturing industries, based on the Standard Industrial Classification (SIC) system, with each industry represented
according to its contribution to gross domestic product (GDP).
The twenty industries include the following:
 Apparel
 Chemicals
 Electronic Components & Equipment
 Fabricated Metals
 Food
 Furniture
 Glass, Stone & Aggregate
 Industrial & Commercial Equipment & Computers
 Leather
 Miscellaneous (e.g., jewelry, toys, sporting equipment, and
musical instruments)
 Paper
 Petroleum
 Photographic Equipment
 Primary Metals
 Printing & Publishing
 Rubber & Plastic Products
 Textiles
 Tobacco
 Transportation & Equipment
 Wood & Wood Products

110



The Trader’s Guide to Key Economic Indicators

The Institute for Supply Management™ 2003 Survey Questions
1. GENERAL REMARKS: Comment regarding any business condition (local, national, or international) that affects your purchasing operation or the outlook for
your company or industry. Your opinion and comments are very important.
2. PRODUCTION:
Check the ONE box that best expresses the current month’s level (units, not dollars) compared to the previous month.
 Better than a month ago  Same as a month ago  Worse than a month ago

3. NEW ORDERS:
Check the ONE box that best expresses the current month’s new orders (units, not
dollars) compared to the previous month.
 Better than a month ago  Same as a month ago  Worse than a month ago

4. BACKLOG OF ORDERS:
Check the ONE box that best expresses the current month’s backlog of orders
(unfilled sales orders) (units, not dollars) compared to the previous month.
 Do not measure backlog of orders
 Same as a month ago

 Greater than a month ago
 Less than a month ago

5. NEW EXPORT ORDERS:
Check the ONE box that best expresses the current month’s new export orders
(units, not dollars) compared to the previous month.
 Do not export
 Same as a month ago

 Better than a month ago
 Worse than a month ago

6. COMMODITY PRICES:
Check the ONE box that indicates the current month’s level of change in approximate net weighted average prices of the commodities you buy compared to the
previous month.
 Higher than a month ago  Same as a month ago  Lower than a month ago

List specific commodities (use generic terms, not proprietary) which are up or
down in price since the last report.
DOWN IN PRICE:___________________
Source:

UP IN PRICE:___________________

7. INVENTORIES OF PURCHASED MATERIALS:
Check the OVERALL inventory level (units, not dollars) including raw, MRO

Institute for Supply Management Indices



111

(Maintenance, Repair, Operating Supplies), intermediates, etc. (not finished
goods, unless purchased) compared to the previous month.
 Higher than a month ago  Same as a month ago  Lower than a month ago

Do you perceive THIS MONTH, your customers’ inventories of products they
order from you, as being:
 Too High
 About Right
 Too Low
8. IMPORTS:
Check the ONE box that best expresses the current month’s OVERALL imports
(units, not dollars) including raw, MRO (Maintenance, Repair, Operating Supplies), components, intermediates, etc. (not finished goods unless purchased)
compared to the previous month.
 Do not import
 Same as a month ago

 Higher than a month ago
 Lower than a month ago

9. EMPLOYMENT:
Check the OVERALL level of employment compared to the previous month.
 Greater than a month ago  Same as a month ago  Less than a month ago

10. SUPPLIER DELIVERIES:
Check the ONE box that best expresses the current month’s OVERALL delivery
performance compared to the previous month.
 Faster than a month ago  Same as a month ago  Slower than a month ago

11. ITEMS IN SHORT SUPPLY:
Report specific commodities (use generic names, not proprietary) you purchase
that are in short supply, even if mentioned in previous reports.
12: BUYING POLICY:
Indicate by checking ONE appropriate box for each category of purchases and
the approximate weighted number of days ahead for which you are committed.
Do not report hedging or speculative purchases.
Production Materials
 Hand to Mouth  30 Days  60 Days  90 Days  6 Months  Year

MRO Supplies
 Hand to Mouth  30 Days  60 Days  90 Days  6 Months  Year
Source:

Capital Expenditures
 Hand to Mouth  30 Days  60 Days  90 Days  6 Months  Year
Reprinted with permission from The Institute for Supply Management™, 2003.

112



The Trader’s Guide to Key Economic Indicators

Each month, the ISM asks its member companies within the
above industries twelve questions (reprinted on the preceding pages
with permission from The Institute for Supply Management™,
2003). The responses are then aggregated by around the 20th or
21st of the month.
In addition to direct responses, which are used in the calculation of the diffusion indices, the ISM asks for remarks after each
question regarding the reasons for higher or lower commodity prices, or greater or less employment. The ISM also asks its
participants to report specific commodities (using generic, not
proprietary, names) they purchase that are in short supply, even if
mentioned in previous reports. These remarks are used in the text
prepared for the monthly report, as well as the summary provided
in the beginning of the report that precedes the presentation of
the individual components.
The ISM separates the responses to each question into positive,
neutral, and negative groups and calculates the percentage of the
whole that each represents. It then plugs the appropriate percentages for each question into the following formula:
Percentage of Positive Responses + [1/2]
(Percentage of Neutral Responses)

The result is a diffusion index. Say the question regarding the
level of employment received 350 responses, of which 20 are negative, or “lower than a month ago”; 275 neutral, or “same as a month
ago”; and 55 positive, or “higher than a month ago.” Plugging those
numbers into the formula would give a value for the ISM Employment Index as follows:

Source:

= [55 ÷ 350] + [1/2] [275 ÷ 350]
= 15.7 + [1/2] [78.57]
= 15.7 + 39.28
= 54.9

Institute for Supply Management Indices



113

Indices calculated in this manner have values between 0 and
100. Values above 50 are interpreted as predicting expansion; those
under 50, contraction. Thus, the expansion/contraction cut-off
level is 50, and not 0. Historically, the individual ISM indices tend
to fluctuate between 35 and 70, depending on the current phase of
the business cycle and the individual index. For example, the Price
Index has fluctuated between 30 and 90, whereas the Export Index
has maintained a range of 40 to 70.
Five of these indices, as was noted earlier, are weighted and summed
to create the PMI. The indices and their associated weightings are:
new orders (30 percent), production (25 percent), employment (20
percent), supplier deliveries (15 percent), and inventories (10 percent).
The September 2002 PMI, for instance, was calculated as follows:
PMI (September 2002) = new orders (0.30 x 49.7)
+ production (0.25 x 55.6)
+ employment (0.20 x 45.8)
+ supplier deliveries (0.15 x 53.4)
and inventories (0.10 x 45.2)
= 14.91 + 13.90 + 9.16 + 8.01 + 4.52
= 50.5

The heaviest weighting is given to new orders and reflects the
fact that it has the greatest predictive value of the five components.
This makes economic sense because orders are, by definition, a representation of intended purchases. This weighting, in turn, contributes to the accuracy of the composite PMI in forecasting turning
points in the business cycle.

Source:

WHAT DOES IT ALL MEAN?
The ISM manufacturing report is valued not only for the diffusion
indices but also for the accompanying discussion and comments
made by the purchasing and supply executives participating in the
survey. Together, the indices and executives’ anecdotal insights
paint a detailed picture of the state of the manufacturing sector.

114



The Trader’s Guide to Key Economic Indicators

The information is also timely, because the report is released on
the first business day of the month, thus unofficially kicking off the
monthly cycle of economic indicators.
All these features make the report a valued tool for Wall Street
economists, government forecasters, and business planners. Because these professionals’ concerns, actions, and pronouncements
all affect the stock and bond markets, investors, too, need to pay
close attention to the report.
When strong postings are registered in the PMI, production,
new orders, and employment indices, it is generally a safe bet that
many of the nation’s manufacturers are experiencing some positive growth. Granted that countless factors and risks influence the
value of a particular company and industry, but when economic
fundamentals are strong (i.e., low inflation, low interest rates, solid
employment growth, and increased global demand for U.S. manufactures), and the ISM’s indices are on the rise, industrial companies
like Caterpillar, Ingersoll-Rand, International Paper, United Technologies, Eaton Corp., and Leggett & Platt tend to prosper.
The following sections discuss the economic significance of the
PMI and a few of the sub-indices contained in the Report on Business, as well as some of the associations that they possess with other
economic indicators. There are countless relationships between
each of the sub-indices and various other economic indicators, far
too many, in fact, for inclusion in this introductory book. It is highly
recommended that readers attempt to discover some of these relationships by charting the data from the ISM’s website against other
economic indicators. Let’s examine a handful of the more popular
relationships studied by Wall Streeters.

A few characteristics that set the PMI apart from most other economic indicators discussed in this book, and that contribute greatly
to its appeal for analysts, economists, and investors, are its relative
simplicity, its strong correlation with macroeconomic trends, and
its unique perspective. The fact that it is more of an anecdotal

Source:

PMI

Institute for Supply Management Indices



115

representation of the goings-on in the economy, from the perspective of those responsible for the actual purchasing of manufactured
goods, than a calculated measure of output or volume of production
is indeed a distinctive quality of this index. Perhaps the most attractive feature of the PMI and its sub-indices, however, is its ease of
interpretation, permitting the trader to interpret month-to-month
changes almost instantly.
The PMI is not solely an indicator of manufacturing activity.
FIGURE 5-1 shows that monthly movements in the index closely mirror the year-over-year percentage change in nominal GDP. This
makes the PMI extremely valuable as a predictor of total macroeconomic activity.
The chart depicts the relationship between the PMI and GDP as
a concurrent or coincidental association. That’s because the PMI in
Figure 5-1 is charted on a quarterly basis, with the quarterly calculation of PMI derived by averaging the individual months’ data for
each respective quarter. But because the ISM’s PMI is released on a
monthly basis, and the GDP report is on a delayed, quarterly basis,
the PMI in fact assumes a leading, or predictive, tenor.
As discussed previously, a level of 50 is considered the cut-off
between expansion and contraction for manufacturing conditions.

Sources: ISM; U.S. Department of Commerce, Bureau of Economic
Analysis

Figure 5-1

GDP Compared Against ISM’s Purchasing Managers Index

GDP YOY%

PMI

20
15

GDP YOY%
PMI

75
70
65
60

55
50

45
40

–5

30
1948 1952 1956 1960 1964 1968 1972 1976 1980 1984 1988 1992 1996 2000

116



The Trader’s Guide to Key Economic Indicators

But when it comes to movements in the broader macroeconomy,
different levels are associated with expansion and contraction. For
example, a reading of about 46.0 in the PMI serves to identify turning points in the overall business cycle.
The reason for the lower level is due to the underlying composition of the economy. From the 1940s through the 1970s, manufacturing activity was a much greater influence on the total U.S.
economy than during the 1980s, when the economy became less
industrialized and more services-oriented. So, in earlier times when
manufacturing fell into a slump, it dragged the entire economy into
recession. Today, manufacturing accounts for only 20 percent or
so of total economic output. As a result, declines in manufacturing
activity don’t always result in macroeconomic recessions. This is
the reason why, in the case of the PMI, lower index levels equate to
macroeconomic recessions.
FIGURE 5-2 shows that post–World War II recessions in the United
States have been associated with steep declines in the PMI, which
sometimes preceded the downturn and always continued through
them. Before the 1980s, virtually every time the PMI fell below
50, the economy slipped into recession. After the 1980s, however,
Figure 5-2

Recessionary Periods Compared Against ISM’s Purchasing
Managers Index

PMI

80
Shaded areas = Recession

70
60

40
30
1948

1953

1958

1963

1968

1973

1978

1983

1988

1993

1998

2003

Sources: ISM; NBER

Institute for Supply Management Indices



117

sub-50 readings were associated with only two, very short and mild,
recessions. If you were to go back and review press clippings about
the sub-50 postings that occurred in 1995–96 and in late 1998,
for example, you’d find many an economist predicting recession.
Those recessions never developed—again, underlining the fact that
manufacturing slumps no longer designate overall recession. That
is, the manufacturing sector can experience a recession without the
broader economy falling into recession.
ISM EMPLOYMENT INDEX
The PMI by itself depicts the general state of the manufacturing
sector and of the larger economy. It does so in broad strokes, however. For a more detailed view, economists and traders look at the
sub-indices, each of which possesses a considerable amount of predictive power with respect to the particular aspect of manufacturing
condition and activity that it portrays.
The ISM Employment Index, which is compiled from the
answers to how the current level of employment compares with
last month’s, shows the employment trends at U.S. manufacturers.

Sources: ISM; U.S. Department of Labor, Bureau of Labor Statistics

Figure 5-3

ISM’s Employment Index Compared Against Manufacturing
Payrolls

ISM

Payroll growth YOY%

-2

-4

Employment Index
Mfg. payroll growth

-6
-8
-10

20
1980

1983

1986

1989

1992

1995

1998

2001

118



The Trader’s Guide to Key Economic Indicators

Source:

For instance, FIGURE 5-3 charts the Employment Index against the
year-over-year percentage change in manufacturing payrolls as reported by the Bureau of Labor Statistics and gives more evidence
of the movement in the U.S. economy away from manufacturing
and toward service industries: From 1980 through 2002, the index
was below the 50 level for a staggering 190 of the 276 months. In
other words, for roughly 69 percent of the last quarter century,
manufacturing employment has been contracting, rather than
expanding.
The relationship between manufacturing payroll growth and
the ISM’s Employment Index is indeed close. Figure 5-3 also
shows that the Employment Index is good at predicting declines
in the growth rate of manufacturing payrolls but not so good at
signaling upturns. During the period covered, whenever the index
fell below 50 and stayed there for at least five months, manufacturing payroll growth turned negative about six months after the
first sub-50 posting. In contrast, since 1980, the Employment
Index rarely, if ever, predicted positive manufacturing payroll
growth. The reason behind this inability probably lies in the fact
that manufacturers shed an overwhelming proportion of workers
during the 1980–2003 period—with many manufacturing positions sent abroad.
Economists have recently begun to conclude that there has
been a structural change in the U.S. economy, particularly in the
employment of manufacturing workers. With soaring productivity
rates beginning in the latter half of the 1990s—economists define
productivity as the amount of output produced per hour worked—
businesses could get away with employing fewer workers. What’s
more, multinational manufacturers had looked overseas to nations
like China, India, the Czech Republic, and Mexico for lower-cost
labor. It has been argued that these job eliminations are permanent
and the jobs are never coming back. This trend is supported by the
extended sub-50 postings in the ISM’s Employment Index from the
mid-1990s to today.

Institute for Supply Management Indices



119

ISM PRICE INDEX
The ISM Price Index, which is compiled from the answers to
the question about current prices paid by manufacturers of commodities used in the production process compared with last month’s
prices, provides an indication of possible inflationary pressures
faced by manufacturers. The index receives a considerable amount
of attention from the Street, especially the credit markets, because
inflation erodes the value of fixed-income securities. And because
the Federal Reserve is so concerned with the possibility of rising
inflation, it is atop the Fed’s economic indicator watch list as well.
Of course, the index directly indicates only how prices are moving in the manufacturing industry. How accurate a reflection it is of
inflation in this sector is shown in FIGURE 5-4, which charts the ISM
Price Index against the year-over-year rate of growth in the intermediate goods producer price index. The intermediate goods index
is one of the wholesale price measures calculated by the Bureau of
Labor Statistics (BLS) as the Producer Price Index (PPI). It is a
representation of price activity of goods that are one stage after raw
materials, in other words, intermediate goods that have received
some processing, and includes products such as flooring, newsprint,
rubber tires and inner tubes, steel wire, and refined sugar. These are

Sources: ISM; U.S. Department of Labor, Bureau of Labor
Statistics

Figure 5-4

ISM’s Price Index Compared Against Intermediate PPI

ISM price

PPI YOY%

20
Price Index

PPI %

–5

1980

1983

1986

1989

1992

1995

1998

2001

–10



The Trader’s Guide to Key Economic Indicators

the types of materials that manufacturing companies are most likely
to buy. Once again, who better to know about price developments
in this area than the purchasing managers of those goods? That’s
why the ISM Price Index is so closely respected and watched by the
financial markets, as well as the Federal Reserve.
One question that reasonably might arise is, if these two indices
are so closely related, why not just watch the intermediate goods
PPI? Because the ISM’s Report on Business is released on the first
business day of the month, it precedes the release of the BLS’s PPI
report by about two weeks. To a bond trader, two weeks is an eternity. Decisions regarding the purchase and sale of fixed-income
securities are made in a matter of seconds. The faster a bond trader
can get the skinny on developing inflation trends the more money
he stands to make.
Investors and economists watch the Price Index to glean information beyond the cost of intermediate goods. The greatest concern
for policymakers is that manufacturers may pass any higher costs on
to the final user of the goods being produced. If any such increased
prices are widespread and sustained, the general price level could
increase, which by definition, signals inflation. In addition to the
corrosive influences in the fixed-income securities market, inflation
impedes the pace of consumer spending because workers have to
earn more (i.e., by working longer) just to afford the same amount
of goods and services they used to get at lower prices.
The Price Index is thus a useful indicator of the potential for inflation, in both the manufacturing sector and the broader economy.
According to a recent Report on Business, a Price Index below 46.9
for a sustained period of time is generally consistent with a decline
in the BLS’s Index of Manufacturing Prices, which is one of the
measures of price inflation on the wholesale level.
As with all sections of the ISM’s Report on Business, the anecdotal commentary and remarks following each sub-index contribute
greatly to this report’s value. Because of the detail and insight that the
monthly report provides, basic materials and commodities analysts—
for example, those covering metals, chemicals, cement, lumber, paper
and packaging—scurry each month to get their hands on a copy.

Source:

120

Institute for Supply Management Indices



121

Investors can benefit by comments such as those in the September 2003 report, which stated that the twelve industries paying higher prices for that month were Tobacco; Wood & Wood
Products; Textiles; Primary Metals; Instruments & Photographic
Equipment; Food; Fabricated Metals; Furniture; Transportation
& Equipment; Chemicals; Industrial & Commercial Equipment &
Computers; and Rubber & Plastic Products.
If investors know which industries are paying higher prices during slower economic periods, they can tell that activity in these
respective industries may be beginning to turn around. One of the
most common characteristics of slower economic times is the inability—and undesirability—of businesses to raise prices. Because
slower economic times mean income growth is sluggish and job
creation scant, businesses rarely get away with raising prices during such periods. (The only time a company can raise prices during
gloomy economic conditions is if that company is selling a necessity,
like health or medical care.) So when prices are rising, it’s generally
a signal of better times to come.
A high correlation also exists between the ISM Price Index and
the spot price of West Texas intermediate crude oil. Again, this
isn’t surprising because manufacturers employ a great deal of oil
in the production process. Smelters, kilns, compressors, furnaces,
and machinery at countless plants and factories utilize some form
of crude oil or one of its derivatives. So when the price of crude oil
rises, so too does the cost of manufacturer’s inputs. These increases
are usually reflected in the ISM’s Price Index, because it is a measure of prices that manufacturers pay. Again, the concern among
policymakers is that if input prices are rising, there is an increased
likelihood that those prices will be transferred to the final product,
which would be a higher cost to the end-user, the consumer.

Source:

ISM SUPPLIER DELIVERIES INDEX
The ISM’s Price Index isn’t the only inflation barometer in the
ISM’s monthly Report on Business. The ISM Supplier Deliveries Index, also referred to as the vendor supplier index, provides

122



The Trader’s Guide to Key Economic Indicators

clues to the future direction of prices.
The Supplier Deliveries Index, which is compiled from the responses to the question regarding delivery performance, is a measure of how long it takes suppliers to deliver parts and materials that
are integral to the production process. When the index exceeds 50,
it means that delivery has slowed, indicating that greater demand is
making it more difficult for suppliers (vendors) to get crucial materials to manufacturers. Sub-50 postings alternatively indicate faster
deliveries.
Lengthening delivery times often result from labor disputes that
lead to strikes or lockouts. Such disputes occur more frequently
among companies involved in the manufacturing process—particularly those in the machinists, rail workers, dockworkers, and trucking industries—than in the services sector. When the delivery time
for supplies (inputs in the production process) lengthens, bottlenecks could be developing that can, if sustained, result in higher
prices because of the inability to fabricate and deliver necessary
inputs.
This process is in agreement with the basic laws of supply and
demand. Impeded deliveries, for whatever reason, limit the amount
of production that can be performed and consequently reduce the
Figure 5-5

ISM’s Supplier Deliveries Index Compared Against JOC-ECRI
Industrial Materials Price Index

ISM

JOCICPI

110

70
65

105

SDI
JOC-PI

100

80
75

35
30

65
1982

1985

1988

1991

1994

1997

2000

2003

Sources: ISM; ECRI

Institute for Supply Management Indices



123

availability of final goods produced. This, by definition, is a decline
in supply. When the supply of a good declines, its price rises. The
longer the labor dispute, the longer the delivery time, the greater
the potential for inflation. Moreover, any time supplier backlogs
develop, the manufacturer incurs greater costs as well as the increased potential for lost business, which may exacerbate the inflationary environment.
The Journal of Commerce’s Industrial Materials Price Index is an
index calculated on a daily basis by the Economic Cycle Research
Institute, representing eighteen industrial commodity materials,
including nickel, cotton, polyester, burlap, copper, red oak, plywood, tallow, steel, crude oil, benzene, and ethylene. As previously
stated, when these industrial commodity prices rise, it is often a sign
that conditions are improving. This index, presented in FIGURE 5-5,
is simply another proxy for raw material or commodity prices for
comparison with the ISM Supplier Deliveries Index. As the chart
illustrates, higher postings in the Supplier Deliveries Index are soon
accompanied by increases in the prices of industrial commodities, as
represented by the Journal of Commerce’s Industrial Materials Price
Index. Conversely, precipitous declines and sub-50 postings in the
Index result in lower industrial price levels.

Source:

ISM NON-MANUFACTURING INDICES
Because manufacturing currently accounts for only 20 percent or so
of total economic output, not surprisingly, economists have wanted
a measure similar to the PMI that would address the condition of
the businesses constituting the other 80 percent of the economy.
In response, the ISM created the Non-Manufacturing Report on
Business in July 1997. The non-manufacturing survey is similar to
the manufacturing survey and, like its older sibling, possesses some
degree of predictive power. It is based on data from responses to
questions asked of more than 370 purchasing and supply managers
in approximately 62 industries, including entertainment, utilities,
hotels, real estate, retail, insurance, finance and banking, accounting, communications, mining, agriculture, engineering, educational

124



The Trader’s Guide to Key Economic Indicators

Parallel to the ISM manufacturing report’s PMI, the non-manufacturing report has a headline index called Total Business Activity.
This index is frequently referred to as the non-manufacturing, or
services, PMI. Unlike the manufacturing report’s PMI, however,
the non-manufacturing Total Business Activity index is not a composite index or a weighted average of sub-indices.
Because the non-manufacturing report is, relatively speaking,
in its infancy, the monthly report—usually released two business
days after the manufacturing report, or the third business day of
the month—doesn’t possess the same ability to move the financial
markets as its older, more esteemed manufacturing counterpart. Its
predictive nature isn’t exactly known, considering that the number
of economic downturns since its initiation in 1997 has been rather
limited.
FIGURE 5-6 shows how closely the ISM’s non-manufacturing PMI
tracks the Commerce Department’s twelve-month rate of consumption of services in the economy. This relationship is to be expected
because services dominate the non-manufacturing sectors surveyed.

Source:

services, construction, and health services—in other words, just
about anything that doesn’t fall under the heading of manufacturing. Because a considerable number of service businesses are
represented in this survey, the business press often refers to the
non-manufacturing survey as the ISM services report.
Among the indices represented in the Non-Manufacturing Report on Business are the following:
 total business activity
 new orders
 backlog of orders
 new export orders
 imports
 prices
 employment
 supplier deliveries
 inventory sentiment
 inventory change

Institute for Supply Management Indices
Figure 5-6



ISM’s Non-Manufacturing PMI and Consumption Spending on Services

ISM

Spending YOY%

65
Sources: ISM; U.S. Department of Commerce

125

55
5

50
45

Non-Mfg PMI
Spending

40
July 1997

July 1999

July 2001

July 2003

Economists have found that an index reading of 50 equates to a
growth rate of about 3.5 percent (year-over-year) in the consumption of services.
The indices in the non-manufacturing survey clearly possess
a high degree of correlation with a considerable portion of U.S.
economic activity. There’s no doubt that the ISM’s non-manufacturing survey will eventually become a top-tier economic indicator,
especially because the United States has steadily evolved from a
smokestack industrialized nation to a more services-dominated one.
But before than can occur, a longer history, as well as a weighted
composite index similar to the Manufacturing ISM Report on Business PMI, will be needed.

HOW TO USE WHAT YOU SEE
The stock market tends to react in sync with movements in the PMI
and Production Index. This is understandable, because manufacturing activity in spite of its reduced role has proven to provide a
good representation of total economic conditions. When businesses
foresee stronger consumer demand—as evidenced through declining inventories and acceleration in consumer spending—they pick

126



The Trader’s Guide to Key Economic Indicators

up the pace of production. So, upticks in the PMI or the Production
Index and postings higher than 50 usually portend better things for
equity issues because greater demand usually results in higher sales
and, subsequently, profitability.
Normally, the fixed-income market would react adversely to
large increases in the PMI. In other words, higher postings, above
50, would be regarded as inflationary, thereby eroding the value of
fixed-income securities, sending prices of bonds lower (and yields
higher). Bear in mind that the ISM’s Report on Business contains as
well the Price Index that also captures the attention of bond traders.
The loftier the posting of the Price Index, the greater the inflation
fear and the greater the sell-off in bonds. So, to better understand
the fixed-income market reactions, Wall Streeters usually look
squarely to the Price Index and its direction as the ultimate inflation
barometer of this report.
THE

TRENCHES

When economists get their hands on a proven set of economic
indicators like those contained in the ISM’s Report on Business,
they tend to get creative and conjure up their own indicators, usually derivations of the more successful components in the series.
One of the more popular tricks performed by Wall Streeters is to
take the difference between the ISM’s New Orders Index and the
Inventories Index. The resulting statistic, depicted in FIGURE 5-7, has
exhibited a relatively strong correlation with the year-over-year
percentage change in real GDP.
Because gross domestic product is reported on a quarterly basis,
and is prone to rather lengthy delays, some economists find this
measure a timely (available monthly as opposed to quarterly) and
accurate depiction of the economic growth rate.
Like the many diffusion indices examined in this chapter, the
New Orders minus Inventories (NO-INV) Index doesn’t project a
specific magnitude for real economic growth, or GDP, but it is quite
telling about the likelihood of economic contraction or expansion.
During some phases of the business cycle when the identification

Source:

TRICKS FROM

Institute for Supply Management Indices

Source:

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; ISM

Figure 5-7



127

ISM’S New Orders Minus Inventories Versus Real GDP

NO-INV

GDP YOY%

25
20

15
10

5
0

–5
NO-INV

–10

GDP

–15
1980

–3
1983

1986

1989

1992

1995

1998

2001

of recession isn’t exactly clear, this trick becomes another helpful
indicator on the state of economic affairs.
Economists look out for low or negative readings in the NOINV Index. Anytime there is no difference between the Inventories
and New Orders Indices, that is, when the NO-INV Index is zero,
the economic growth rate tends to contract. Historically, negative
readings have been associated with economic recessions. Readings
in the NO-INV of 10 or higher historically have equated to real
economic growth rates in excess of 3.5 percent—a rate generally
considered to be strong.
Are these levels gospel? Absolutely not. This is nothing more
than a crude model and should only be used as a guide to projecting
economic activity on a monthly basis (because GDP is not readily
available). However, as this indicator has, on occasion, possessed
somewhat of a leading quality, some economists look for trends in
the index to forecast GDP growth.
Another crude model applied by economists is formed by taking
the difference between the ISM’s New Orders Index and the Price
Index. The resultant measure has displayed a relationship with the
year-over-year percentage change in the total return of the S&P
500. The possible economic explanation behind this association

128



Figure 5-8

The Trader’s Guide to Key Economic Indicators
ISM’s New Orders Minus Price Index Versus S&P 500 Total Return

S&P 500 YOY%

ISM NO-P

30
20

10
40

–10
–20

–30
–20

S&P 500
NO-P

–40

–50
1971

1975

1979

1983

1987

1991

1995

1999

2003

may be that the Price Index is in effect a proxy for costs, and New
Orders for revenues. The difference between these two measures
would be profits. In this case, profits are reflected in the year-overyear total return of the S&P 500 Index. As suggested in FIGURE 5-8,
there is indeed a close association between these indicators.
Problems with this little-known index include that input prices
as measured by the ISM Price Index are not a highly accurate proxy
for costs. A number of costs manufacturers incur are not captured
by the Price Index, the largest being labor, for example. Furthermore, new orders are hardly the best representation of revenues.
Orders aren’t always filled, many are delayed, and ultimately different volumes may be shipped. Still, the so-called NO-P makes for a
“quick and dirty” model for predicting the total return of the S&P
500. A simple regression analysis reveals that the total return on the
S&P 500 (on a year-over-year basis) will be positive as long as the
difference between the ISM New Orders Index and the Price Index
is greater than –20.
Bear in mind that this index doesn’t always pan out as the greatest forecasting model of stock market levels. Despite the aftermath
of the burst stock market bubble in early 2000, the NO-P began
to register some hefty readings in early 2001, suggesting the total

Source:

Sources: Standard & Poor’s; ISM

Institute for Supply Management Indices



129

Source:

return of the S&P 500 would be strong. Clearly this was a false signal, as stocks continued their descent well into 2003. Nevertheless,
as Figure 5-8 suggests, for a simple, back-of-the-envelope type of
forecast, the tool isn’t too bad a predictor of stock market returns.
Regional economists and all those who are looking for a more
detailed, targeted representation of manufacturing conditions are
encouraged to visit the ISM’s website and view the various regional
business survey reports. These surveys are conducted by the many
local purchasing management associations, and should not be confused with the national survey that has been examined in this chapter. Some of the more popular regional reports observed by Wall
Street include the Arizona, Austin, Buffalo, Chicago, Cleveland,
Dallas, Denver, Georgia, Houston, New York, Northwest Ohio,
Pittsburgh, and Western Washington reports.
These reports have helped many Wall Street economic departments create a mock Beige Book. The real Beige Book, so called by
the Street because of the color of its cover, is one of the three books
that the Federal Reserve creates and uses during its eight-timesa-year monetary policy deliberations. It is the only information at
those meetings released to the public; this generally occurs two
weeks prior to the Fed’s Open Market Committee meeting.
Formally known as the Summary of Commentary on Current
Economic Conditions by Federal Reserve District, the Beige Book
contains anecdotal commentary that has been accumulated by each
of the twelve Federal Reserve districts. The regional ISM surveys
are frequent providers of information to those surveys and generally offer excellent insight as to the manufacturing and economic
climate in each region.

This page intentionally left blank

Manufacturers’
Shipments, Inventories,
and Orders

he Manufacturers’ Shipments, Inventories, and Orders,
or M3, survey is one of the most respected economic indicators on Wall Street. Published monthly by the U.S. Department of
Commerce’s Census Bureau, the report measures current activity
and future commitments in the U.S. manufacturing sector. Using
data supplied by some 4,700 reporting units of businesses in eightynine industry categories, it provides statistics on the value of factories’ shipments, new orders, unfilled orders, and inventories. The
survey is closely followed by economists, members of the business
community, and various government organizations, including the
Bureau of Economic Analysis (BEA), which employs the survey’s
figures in preparing its gross domestic product, or GDP, report,
particularly the investment and inventory sections.
The M3 survey is published in two parts. The Advance Report
on Durable Goods is released about four weeks after the reference
month, on about the eighteenth business day of the month (the
date varies somewhat to avoid overlapping with other economic
releases). The revised and more comprehensive Manufacturers’
Shipments, Inventories, and Orders appears about a week later and
provides greater detail about production, by industry group, as well
as including for the first time information about nondurable, in addition to durable, goods. In both the advance and the later report,
it is the orders component that garners the most attention from
market participants.
Manufacturing orders constitute a leading economic indicator,
because they reflect decisions about optimal inventory levels given
131

132



The Trader’s Guide to Key Economic Indicators

the demand businesses anticipate based on their economic forecasts.
In this regard, new orders for durable goods have proved to be particularly accurate predictors, because demand for such products is
especially dependent on economic health.
Manufacturing is an important sector of the U.S. economy, accounting for roughly 20 percent of GDP and about the same percentage of overall employment. But the significance of the demand
and pace of production figures for trading floor economists and
traders is even greater than those percentages imply. Trends in production are usually experienced as well by the services sectors associated with shipments of manufactures and can be quite accurate
in marking turning points in the overall economy. For that reason,
the Conference Board’s index of leading indicators (LEI) includes
components of the M3 survey: manufacturer’s new orders for nondefense capital goods and manufacturers’ new orders for consumer
goods and materials.
The M3 survey can be found on the Census Bureau’s website,
at www.census.gov/indicator/www/M3/index.html, together with
the historical series for all its components. Larger increases—of the
magnitude of 1.0 percent or more—in the monthly percent change
of new orders for durable or factory-produced goods are usually
interpreted as positive for equity markets and somewhat undesirable for fixed income security holders. Businesses only order if they
expect demand to increase. Conversely, declines in new orders are
perceived as slower times to come, which generally cause the stock
market to decline and bond prices to rally. Investors must keep in
mind that both reports are extremely volatile and hard to predict,
so the markets may not always place the utmost emphasis on the
month-to-month postings.

The M3 survey grew out of the Department of Commerce’s Current Industrial Report (CIR) program, in place since 1904. In
1939, the Commerce Department’s Office of Business Economics,
working through the CIR program, established the first monthly

Source:

EVOLUTION OF AN INDICATOR

Manufacturers’ Shipments, Inventories, and Orders



133

Source:

Industry Survey. The forerunner of the M3 survey, it contained
broad-based measures of inventory changes and information about
the ratio of new and unfilled orders to current sales. In following
years, various changes were made in how the data were calculated
and presented, as well as in which industry groups were represented
and their composition. Revisions instituted from 1947 to 1963, for
example, included adding seasonally adjusted dollar estimates of the
data to better reveal non-seasonal features, distinguishing durable
from nondurable and household from business-related goods, and
breaking down market categories into final products and materials. Seasonal effects include a decline in motor vehicle production
during summer months as factories retool for the new model year,
and the increase in the production of heating oil during September
ahead of the winter special season.
In 1997, as part of a broader revision involving many Commerce
Department economic reports, the Census Bureau responded to
the development of new products and industries by switching
the M3 series of data to the more current and generally accepted
North American Industry Classification System (NAICS) from the
outdated Standard Industrial Classification (SIC) system. These
systems are simply uniform systems of classification. By adopting
the NAICS, manufacturing, trade, and inventory data can be compared throughout all of North America, rather than just the United
States. The NAICS system resulted in a number of regroupings.
For example, some activities that were not previously classified
as manufacturing under the SIC system—such as bottled spring
water, retail bakeries, and software reproduction—are now formally
counted under the new NAICS system. During this benchmark
reclassification, data were revised only back to February 1992, when
the U.S. economy was in the process of emerging from the 1990–91
recession. Because the NAICS data are only available from that date
and previous data are classified by different, SIC industries, historical analysis is limited.

134



The Trader’s Guide to Key Economic Indicators

DIGGING FOR THE DATA
The Census Bureau obtains its data on domestic manufacturing
through surveys of manufacturing companies with annual shipments totaling $500 million or more. Participation is voluntary,
and responses may be submitted over the Internet, by telephone,
or by fax. Relevant data received by the eighteenth day of the
month following the month covered by the survey are included in
the Advance Report on Durable Goods. Additional data, collected
through the thirtieth of the month, are consolidated with the previously reported data and included in the more complete Manufacturers’ Shipments, Inventories, and Orders.
The reports contain both seasonally adjusted and nonadjusted
figures for the record month and for the previous three months,
together with percentage changes from month to month. All the
values are nominal, given in constant-dollar terms.
The Department of Commerce’s Census Bureau presents the
collected data both by industry category, such as industrial machinery and computers, and by topical series. Topical series aggregate
industries into broad market groupings, such as home goods and
apparel, and into special series, such as nondefense capital goods
and defense capital goods. For example, the nondefense capital
goods series includes small arms; farm machinery and construction
equipment; turbines, generators, pumps, and compressors; oil- and
gas-field machinery; computer storage devices; office and institutional furniture; and medical equipment and supplies. Wall Street
industry analysts monitor the topical series carefully. Aerospace and
defense analysts, for instance, watch defense capital goods closely,
whereas hardware and peripheral equipment analysts scrutinize the
information technology series.

Durable goods are goods expected to last three years or more.
They include lumber and wood products; furniture and fixtures;
stone, clay, and glass products; and industrial machinery and equip-

Source:

DURABLE GOODS REPORT

Source:

Manufacturers’ Shipments, Inventories, and Orders



135

ment. These products tend to be quite pricey and are not usually
purchased on a regular basis. As a result, data connected to their
manufacture fluctuate significantly from month to month and are
difficult to predict. This is particularly true with regard to defenserelated goods, such as ships and aircraft, whose valuation is exceptionally complex.
The Advance Report on Durable Goods contains four categories of data: shipments, new orders, unfilled orders, and inventories.
The table in FIGURE 6-1, from the May 2003 report, illustrates how
the seasonally adjusted data (and the monthly percentage change)
for shipments and new orders of durable goods in different sectors
are reported. Another table in the report (not shown) presents the
same breakdown for unfilled orders and total inventories.
Shipments comprise products actually sold by establishments.
The dollar figures reported are the net sales values of domestically manufactured goods shipped to distributors during the record
month. (For larger goods with lengthy fabrication schedules, such
as aircraft and tanks, the reported figures are estimates of the value
of work performed during the survey period.)
Some of the larger categories of shipments include capital goods
(products, such as machinery, that are used to make other products),
which amount to 37.3 percent of the total; of transportation products, representing 29 percent; and of machinery, 13 percent. (The
percentages don’t add up to 100 percent because the categories
overlap, i.e., both nondefense capital goods and machinery include
textile and paper industry machinery.)
New orders are product orders received during the record
month, including both those to be filled during the month and
those for goods to be delivered some time in the future. Because
these figures indicate businesses’ intentions with regard to purchases, they are the most forward-looking in the release. Unfilled
orders are orders that haven’t yet been shipped or reported as sold.
They are a measure of order backlog.
Businesses understandably want to keep information regarding
the level of new orders close to their vests. Because some survey
participants (in this voluntary survey) are reluctant to provide this

136



Figure 6-1

The Trader’s Guide to Key Economic Indicators
Manufacturers’ Shipments and New Orders for Durable Goods
Seasonally Adjusted
Monthly
Mar
2003*

Percent Change
Mar–
Apr

Feb–
Mar *

Shipments ........................... $170,294 $171,564 $170,675

–0.7

0.5

–1.9

New orders ......................... 168,929

–2.4

1.4

–1.1

Item

Apr
2003

Feb
2003

Jan–
Feb

DURABLE GOODS
Total
173,159

170,833

Shipments. .......................... 120,546

120,393

118,694

0.1

1.4

–3.0

New orders ......................... 119,331

120,745

119,485

–1.2

1.1

–2.3

Shipments ........................... 161,426

163,060

161,964

–1.0

0.7

–2.1

New orders ......................... 158,903

161,365

160,283

–1.5

0.7

–2.9

Shipments ........................... 118,998

119,382

118,902

–0.3

0.4

–1.6

New orders ......................... 117,633

120,977

119,060

–2.8

1.6

–0.4

Excluding transportation

Excluding defense

Manufacturing with
unfilled orders

Primary metals
Shipments ...........................

10,719

10,957

10,985

–2.2

–0.3

–3.2

New orders .........................

10,553

10,363

10,871

1.8

–4.7

–2.8

Shipments ...........................

20,408

20,641

20,606

–1.1

0.2

–1.3

New orders .........................

20,236

20,496

20,247

–1.3

1.2

–1.7

Shipments ...........................

22,296

22,320

21,905

–0.1

1.9

–2.2

New orders .........................

21,849

22,686

21,836

–3.7

3.9

–2.5

Shipments ...........................

27,023

26,075

25,492

3.6

2.3

–4.7

New orders .........................

26,829

26,824

26,177

0.0

2.5

–4.2

Shipments ...........................

7,489

6,233

6,608

20.2

–5.7

–14.0

New orders .........................

7,415

6,431

6,451

15.3

–0.3

–13.1

Shipments ...........................

5,473

5,595

5,496

–2.2

1.8

–1.6

New orders .........................

6,039

6,365

6,371

–5.1

-0.1

0.8

Fabricated metal products

Computers and electronic
products

Computers and
related products

Communications
equipment

Source: U.S. Department of Commerce, Bureau of Census

Machinery

Manufacturers’ Shipments, Inventories, and Orders
Figure 6-1



137

(continued)
Seasonally Adjusted
Monthly
Apr
2003

Item

Mar
2003*

Percent Change
Feb
2003

Mar–
Apr

Feb–
Mar *

Jan–
Feb

Computers and electronic
products (continued)
Semiconductors
Shipments ...........................

(NA)

New orders .........................

(NA)

Shipments ...........................

$ 8,609

$ 8,644

$ 8,568

–0.4

0.9

–4.3

New orders .........................

8,553

8,871

8,574

–3.6

3.5

–1.7

Shipments ...........................

49,748

51,171

51,981

–2.8

–1.6

0.8

New orders .........................

49,598

52,414

51,348

–5.4

2.1

1.8

Shipments ...........................

35,538

36,655

37,013

–3.0

–1.0

–2.6

New orders .........................

35,412

36,515

37,117

–3.0

–1.6

–2.2

Shipments ...........................

5,306

5,980

6,071

–11.3

–1.5

28.0

New orders .........................

4,427

2,980

3,588

48.6

–16.9

–30.5

Shipments ...........................

3,362

3,299

3,249

1.9

1.5

0.4

New orders .........................

4,285

5,822

2,449

–26.4

137.7

3.4

Shipments ...........................

31,491

31,756

31,138

–0.8

2.0

–2.7

New orders .........................

31,311

31,505

31,780

–0.6

–0.9

Shipments ...........................

63,567

62,997

62,328

0.9

1.1

–1.2

New orders .........................

64,404

66,665

63,431

–3.4

5.1

–0.4

Shipments ...........................

56,121

55,817

55,035

0.5

1.4

–1.4

New orders .........................

55,644

55,801

54,098

–0.3

3.1

–4.9

Shipments ...........................

54,042

53,155

52,193

1.7

1.8

–3.5

New orders .........................

53,874

55,516

53,006

–3.0

4.7

–2.4

Electrical equipment,
appliances, and components

Transportation equipment

Motor vehicles and parts

Nondefense aircraft and
parts

Defense aircraft and parts

All other durable goods

Capital goods

Nondefense capital goods

Excluding aircraft

Source:

Defense capital goods
Shipments ...........................

7,446

7,180

7,293

3.7

–1.5

0.5

New orders .........................

8,760

10,864

9,333

–19.4

16.4

37.4

NA Not Available

* Revised

138



The Trader’s Guide to Key Economic Indicators

information, Commerce Department’s Bureau of the Census is
forced to estimate the level of new orders.
New orders numbers are derived by adding together the dollar values of shipments and of unfilled orders from the current
month and subtracting the value of unfilled orders from the previous month (which can’t be “new”). For example, in April 2003, the
value of durable goods shipments was $169.791 billion and that of
unfilled orders was $477.388 billion, for a total of $647.179 billion.
Subtracting March’s $478.035 billion of unfilled orders yielded a
new orders number for April of $169.144 billion. (For nondurable
goods, new orders and shipments are generally the same. Because
many of these products are perishable, they are usually shipped, and
in many cases, consumed, as soon as they are ordered.)
The monthly inventory postings in the M3 report indicate the
dollar value of goods stockpiled at factories and in sales branches,
regardless of the stage of fabrication. These manufacturing inventories account for about 40 percent of the total U.S. inventories.
(The total business inventory position is discussed in detail in
Chapter 7.)

The more comprehensive factory orders report contains revisions
to the Advance Report on Durable Goods, as well as the first presentation of nondurable goods data and greater detail on all of these
data by stage of fabrication.
What’s a nondurable good? An old economist friend once gave
the following nontechnical definition: If you leave it outdoors in the
elements and after three years it disappears, it’s a good bet that it was
a nondurable good. Included in this category are foods and beverages, tobacco products, textiles, apparel, paper and allied products,
chemicals and allied products, and petroleum and coal products.
Nondurables don’t enjoy the headline status of their cousins,
durables. Durable goods such as turbines, engines, household appliances, and machinery are sexier and easier to visualize as factorproduced manufactures than are nondurables such as pesticides,

Source:

FACTORY ORDERS REPORT

Manufacturers’ Shipments, Inventories, and Orders



139

tobacco, and paints. That doesn’t mean they aren’t an integral part
of the economy, however. Nondurable goods account for just a little
less than half of total manufacturing industry shipments. In April
2003, for example, nondurables made up 46.95 percent of total
shipments, compared with 52.95 percent for durables.
In addition to industry category and topical series, the factory
orders report also groups data by stage of fabrication: materials and
supplies, work in process, or finished goods. Most economists like
to look at the work in progress stage, because it includes material
already in the pipeline. About three-quarters of the products at this
stage of fabrication are durable goods. The fabrication of shortlived nondurables tends to be similarly short; these goods are heavily represented in the finished-goods numbers.

WHAT DOES IT ALL MEAN?
The M3 report is a gold mine of economic information for investors, traders, analysts, and economists. The durable new orders data
represent a particularly rich lode, because of the insight they provide into a large component of personal-consumption and capital
expenditures.

Source:

DURABLE GOODS REPORT
The equity market reacts positively to increases in new orders for
durables, because they indicate new demand and an optimistic economic outlook. Conversely, a slowdown or decline in new orders
implies a softer economic climate and less likelihood of a pickup in
corporate profitability, which generally hurts the value of stock. In
the land of fixed income, the situation is reversed: Strong monthly
postings fuel worries about inflation, which erodes the value of coupon payments and sends interest rates higher, both of which depress
bond prices and push yields up. The bond market welcomes weaker
data, because they are generally associated with a benign inflationary environment.
Interpretation of these data is complicated, however, by their ex-

The Trader’s Guide to Key Economic Indicators

Figure 6-2

Month-to-Month Percentage Change in New Orders for Durable Goods

Percent

15
10
5
0
–5
–10
–15
1992

1994

1996

1998

2000

2002

treme volatility. Postings in the durable goods and factory orders reports whipsaw from month to month. This is illustrated by the chart
in FIGURE 6-2, which shows the monthly changes in new orders for durable goods. Virtually all the new orders data, whether for fabricated
metals, communications equipment, or machinery, exhibit this type
of irregular behavior, which inhibits accurate trend spotting.
To correct for this instability, economists adopt two approaches.
One is to look at the data without the most volatile components:
the defense and transportation sectors. Demand for goods such
as ships, military armored vehicles, and guided missiles is highly
unpredictable and unstable, especially on a monthly basis; an unforeseen order for twenty-five commercial airplanes can really
blow an economic forecast. The other approach is to apply some
smoothing techniques, such as moving averages, to the data. To
compute a moving average, you start by averaging the data over
a certain period—for a three-month moving average, for example,
you would add together the figures (new orders, say) from three
monthly reports and divide by three. To make this average move,
you recalculate it each month, dropping the oldest data point and
adding the newest one. Another smoothing technique is to look at
year-over-year (rather than month-to-month) changes.

Source: U.S. Department of Commerce, Bureau of the Census



Source:

140

Manufacturers’ Shipments, Inventories, and Orders
Figure 6-3



141

New Orders for Durable Goods

Source:

Sources: U.S. Department of Commerce, Bureau of the
Census; NBER

YOY%

20
Shaded areas = Recession

15
10
5
0
–5
–10
–15

Durable goods less transportation
Durable goods

–20
1993

1995

1997

1999

2001

2003

The chart in FIGURE 6-3 illustrates the smoothing effects of two
approaches: eliminating the volatile transportation sector and taking the longer, year-over-year view. Although the data still reveal
plenty of volatility, the sharp swings are minimized.
The chart illustrates another point: the importance of using
additional sources of information, such as business reports in
newspapers and on TV, in interpreting the data. In Figure 6-3, the
line graphing new orders bottoms out in October, one of the final
months of the 2001 recession, and heads higher during the year’s
final quarter. Anyone following the economic news would know
that this recovery was prompted by the zero percent financing that
automakers introduced for several makes and models in October
2001. This move sparked the highest automotive demand, production, and sales in history: In October alone, some 21.1 million cars
and lightweight trucks were sold. In the period of sluggish economic growth that followed the 2001 recession, big-name retailers
such as Home Depot and Sears also offered zero percent financing
for up to eighteen months on large durable goods such as washing
machines, refrigerators, carpeting, and lawn mowers. These actions
served not only to spur activity in the sectors directly involved but
also to boost aggregate demand in 2001 and so help keep the reces-

142



The Trader’s Guide to Key Economic Indicators

sion mild and short-lived. Economists and analysts who took the
extra step of investigating the automakers’ and retailers’ business
decisions were able to forecast economic activity more accurately
and perhaps avoid being overoptimistic about any continuing impetus from the auto sector, since its strength is somewhat artificial,
built on unsustainable zero financing and discounts that in effect
cannibalized its future sales.
Of the categories represented in the durable goods report,
the primary metals sector—which includes iron, steel, aluminum,
and copper production—is among those most closely watched by
economists. Although shipments of these products represent a
relatively small slice of the total for durable goods (about 6 percent in April 2003, as shown in Figure 6-1), the materials are vital
ingredients in the manufacturing process. Consequently, increases
in new orders for the metals generally hint at greater future manufacturing activity.
Most of the other categories in the durable goods report involve
manufacturing. The goods produced by these sectors are generally
expensive, and buyers of them (whether individuals or companies)
usually need to finance their purchases. New orders for and shipments of these products are thus very sensitive to interest rates.
High rates mean lower numbers. Lower rates, in contrast, boost the
figures, by making it easier for consumers to buy furniture, appliances, automobiles, and personal computers and for businesses to
buy machinery, power transmission generators, and capital equipment. When the Federal Reserve reduces its overnight borrowing
target, it is attempting to stimulate these sectors.

By the time the report on factory orders comes out, the markets
have had time to absorb and react to the critical data in the durable
goods report, released a week earlier. The more detailed information on inventories and unfilled orders contained in the later report
doesn’t exactly light a fire under traders. Therefore the markets
don’t move much when it is released. The report contains several

Source:

FACTORY ORDERS REPORT

Manufacturers’ Shipments, Inventories, and Orders



143

components, however, that provide important measures of current
and future demand and so bear careful monitoring.
Unfilled orders are a reasonably accurate gauge of the strength
of demand for the particular industry, and so of the strength of the
underlying economy. Economists view a rising backlog as a sign of
accelerating demand and economic strength and a declining backlog as an indication of a weakening economy. The story behind
this relationship is somewhat complicated. When the economy is
expanding rapidly, businesses tend to be optimistic about the future
and place large numbers of new orders for goods. They want the
products on their shelves and in their showrooms when consumers
come calling. The manufacturers of these goods attempt to fill as
many of the orders as possible but often can’t do so expeditiously
with their existing staff levels. Most employers put off hiring new
workers until they see demand increasing. Companies adopt this
“just-in-time” approach to hiring because of the expenses related to
employment and employees—perhaps their greatest business costs.
Manufacturing is largely unionized, and specialized craftsmen with
finely honed skills are often needed to fill skyrocketing new orders.
Manufacturers must consult union officials, adopting specific pay
rates and working schedules. They are often forced to hire a minimum number of workers. If the economy turns south, triggering
widespread order cancellations, the companies are saddled with
newly hired workers standing around idled machinery. Only when
the union-negotiated contracts expire are they free to eliminate
staff. No wonder they’re reluctant to add employees until demand
makes it necessary. But the process of advertising for, interviewing,
hiring, and training new workers is lengthy. In the meantime, backlogs of unfilled orders thus pile up.

Source:

HOW TO USE WHAT YOU SEE
The basic trading strategy related to the M3 report involves identifying potential turning points in demand for U.S. manufactures and
related products. Obviously, the greater the volume of shipments,
the greater the strength of current demand. For future demand, the

The Trader’s Guide to Key Economic Indicators



Figure 6-4

Shipments of Nondefense Capital Goods Excluding Aircraft
Versus Software and Equipment Investment

NDCGXA ($ in millions)

Software & Equipment ($ in billions)

70,000

1,000

65,000

900

60,000

800

55,000

700

50,000

600

45,000
NDCGXA

40,000

500

Software & equip.

400

35,000
1992

1994

1996

1998

2000

2002

Sources: U.S. Department of Commerce, Bureau of the Census
and Bureau of Economic Analysis

144

reasoning is similar: The higher the number of new orders for goods,
the greater the probability of strong activity down the road; the lower
the number, the bleaker the outlook. Particularly potent in predicting the future are manufacturers’ new orders of consumer goods
and materials and of nondefense capital goods. These components
have proved so accurate that the Conference Board includes them
in its index of leading economic indicators. Remember, however,
that in applying this approach, it is best to look at year-over-year
percentage changes as opposed to the monthly figures, which, as
noted above, are too volatile to allow accurate trend spotting.
THE

TRENCHES

This chapter’s trick is to watch the value of nondefense capital goods
spending excluding aircraft (NDCGXA). The BEA uses the reports’
figures for shipments of NDCGXA in determining the equipment
and software investment component of the GDP report. As FIGURE 6-4
shows, the volume of NDCGXA shipments shows a high correlation
with the level of software and equipment spending.
Because of this relationship, shipments of nondefense capi-

Source:

TRICK FROM

Manufacturers’ Shipments, Inventories, and Orders

Source:

Sources: U.S. Department of Commerce, Bureau of the Census
and Bureau of Economic Analysis

Figure 6-5

145



New Orders for Nondefense Capital Goods Excluding Aircraft
Versus Capital Spending

Investment %

Orders %

6
4

–2

–5

–4

Spending

–10

–6

Orders

–8

–15
1993

1996

1999

2002

tal goods excluding aircraft can serve as a proxy for the level of
total nonresidential business investment in the GDP report, also
referred to as capital goods investment. This is very useful; because
investment in capital drives economic activity, economists want
to spot trends in this component as early as possible. But capitalinvestment figures are available only on a quarterly basis, in the
GDP report, and three months is a very long time to wait. NDCGXA shipments are among the few monthly surrogates for this
important indicator—and the only one that is given in dollardenominated values, rather than as a representative index, like the
industrial-production or the Institute for Supply Management New
Orders diffusion indices.
As mentioned above, shipments data paint a picture of current
demand; for future trends you have to look at new orders. New orders for nondefense capital goods excluding aircraft are particularly
significant. Only when businesses are confident about the economic
outlook and future demand will they make costly investments in
new machinery and innovative processes. As FIGURE 6-5 illustrates,
NDCGXA new orders anticipate this investment by about three
to six months. On a macroeconomic level, prior to the commence-

146



The Trader’s Guide to Key Economic Indicators

Source:

ment of the recession in April 2001, for instance, new orders began
a sharp decline in July of 2000 replicated by capital spending during
the fourth quarter of 2000.
So watch for trends in the NDCGXA data. These reveal corporate perceptions of the near-term economic outlook and so can
help you to predict manufacturing activity and overall economic
growth.

Manufacturing and Trade
Inventories and Sales

alk about the red-headed stepchild of economic indicators! Few investors, either professional or lay, pay much attention to inventories. Even popular economics texts largely ignore
them. Not so corporate managers and economists. Both these
groups are well aware of the importance of inventory levels—and of
the dangers of miscalculating them.
Business inventories are “waiting room” goods—products that
have been manufactured, processed, or mined but have not yet been
sold to a final user. As such, they are very important in the calculation of the Gross Domestic Product report (see Chapter 1). GDP
is the total amount of final goods and services produced in an economy in a given period. That includes goods that haven’t found a
final purchaser—in other words, inventory. Accordingly, the Bureau
of Economic Analysis incorporates inventories in the aggregate
expenditure formula—C + I + G + (X–M )—used to calculate GDP.
The I in the formula stands for gross private domestic investment,
which includes businesses’ spending on inventories. That’s one reason economists keep close track of this factor. Miscalculating the
value of inventories can throw off an estimate of economic growth
by up to two percentage points—a mistake that can ruin a Wall
Street economist’s career.
But inventories’ role in the GDP calculation is not the sole
reason economists and managements monitor them carefully.
Failure to balance inventories against demand can, and has, hurt
businesses and destabilized the economy. Companies that overstock their shelves in anticipation of orders that fail to materialize
147



The Trader’s Guide to Key Economic Indicators

find themselves in a hole, forced to cut production and lay off
workers. Some prominent economists have even implied that the
Great Crash of 1929 was provoked, at least in part, by the misalignment of inventory positions. Businesses whose inventories
are too lean, on the other hand, miss potential profit during a
boom. To avoid such opportunity loss, companies with low stocks
may pick up the pace of new orders, spurring manufacture and
boosting the economy. Because of their relation to production
activity, inventory levels are also of interest to traders. Falling levels, with their promise of increased production and thus
higher earnings, can boost equity prices. Fixed-income prices,
meanwhile, may decline because of fear of heightened inflation
that often accompanies an accelerating economy. On the other
hand, rising inventories, implying slowing production, depresses
stocks and buoys bonds.
Several economic reports address the inventory position of U.S.
companies, but the most comprehensive is the monthly Manufacturing and Trade Inventories and Sales report. The MTIS report,
also referred to on the Street as the total business inventories
report, contains the most recent available inventory data for the
manufacturing, wholesale, and retail sectors. It is assembled by the
U.S. Department of Commerce’s Census Bureau and released at
8:30 a.m. ET, about six weeks after the reference month. The report for June 2003, for example, was distributed on August 13. Both
the current report and historical data are available on the Census
Bureau’s website at www.census.gov/mtis/www/current.html.
Despite the importance of inventories to GDP and business
management, the monthly MTIS report is generally ignored by
the investment community and the business media. One reason is
that all the sales statistics it contains and two-thirds of the inventory
data—those for manufacturing and wholesalers—have already been
reported elsewhere. The only new information is on retail inventories, and all the figures are a month and a half old.
Another reason for the MTIS report’s lukewarm reception
is the difficulty of making the historical comparisons necessary
for extensive analyses. The numbers presented are only for the

Source:

148

Manufacturing and Trade Inventories and Sales



149

record month and the preceding one. Moreover, because all the
data were converted to the North American Industry Classification System (NAICS) in 2001, the statistical history extends only
to 1992.
These drawbacks notwithstanding, the MTIS report is definitely worth a careful read. Beyond presenting the first monthly
numbers for retail inventories, it contains the very useful inventories-to-sales ratios. And its breakdown and presentation of the
data make it relatively easy to trace trends and evaluate the significance of movements.

EVOLUTION OF AN INDICATOR
The MTIS survey was originally implemented and the results published and analyzed by the Department of Commerce’s Bureau of
Economic Analysis. The Census Bureau took on those duties in
March 1979, with the release of the January inventory and sales data
of that year. In 1997 the bureau, responding to the scores of new
products and industries that needed to be catalogued, decided to
convert to the NAICS from the outdated Standard Industrial Classification (SIC) system. It wasn’t until 2001 that the data contained
in the MTIS report was converted to NAICS. This necessitated
some adjustment of data reported under the old system, so that
historical comparisons could be made. But the adjustment has been
completed back to only 1992.

Source:

DIGGING FOR THE DATA
The MTIS report compiles sales data previously reported in the
Census Bureau’s Advance Monthly Sales for Retail Trade and Food
Services report (see Chapter 10) together with inventory and sales
information from its Wholesale Trade Survey and its Manufacturers’ Shipments, Inventories, and Orders survey (see Chapter 6).
The only new information, as noted earlier, is on retail inventories,
which is obtained from retail firms regarding the value of their endof-month inventories.

150



Figure 7-1

The Trader’s Guide to Key Economic Indicators
Estimated Monthly Sales and Inventories for Manufacturers,
Retailers, and Merchant Wholesalers
June 2003
( p)

Total business
Manufacturers
Retailers
Merchant wholesalers

$845,230
324,611
283,096
237,523

Sales ($ in millions)
May 2003
(r)

June 2002

$835,780
321,153
280,578
234,049

$819,478
320,810
270,093
228,575

(p) Preliminary, (r) Revised

The report organizes the data into three tables. Table 1 contains sales and inventory numbers and inventories-to-sales ratios
for business as a whole as well as for the three primary subgroups:
manufacturers, retailers, and merchant wholesalers. The numbers
are given, in both seasonally adjusted and nonadjusted form, for
the reference month and the month preceding, as well as for the
same month a year earlier. FIGURE 7-1 shows part of Table 1 from the
June 2003 report, released in August. In this table, manufacturing
accounts for 38 percent of total inventories, retailers for about
37 percent, and merchant wholesalers about 25 percent. These
percentages tend to remain relatively stable, although they can be
expected to change somewhat as the composition of the American
economy evolves.
Table 2 from the report shows the month-over-month and yearover-year percent changes in sales and inventories (both seasonally
adjusted and not) for the reference month and the month-overmonth changes for the month preceding. Table 3 presents the detail
for the retail sector, breaking down the inventory and sales numbers, percent changes, and ratios by retail business: motor vehicle
and parts dealers; furniture, home furnishing, electronics, and appliance stores; building materials, garden equipment, and supplies
stores; food and beverage stores; general merchandise stores; and,
within the last category, department stores excluding leased stores.
It also gives two totals: for the retail sector as a whole and for the
sector excluding motor vehicles and parts dealers.

Manufacturing and Trade Inventories and Sales

June 2003
(p)

$1,168,312
430,508
448,755
289,049

Inventories ($ in millions)
May 2003
(r)

$1,167,232
431,356
446,914
288,962

June 2002

$1,130,803
428,230
418,653
283,920



151

Inventories/Sales Ratio
June 2003
May 2003 June 2002
(p)
(r)

1.38
1.33
1.59
1.22

1.40
1.34
1.59
1.23

1.38
1.33
1.55
1.24

Source: U.S. Department of Commerce, Bureau of the Census

Source:

WHAT DOES IT ALL MEAN?
What do the levels of inventories tell economists? The simple answer is, plenty. Low inventory positions may signal an impending
acceleration in production and manufacturing activity. Conversely,
very high inventories may portend a recession and widespread
layoffs. Wall Street equity analysts likewise can learn much from
inventories. By monitoring the levels in the industries they cover,
they can identify developing imbalances and potential troubles that
other indicators may not reveal.
Inventories are informative because they are central to the production process, which in turn is key to the health of individual
businesses, sectors, and the broader economy. As an illustration,
consider the recent history of the telecommunications industry.
During the dot-com heyday of the late 1990s and early 2000s,
companies desperate to keep up with the twenty-first-century zeitgeist were continually upgrading their Internet infrastructure with the
newest telecom technologies. Communications and fiber-optic equipment went from blueprint to development and production on what
seemed a monthly basis, and businesses and consumers eagerly bought
up the new products. Competing telecom-equipment makers, greedy
for market share, dropped prices. They also boosted production, both
to make up for their shrinking margins and to keep ahead of soaring
demand. Then the bubble burst. Everyone had upgraded as far as they
needed or could afford to, and the telecom manufacturers were left
with massive inventories. The only way to deplete these was by selling

152



The Trader’s Guide to Key Economic Indicators

stock at deep discounts. The resulting downward price spiral squeezed
profits. Companies slashed orders for goods they couldn’t sell, leading
to pullbacks in production and the layoff of thousands of idled workers. Stock prices collapsed, credit deteriorated, companies declared
bankruptcy—and the economy fell into recession.
An interested observer could have identified the potential formulation of a bubble in the telecommunications equipment industry
by merely observing this sharp increase in inventories. The graph
in FIGURE 7-2 shows that inventories in the nondefense communications industry increased steadily but evenly during the boom time
of the mid-1990s, as manufacturers managed to keep just ahead of
heavy demand. Around the turn of the century, though, inventories
suddenly shot up, reflecting the steep reduction in consumer purchases. The following period of recession was marked by companies
slowly clearing their shelves.
Not all accumulation or depletion of inventories is economically
significant; some degree of growth or shrinkage is part of the normal
course of business. The significance of a change in inventory level
depends in part on its cause—whether it was planned or unplanned.
The smart economist determines this by talking with corporate
officers in the industry experiencing these run-ups or depletions.
Inventories of Telecommunication Equipment Manufactures

YOY%

$ in millions

20,000

Shaded area = Recession

18,000

16,000

30
20

14,000

12,000

10,000

–10

8,000

–20

YOY%
$ in millions

–30

6,000
4,000

–40
1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

Source: U.S. Department of Commerce, Bureau of the Census

Figure 7-2

Manufacturing and Trade Inventories and Sales



153

INVENTORIES AND THE BUSINESS CYCLE
Inventories are obviously closely tied to economic activity. But are
they a lagging, coincident, or leading indicator? The answer is: It
depends on how you look at them. The level of inventory investment is generally considered a lagging indicator. This is because
businesses have traditionally been slow to recognize that demand
is drying up and their back stock getting dangerously high and thus
don’t start to draw down inventory until an economic slowdown is
already under way. Once they begin, however, the decline in inventory investment from business cycle peak to trough can be staggering, sometimes actually exceeding that in aggregate demand. At the
other end of the cycle, businesses have generally delayed ramping
up production and restocking their shelves until they’re certain a
recovery is underway.
Much of this lagging association can be identified in FIGURE 7-3,
which depicts the month-over-month change in manufacturing and
trade inventories.
Changes in inventory levels lead the business cycle insofar as
they reflect business expectations. If managements anticipate solid
economic growth down the road, they beef up their stocks. Alter-

Source: U.S. Department of Commerce, Bureau of the Census

Figure 7-3

Month-to-Month Dollar Change in Inventories

$ in thousands

14,000
10,000
6,000
2,000
–2,000
–6,000
–10,000
–14,000

Shaded area = Recession

–18,000
1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

154



The Trader’s Guide to Key Economic Indicators

natively, if they foresee lackluster demand, they permit current sales
to draw down their existing inventories. The object is to maximize
inventory levels with respect to anticipated demand. Achieving this
goal has become easier in recent years with the advent of technologies such as electronic scanners and supply-chain and related
inventory management systems. These innovations have improved
businesses’ ability to monitor merchandise levels, so that they don’t
overorder or understock, and have given them greater control over
inventory movements. The result has been a gentler business cycle
with recessions that are both shorter and milder.
The rate of change in total business inventories is a useful tool
in predicting business cycle turning points. But it can be calibrated
to an even finer instrument by looking at the numbers for individual industries, some of which have much more relevance than
others to economic activity. Inventory levels of nondurable goods
such as food, for instance, don’t offer much insight into the general
economic situation. The demand for these goods is fairly stable,
whether recession is imminent or not, and because the products
are, by definition, perishable, they never sit long on the shelf.
Much more revelatory are the inventory levels of more cyclical
sectors such as nondefense capital goods, automobiles, and consumer durable goods.

Among the most useful numbers in the MTIS report are the inventories-to-sales ratios. They indicate how many months it will take, at
the current sales pace, until inventories are entirely liquidated—that
is, until nothing is left to be sold. An I/S ratio of 1.40, for example,
means that at the current sales rate, businesses have 1.40 months of
inventories left on their shelves. From 1992 through 2002, the average ratio of total business inventory to sales was 1.45.
Because different types of goods have varied shelf lives, production schedules, and sales rates, the inventories-to-sales ratios for
different industries normally move within separate ranges. This
is clearly illustrated in FIGURE 7-4, which graphs the ratios for the

Source:

INVENTORIES-TO-SALES RATIOS

Manufacturing and Trade Inventories and Sales

Sources: U.S. Department of Commerce, Bureau of the Census;
NBER

Figure 7-4



155

Inventory-to-Sales Ratios: Retail, Wholesale, Manufacturing

I/S

1.8
Shaded area = Recession

1.7

Retailers
Manufacturers
Wholesalers

1.6
1.5
1.4
1.3
1.2
1992

1994

1996

1998

2000

2002

manufacturing, retail, and wholesale sectors. The chart shows that
from 1992 through 2003, retailers had the highest ratios of the
three groups, followed by manufacturers and then wholesale merchants. Security analysts who are aware of the range for their area of
specialization can check the MTIS report to see where the current
ratio fits within it.

Source:

HOW TO USE WHAT YOU SEE
Because the data in the MTIS are dated, it is more productive to
look at long-term trends in the inventory numbers than to try to
analyze those contained in individual reports. Many Wall Street
economists maintain running databases of inventories, sales, and
the inventories-to-sales ratios from the reports, updating the numbers with each release. This practice can be a bit tedious, but it is
well worth the undertaking.
The economics departments at most Wall Street research institutions provide monthly chartbooks to their security analysts.
These chartbooks inventory data in the Manufacturing and Trade
Inventories and Sales report so that potential production changes
can be identified. A sustained, abnormal run-up in monthly auto-

156



The Trader’s Guide to Key Economic Indicators

mobile inventory levels, for instance, could mean a manufacturing
slowdown in the industry down the road, together with slashed
prices or staff reductions. An analyst or economist noticing such
a trend might talk with automotive industry insiders to confirm or
refute the apparent production overhangs and identify their causes.
It’s always useful to have a Rolodex of insiders’ numbers.

TRICK FROM

THE

TRENCHES

Source:

The trick for this chapter involves a different way of looking at
data that are there in the MTIS report for anyone to see: the inventories-to-sales ratios. As was noted, these have long been considered good indicators of future economic activity. In the period
before the conversion to NAICS, an I/S ratio between 1.55 and
1.60 was taken to mean that recession was imminent. That indication no longer seems to be the case. As also was noted, innovations
such as just-in-time inventory and supply-chain management systems have given businesses a better grasp of appropriate inventory
levels and better control over the production-shipment-inventory
process. As a result, no specific I/S number can be an accurate
recession signal.
Instead, economists look for pronounced movements in the
ratio that are sustained over several months, such as the rapid runup from 1.38 in March 2000 to 1.46 in April 2001. When the I/S
ratio rises over time, it means that sales are not strong enough to
reduce inventories or that goods are being accumulated at too fast
a pace. The bottom line is that sales are slower than companies had
anticipated. This is a bad sign for the economy. As FIGURE 7-5 shows, a
recession soon followed the 2000–2001 buildup. Conversely, when
the ratio of inventories to sales falls over several months, it means
that sales are growing faster than inventories and that manufacturers may soon have to boost production. This, of course, is good for
overall economic activity.

Manufacturing and Trade Inventories and Sales

Source:

Sources: U.S. Department of Commerce, Bureau of the Census;
NBER

Figure 7-5



157

Total Business Inventory-to-Sales Ratio

I/S

1.6

1.5

1.4
Shaded area = Recession

1.3
1992 1993

1994

1995

1996 1997

1998

1999 2000

2001

2002

2003

This page intentionally left blank

New Residential
Construction

f you learn one lesson from this book, let it be this: never underestimate the economic importance of housing. The realization of the
American Dream of home ownership is one of the primary drivers of
the economy. Housing activity affects the investment (I ) component
of the aggregate expenditure formula for calculating gross domestic
product (see Chapter 1): C + I + G + (X–M ). The construction of
new, privately owned residential structures, particularly single-family
homes, is very informative about consumer sentiment and the health
of the economy. After all, a purchase of this magnitude requires the
utmost confidence in one’s personal financial situation, including
employment security and earnings prospects. It also implies favorable
conditions in the broader economy. Beyond personal income, the primary influences on housing activity are the level of interest rates and
demographics. When these factors are aligned correctly, a buoyant
housing market can boost total economic activity; when they’re not, a
slumping market can drag the overall economy into deep recession.
How can new housing construction, which accounts directly
for only 3 percent of GDP, influence the economy so profoundly?
Because of the multiplier effect of related spending and other indirect contributions. Once a home is bought, it must be furnished
and decorated. All of this activity means new jobs for construction
workers, retail salespeople, and manufacturers; increased tax revenues for local and state municipalities; and greater spending on
goods such as carpeting, furniture, and appliances. Of course, these
jobs and revenues fail to materialize if unfavorable conditions, such
as rising interest rates, stifle demand for new homes.
159

160



The Trader’s Guide to Key Economic Indicators

Several important housing indicators exist, including the
Census Bureau’s new home sales and the National Association
of Realtors’ existing home sales. The most influential, however,
are new-housing starts and building permits. These statistics are
contained in New Residential Construction, which is released
jointly by the U.S. Department of Commerce’s Census Bureau
and the U.S. Department of Housing and Urban Development,
at 8:30 a.m. ET on approximately the fifteenth day of the month
following the reference month. The release, as well as detail on
the different stages of the construction process and a description
of the methodology used, is available on the Census Bureau’s
website, at www.census.gov/const/www/newresconstindex.html.
The data contained in New Residential Construction, informally
called “housing starts” by market participants, are among the most
respected economic indicators on Wall Street and should be included in any serious model of U.S. economic activity.

Information regarding the origins of the new residential construction report is limited. Truth be told, there is little meaningful history to speak of. The Commerce Department’s U.S. Census Bureau
assumed the responsibilities of the collection of housing starts data
back in 1959. Previously, the duties of aggregating, calculating, and
distributing the starts and permits data belonged to the Department
of Labor’s Bureau of Labor Statistics (BLS). The data that the BLS
had gathered dated back to 1889.
The report has remained basically the same, in form and methodology, since 1959. A few minor changes have occurred along
the way, such as the inclusion of farm housing and the exclusion
of public housing. Some definitions have also been modified. In
the wake of Hurricane Andrew in 1992, for instance, the meaning
of “housing starts” was expanded to include rebuilding on existing
foundations. In the 1990s, a correction was made so that starts
without legal permits would be included, resulting in a break in
the series. This break has had minor consequence on the conti-

Source:

EVOLUTION OF AN INDICATOR

New Residential Construction



161

nuity of the series. In 1968, the bureau began reporting detailed
housing characteristics, such as completions. For the most part,
however, any and all of these changes have had a negligible effect
on the monthly reports.

Source:

DIGGING FOR THE DATA
The New Residential Construction report contains data on privately
owned residential structures, both single family and multifamily.
Each unit in the multifamily structures is counted separately; so a
building with ten apartments, for example, is counted as ten units.
Excluded are public housing and hotels, motels, group residential
structures such as college dormitories and nursing homes, and mobile homes.
The release consists of a summary of the month’s figures followed by five tables. Each table corresponds to five stages in the
residential construction process. The first contains data on the
number of new, privately owned housing units for which applications for building permits, required for construction in most U.S.
regions, have been filed. The figures are compiled from responses
to mail surveys sent to building permit officials in 8,500 permitissuing localities, out of a universe of some 19,000.
The Census Bureau derives a sample of 900 permit-issuing
places in the Survey of Construction (SOC), which is the source for
the Housing Units Authorized, but Not Started; Housing Starts;
Housing Units Under Construction; and Housing Completions
(Tables 2 through 5) data. This sample is chosen with respect to
the labor force, race/ethnic origin, population change, and family
and housing characteristics in the respective permit-issuing place.
Census Bureau field representatives canvass these 900 places, as
well as the less than 3 percent of areas that don’t require permits,
and estimate the characteristics of the several stages of construction
(i.e., was the unit started, completed, and so forth).
The second table contains data for units whose construction has
been authorized but not yet begun. The Census Bureau produced
this table because the majority of new construction typically begins



The Trader’s Guide to Key Economic Indicators

during the month of permit issuance, and is included in Table 1.
The difference between the two tables is the number of housing
units that have been authorized by zoning or building permit,
where construction was and wasn’t started.
The third table presents the number of housing starts—that
is, the number of units for which excavation of the footings or
foundation has begun. This is the headliner of the New Residential Construction report. By this stage, money has generally been
exchanged, implying that owners are committed to the construction
project. The fourth table includes units “under construction”—that
is, where work has begun but not yet been completed. These data
closely parallel the changes in residential construction figures contained in the Commerce Department’s Value of Construction Put in
Place report, referred to on the Street as construction spending.
The fifth table analyzes completions.
By breaking down the numbers according to stages in the housing construction process, the tables allow economists, investors,
and other market participants to pinpoint where the strengths and
weaknesses lie. For example, should permits, starts, and construction all pick up but completions remain flat, homebuilders may be
running up against troubling economic conditions such as an increase in the unemployment rate that could hinder would-be home
buyers from purchasing a new home. Also, a rising interest rate
environment could create difficulties for those buyers attempting to
obtain financing at reasonable terms.
The tables present twelve months of data—preliminary figures for
the record month, revisions for the two previous ones, and final figures going back a year—in both seasonally adjusted annualized form
and as unadjusted monthly numbers. They also include the unadjusted
annual numbers for the previous two years and the year-to-date totals
for the current year and the one before, as well as the adjusted annualized percent changes between the record month and the previous
month and between the record month and year-earlier one.
The numbers of units at each stage of construction are given
for the entire United States as well as for four regions: Northeast,
Midwest, South, and West. The figures for the entire country are

Source:

162

New Residential Construction



163

broken down by size of structure, into single-family dwellings, residences with two to four units, and those with five or more units.

Source:

WHAT DOES IT ALL MEAN
Residential housing investment has a ripple effect on the overall
economy that is far greater than its direct contribution to GDP.
This is because of the amount of labor and the number and volume
of products involved in constructing and furnishing a home. Just
consider the following list, from the National Association of Home
Builders (NAHB), of resources used in an average 2,000-squarefoot single-family home:
 13,127 board feet of framing lumber
 6,212 square feet of sheathing
 13.97 tons of concrete
 2,325 square feet of exterior siding material
 3,100 square feet of roofing material
 3,061 square feet of insulation
 6,144 square feet of interior wall material
 120 linear feet of ducting
 15 windows
 13 kitchen cabinets, 2 other cabinets
 1 kitchen sink
 12 interior doors
 7 closet doors
 2 exterior doors
 1 patio door
 2 garage doors
 1 fireplace
 3 toilets, 2 bathtubs, 1 shower stall
 3 bathroom sinks
 2,085 square feet of flooring material, such as carpeting, resilient sheet, resilient tile, ceramic tile, or wood plank
 1 range, 1 refrigerator, 1 dishwasher, 1 garbage disposal, 1 range
hood
 1 washer, 1 dryer

164



The Trader’s Guide to Key Economic Indicators

Of course, those timbers have to be joined, the concrete poured,
the sheathing, siding, roofing, wall material, and insulation applied,
and the windows, doors, appliances, and other amenities installed.
The NAHB has concluded that building 1,000 single-family homes
generates 2,448 full-time jobs, $79.4 million in wages, and $42.5
million in government revenues.
Even after houses are built and furnished, they can continue to
stimulate economic activity—and not just through remodeling and
other forms of upgrading. People’s homes are often their largest
asset, and thus their largest potential source of capital. When mortgage rates fall, home owners refinance at the lower rates, either
reducing their monthly payments or increasing the amount of their
loans, or some combination of the two.
Research has shown that most of the savings realized are poured
back into the home. It should thus come as no surprise that do-ityourself home centers such as Home Depot and Lowe’s Corporation performed so well during the mortgage refinancing boom of
the late 1990s and early 2000s. Furthermore, because consumer
spending is a large component of GDP, the entire economy benefits
from this activity. In fact, many economists credit the mildness of
the recent recession to the surge in mortgage refinancings driven by
the low interest rates during the period.

Clearly, housing investment has a profound effect on the economy.
The reverse is also true: Residential construction is one of the most
economically sensitive and cyclical sectors. Two important influences on housing are interest rates and demographics.
Because their homes are the most expensive purchases that most
people make, borrowing is almost a certainty—unless, of course,
they’re big lottery winners. It’s not surprising, therefore, that housing starts are correlated with the level of interest rates: the lower the
30-year fixed mortgage rate, the rate at which most wannabe home
owners borrow, the less onerous the loan and the higher the number
of starts. FIGURE 8-1 shows the relationship between mortgage rates

Source:

INFLUENCES ON RESIDENTIAL CONSTRUCTION

New Residential Construction



165

Figure 8-1 Single-Family Housing Starts and the 30-Year Fixed Mortgage Rate
Starts in thousands

Rate %

Source:

Sources: FHLMC; U.S. Department of Commerce

1,600

8
1,200
12
800
1-Family starts
Mortgage rate

400

20
1972

1975

1978

1981

1984

1987

1990

1993

1996

1999

2002

and starts (an inverse one—note the inverted right-hand scale). A
general rule of thumb is that as long as income and employment are
both growing steadily and home prices remain reasonable, rates of
7.25 percent or lower will boost housing activity.
Because it indirectly influences the 30-year rate, the Federal
Reserve has the power to boost or dampen housing activity. When
economic growth threatens to spark high inflation, the Fed may
attempt to slow things down by raising its overnight borrowing
90
20
rate (see Chapter 1). This increase usually travels along the maturity80spectrum, eventually resulting in higher mortgage rates. The
15
consequent reduction in house building and sales, combined with
70
10
retrenchments
in other capital spending caused by the Fed tightening, slows economic growth. If the Fed is overzealous and raises
5
60
interest rates too high, a recession can set in. On the other hand,
when
50 the economy is sluggish, the Fed often acts to stimulate it 0by
lowering its overnight target rate, which bolsters demand for and
-5
40
construction
of new homes.
Also important in determining housing activity are the size -10
and
30
1980
1983
1986
1989
1992
1995
1998
2001
composition
of the population.
Obviously,
the more
people
there are of
home-buying age, the greater the potential activity. Economists have
found that the number of people between 30 and 59 has the strongest

166



The Trader’s Guide to Key Economic Indicators

correlation with the number of new and existing home sales, as well as
with the pace of single-family residential construction. Remember, it
generally takes a while after graduation for individuals to accumulate
enough money for a down payment on a home. Furthermore, earnings
for individuals in their 20s generally don’t amount to much.
People between 20 and 25 and those 60 and above are more closely
linked to multiple-unit construction, because recent graduates and
the elderly tend to reside in multifamily structures, such as apartment buildings. The first of the 77 million or so baby boomers—
Americans born between 1946 and 1964—reached the average
home-purchasing age, 32, in 1978. This demographic shift not only
boosted the housing market over the past quarter century, but also
played an integral role in greater economic activity in general.
Immigration increases the population and thus the demand for
housing. The Federal National Mortgage Association (FNMA) has
found that newcomers to the United States typically realize the
American Dream about ten to fifteen years after arriving. Given the
incredible improvement in immigrant incomes and the interest rate
environment, recent immigrants might become first-time home
owners a bit quicker than that. Already, the tsunami of immigration
in the late 1900s and early 2000s has buoyed new construction activity, with some homebuilders crediting the housing boom of late
as being related more to improvements in immigrant welfare than
to the near-record low interest rate environment.
Because of their sensitivity to interest rates, demographics, and
other factors, such as taxes and weather, monthly housing data are
very volatile. To discern trends more easily, economists smooth out
this volatility by looking at new construction activity on a quarterly
or a trending three-month basis.

Many of the factors that can affect construction activity vary considerably from region to region across the United States. Midwestern
agricultural communities, for instance, will typically suffer disturbances in employment and income growth at different times and

Source:

REGIONAL DIFFERENCES

New Residential Construction



167

Source: U.S. Department of Commerce, Bureau of the Census

Figure 8-2 Regional Composition of New Residential Construction (1990–2003)
Northeast 8%

West 24%

Midwest 21%

South 47%

under different economic conditions than areas with more manufacturing industries, such as the Rust Belt, and the West Coast with
its concentration of technology companies. Weather conditions also
vary regionally, with hurricanes in the Southeast, cold winters in the
Northeast and Midwest, and flooding and drought in the South and
Southwest. Furthermore, although monetary policy, which determines interest rates, is national, other policy matters, such as tax incentives, differ from state to state. Not surprisingly, then, there is a
great disparity among the housing statistics for the different regions
represented in the New Residential Construction report.
As shown in FIGURE 8-2, throughout the 1990s and the early 2000s,
the Northeast has accounted for less than 10 percent of the total of new
private housing starts. The Midwest, West, and South, meanwhile,
have contributed approximately 20, 25, and 45 percent, respectively.

Source:

HOUSING AND THE BUSINESS CYCLE
The preceding sections described the close two-way relationship between housing and the economy. But what does housing activity—and,
more precisely, the statistics contained in the new construction report—
tell us about different stages and turning points in the business cycle?



The Trader’s Guide to Key Economic Indicators

Housing’s relationship with the business cycle may best be described as FIFO, or first in, first out: Housing is one of the first
industries to head south before a downturn and among the first to
pull out of recession. In other words, it is somewhat of a leading
indicator. Not all the stages of construction, however, have the same
predictive power, nor do they receive the same attention.
The number of permit applications is the most forward-looking of all the statistics in this report. It is so forward-looking,
in fact, that the Conference Board includes it in the index of
leading economic indicators (see Chapter 2). Although permit
applications don’t always result in starts, economists assume that
contractors won’t go through the trouble, time, and expense of
applying for a permit unless they are very serious about building a home. Because this statistic is a gauge of future demand
for construction and housing, equity analysts covering the home
builders and real estate investment trust (REIT) markets should
keep an eye on it.
Economists have found that privately-owned housing units authorized by building permits generally precede housing starts by
about one month and sales by about three. The number of permit
applications peaks as much as twelve months before the onset of a
recession and bottoms out, for the most part, almost simultaneously
with the overall economy.
The number of units authorized for construction, representing a
later stage in the process, is slightly more accurate than the number
of applications as a measure of future building activity. Economists
have found that this indicator leads new starts by about two months.
Units under construction depict the current state of construction
activity, construction-industry employment, and the demand for
building materials.
Housing completions can point to problems brewing in the
economy. A significant gap between units under construction
and the number completed may be due to inclement weather—
hurricanes in the South and blizzards in the Northeast and West
frequently cause lengthy delays in the building process. But a lag
in completions can also mean that a large number of builders or

Source:

168

New Residential Construction



169

buyers cannot afford to finish their projects. That in turn may signal financial weakness in the overall economy.

Source:

SINGLE-FAMILY HOUSING STARTS
Although the statistics for the other stages of construction are informative, the headliner of the New Residential Construction report
is the number of housing starts—more specifically, those relating
to single-unit structures. Wall Street focuses on single-family residences rather than multiunit ones because the figures for the latter
are more volatile. Multiunit structures typically include apartments,
condominiums, and townhouses. As noted earlier, each apartment
in one of these buildings is counted as a separate unit. So when
construction of a 500-apartment building is started or canceled, 500
units at a time are added to or eliminated from the monthly tally.
This can cause extreme month-to-month swings.
Multiunit housing is also more sensitive to tax policy. The U.S.
government, for instance, provides the incentives for building in
underdeveloped neighborhoods; these incentives usually involve
multi- rather than single-unit structures. Because such incentives
change with the political landscape, they can add to the volatility of
the multiunit numbers. Finally, the pace of multiunit construction
may be partly a function of the growth rate of single-family housing. In the late 1990s and early 2000s, for instance, many apartment
dwellers left them to buy single-family homes, which had become
much more affordable. This boosted single-unit numbers and depressed multiunit ones.
Single-family starts, as illustrated in FIGURE 8-3, typically account
for close to 80 percent of all new private housing starts. This
statistic may be the ultimate gauge of consumer confidence; and
because consumer activity accounts for upwards of 70 percent of
total U.S. economic output, single-family starts deserve the attention they get.
Why is the level of single-family housing starts such an excellent
barometer of the health of the consumer sector? Simply put, when
people are concerned about their economic situation, they may

170



The Trader’s Guide to Key Economic Indicators

2,500
2,250

1-Family starts
Total

2,000
1,750
1,500
1,250
1,000
750
500
1948

1964

1969

1974

1979

1984

1989

1994

1999

still spend on other products but they won’t even consider buying
homes. This is another reason why the housing market is such an
excellent indicator of business cycle activity.
Before the 2001 recession, housing starts were the most reliable
and accurate measure of U.S. economic health. Every post–World
War II economic recession was accompanied by a precipitous decline in new housing starts and housing activity in general, which
triggered similar reductions in consumer expenditures for related
goods and in the employment of workers in associated industries.
The 2001 recession broke this pattern. Housing starts remained
strong during the downturn because historically low inflation
kept mortgage rates low. Multidecade lows in the unemployment
rate and soaring personal incomes from the torrid pace of payroll
growth during the 1990s helped fuel future (i.e., recession-period)
home and related purchases. The same factors—plus the mortgage
refinancing mentioned earlier—kept consumer spending in positive
territory. In fact, the worst showing for expenditures, occurring
during the second quarter of 2001, was a 1.4 percent increase. Not
coincidentally, a great deal of this spending was on home-related
goods, along with renovations, additions, and other home improvements. All this kept the 2001 recession very mild.

Source:

Units in thousands

Source: U.S. Department of Commerce, Bureau of the Census

Figure 8-3 Total and Single-Family Housing Starts

New Residential Construction



171

Another reason that home purchases and related spending
remained strong during the recession and through 2002 was the
dismal performance of the stock market. Recognizing that they
could do little or nothing to improve their portfolios’ value, people
worked to increase the value of their greatest asset, their home. It is
important to remember that at this time, home owners (at 67 percent) outnumbered shareholders (50 percent of U.S. households)
in the United States. Home ownership is defined by the Census
Bureau as the proportion of the number of households that are
owners to the total number of households. After the September
11, 2001, attacks, moreover, Americans became nesters, vacationing within 100 miles of their homes and eliminating unnecessary
business travel. They also started buying second homes as getaways
from primary residences in, or near, big cities. Many of the purchasers were baby boomers looking toward retirement, whose second
homes could become their primary post-retirement residences.

Source:

HOW TO USE WHAT YOU SEE
As always, one of the goals for market pros is to get a jump on the
most market-moving statistics in the report. In the case of the New
Residential Construction report, one strategy would be to gather
anecdotal evidence from the home builders themselves. Many
of the nation’s largest builders—including K&B Homes, MDC
Holdings, Pulte Homes, Inc., Toll Brothers, Lennar Corporation,
Ryland Group, Hovnanian Enterprises, Engle Homes, and Beazer
Homes USA—talk about the issues they are facing in their quarterly statements or in frequent presentations, which are available on
their websites. In addition, the NAHB provides excellent statistical
measures of the housing market, remodeling trends, and housing
affordability. The Mortgage Bankers Association (MBA) has created
a few of the Street’s most respected economic indicators on mortgage applications for purchases and refinancings. It’s also a good
idea to keep an eye on the loan and default data on the Federal
Reserve’s website, which can provide early signals of troubles in
the industry. Reluctance by loan officers to extend housing credit

172



The Trader’s Guide to Key Economic Indicators

may signal that banks are having problems collecting on extended
loans. Specifically, interested readers should watch the household
detail in the Federal Reserve’s Senior Loan Officer Opinion Survey on Bank Lending Practices. It is available on their website,
www.federalreserve.gov.

TRICKS FROM

THE

TRENCHES

Economists have found that the difference between the number of
total privately-owned housing starts and the number of privatelyowned housing units authorized by building permits yields a crude
gauge of housing activity that is more informative than the number
of starts by itself. Permits here can be seen as a proxy for housing
demand and starts as a proxy for housing supply. The resulting
“spread” is most useful as an indicator of boom conditions in the
housing sector.
A negative spread, with permits lagging behind starts, is characteristic of normal or even moderately expansionary housing market.
As FIGURE 8-4 shows, such a spread existed from 1992 to 1997. During
this six-year period, new housing starts averaged about 1.4 million
units (annualized), a reasonably strong pace by historical standards.

In thousands

200
150
100
50
0
–50
–100
–150
–200

1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

Source: U.S. Department of Commerce, Bureau of the Census

Figure 8-4 Difference Between Housing Permits and Starts

New Residential Construction



173

Source:

Positive spreads, when permits exceed starts, existed from 1998
to mid-2002. Economists interpreted these as signals of a boom
in future housing activity. When housing demand exceeds supply
by a significant amount, prices generally rise. From 1998 to 2002,
median new home prices advanced 28 percent. A boom generally
occurs when all economic cylinders are firing, inflation and interest
rates are low, and incomes and employment are on the rise.

This page intentionally left blank

Conference Board
Consumer Confidence and
University of Michigan
Consumer Sentiment Indices

ost, probably, of our decisions to do something positive,
the full consequences of which will be drawn out over many
days to come, can only be taken as a result of animal spirits—of a
spontaneous urge to action rather than inaction, and not as the outcome of a weighted average of quantitative benefits multiplied by
quantitative probabilities.” These words—by pioneering economic
theorist John Maynard Keynes in his 1936 opus, The General Theory
of Employment, Interest, and Money—may be the first written recognition of the importance of consumer sentiment in economics.
The need to quantify these “animal spirits” is the motivating force
behind the consumer confidence measures of today.
When it comes to the economy, consumers are king, accounting for roughly two-thirds of gross domestic product through their
spending on items from abacuses to Zip disks. It is always good to
know how the king is feeling. When people feel confident about
their financial situation and future, they are usually reacting to
some positive economic fundamental, such as solid employment
growth or rising personal incomes. It is not surprising, then, that
measures of consumer attitude have produced an impressive record
of predicting economic turning points.
Many surveys of consumer confidence and sentiment exist.
Some research institutions and investment firms have even created
their own, to get the data into their systems and transformed into
forecasts and strategies as quickly as possible. Although differing in
specifics, the various surveys share one crucial characteristic: They
ask everyday people from different walks of life simple questions
175



The Trader’s Guide to Key Economic Indicators

that probe their feelings about the current and future state of the
economy, inflation, and their plans for vehicle and home purchases.
The survey participants may not know the difference between recession and depression, but their answers provide insights into the
likelihood of these situations occurring.
Of all the surveys and measures, the best known and most respected are the Conference Board’s Consumer Confidence Index
and the University of Michigan’s Index of Consumer Sentiment.
(The weekly ABC News/Money Magazine Consumer Comfort
Index is another outstanding measure with exceptional links to consumer retail sales activity, but it is less popular than the other two,
possibly because its weekly schedule makes it extremely volatile and
therefore less market-moving.)
The Conference Board’s confidence index is generally released
on the last Tuesday of each month and made available to the investing public (in a limited version) on the Conference Board’s
website at www.conference-board.org. A more detailed version,
as well as its history, is available by subscription directly from the
Conference Board.
The University of Michigan usually issues its sentiment index
on the second to last Friday of each month, followed by the revised final estimate two weeks later. This survey is available by
subscription only. An historical series of the Sentiment Index and
the Inflation Expectation Index is available at the Federal Reserve
Bank of St. Louis website at http://research.stlouisfed.org/fred2/.
The results that come across the newswires or are posted at the
more popular business news websites are sufficient for most readers’ purposes.
Every Wall Street economics department that is serious about
forecasting growth subscribes to these surveys. They are also frequently cited by Federal Reserve and White House officials and are
the object of countless studies by government agencies seeking to
determine if economic policies are working as intended. After the
Federal Reserve lowers the overnight interest rate, for instance, it
will want to know if this move has encouraged consumers to buy
interest-rate-sensitive goods, such as houses and automobiles. The

Source:

176

Consumer Confidence and Consumer Sentiment Indices



177

answer could well determine whether the Fed will lower rates further or not.
The markets generally react most sharply to the confidence
measures when the business cycle is close to a turning point. The
indices won’t indicate how much the economy may grow or fall
back. But when they decline sharply, you can bet on tough economic sledding ahead. Conversely, when the indices spike upward,
you can look forward to more prosperous times.

EVOLUTION OF AN INDICATOR
George Katona, a Hungarian-born psychologist and economist
with the Survey Research Center of the University of Michigan,
started asking questions about consumer intentions in the 1940s.
The university began conducting its survey in 1946, using many
of the same questions that it does today. The survey, originally an
annual event, switched to a quarterly schedule in 1960 and then a
monthly one in 1978.
Throughout the 1940s and 1950s, Katona and University of
Michigan economics professor Eva Mueller were virtually alone in
the field, together producing a weighty corpus of studies of consumer attitudes, incomes, and spending habits. The Conference
Board started conducting bimonthly surveys of consumer attitudes
only in 1967. It, too, converted to a monthly schedule ten years
later, in June 1977.

Source:

DIGGING FOR THE DATA
Although both the Conference Board’s and the University of
Michigan’s indices measure consumer confidence and expectations,
the underlying surveys pose different questions and poll sample
groups of different size and breadth. For its indices, the University
of Michigan’s Survey Research Center polls 500 households in the
lower forty-eight states by telephone, asking participants about
their personal finances, general business conditions, and planned
purchases. The entire survey consists of more than two dozen core



The Trader’s Guide to Key Economic Indicators

questions, which serve as the basis for several indices. The Index
of Consumer Sentiment (ICS) is created from the responses to five
of the questions, those asking respondents (1) if they’re better or
worse off financially than a year ago; (2) whether the year to come
will be better or worse for them financially; (3) how businesses will
fare in the next twelve months; (4) whether the country during the
next five years or so will experience good times or widespread unemployment and depression; and (5) whether it is a good or bad
time to buy major household items. Responses are classified as
positive, neutral, or negative. Two sub-indices are formed from different subsets of the five responses: the Index of Current Economic
Conditions (ICC), from replies to questions 1 and 5, and the Index
of Consumer Expectations (ICE), from 2, 3, and 4.
The ICS, ICC, and ICE are all constructed using diffusion
methodology: The positive responses are added up and the result
divided by the sum of positive and negative responses to yield a relative value. The index values are calculated relative to a base month
of January 1966, whose value is set at 100. The Center generally
releases preliminary indices on the second Friday of the month following the record month and issues final versions within two weeks
after that.
The survey on which the Conference Board bases its indices
is conducted by Greenwich, Connecticut–based NFO Research.
NFO polls a panel of about 5,000 households on their assessment
of current economic conditions, their expectations for the future,
and their plans for major purchases in the next six months. Like
Michigan’s Survey Research Center, the Conference Board constructs three diffusion indices from the responses to five of these
questions. These five ask respondents (1) how they rate general
business conditions in their area; (2) what conditions they foresee
in six months; (3) how they would characterize current job availability in their area; (4) how they think availability will compare
in six months; and (5) how they think their family income in six
months will compare with their current income. The Consumer
Confidence Index is constructed from the responses to all five questions; the Present Situation Index from answers to 1 and 3; and the

Source:

178

Consumer Confidence and Consumer Sentiment Indices



179

Expectations Index from 2, 4, and 5. All three indices are calculated
relative to the base year 1985, whose value is set at 100. Unlike the
Survey Research Center, the Conference Board adjusts the survey
statistics for seasonal variations.
The Conference Board receives preliminary results eighteen
to twenty-one days into the record month, so the indices reflect
the conditions of the month’s first three weeks. Any late surveys
are retained and used in the revision contained in the next month’s
release. The index numbers in the monthly report are kept secret
until the official release. The only exceptions are for the Conference Board officials charged with writing the accompanying commentary, the Federal Reserve, and the White House’s Council of
Economic Advisors. The Fed and the CEA receive the report after
4:00 p.m. ET the day before the official release. The Conference
Board also releases, separately, consumer confidence data for nine
geographical regions.

Source:

WHAT DOES IT ALL MEAN?
The differences in the methodologies used by University of Michigan’s Survey Research Center and the Conference Board are small
but still important enough to produce indices with somewhat divergent characteristics and strengths. On the one hand, many economists feel that the larger pool sampled in the NFO survey makes the
Conference Board’s indices more significant statistically. They also
believe that eliciting expectations for the next six months, as the
NFO survey does, is more realistic than the Michigan survey’s practice of asking for a five-year perspective. On the other hand, the
longer history and the twice-monthly reporting of the sentiment
indices make the University of Michigan report one of Wall Street’s
favorites. Serious investors would be best served by subscribing to
both surveys.
That said, the similarities between the two sets of indices are in
many ways more important than their differences. The most obvious similarity is that they all move in a close association with the
business cycle.

180



The Trader’s Guide to Key Economic Indicators

Figure 9-1 University of Michigan Consumer Sentiment Index and Conference
Board Coincident Index

120

110

100
100

90
80

Michigan Sentiment
Coincident

40
1978

1981

1984

1987

1990

1993

1996

1999

2002

FIGURE 9-1 shows the University of Michigan’s Index of Consumer
Sentiment charted against the Conference Board’s coincident indicators index. As explained in Chapter 2, the coincident index reflects the current condition of the economy; it rises during periods
of expanding economic activity and declines during recessions or
periods of retarded growth. Note that current economic activity, as
measured by the coincident index, flat-lines or declines each time
the University of Michigan’s Index of Consumer Sentiment falls
below 80. When the index is below that level for a sustained period,
the economy is extremely sluggish. Such was the case from late 1978
through early 1983. During that stretch, the economy was mired in
recession, plagued by some of the poorest economic fundamentals
in history: Inflation topped 14 percent, unemployment broke 10
percent, and both the prime and the Fed funds rate exceeded 20
percent. No wonder consumer confidence tumbled. Consecutive
index readings above 90, conversely, coincide with periods of relative prosperity, marked by a rising trend in the coincident index.
The picture is similar for the Conference Board’s Consumer
Confidence Index. Lynn Franco, the director of the Conference
Board’s Consumer Research Center, says that a reading of 100 or

Sources: University of Michigan; The Conference Board

Coincident

Source:

Michigan Sentiment

120

Consumer Confidence and Consumer Sentiment Indices



181

Figure 9-2 Conference Board’s Consumer Confidence Index and Recessions
160

Source:

Sources: The Conference Board; NBER

140
120
100
80
60

Shaded areas = Recession

40
1978

1981

1984

1987

1990

1993

1996

1999

2002

above indicates the economy is expanding. During the 120 months
from April 1991 through March 2001, which marked the longest
economic expansion in U.S. history, the confidence index stood
above 100 for 69 months and above 90 for 81. At the onset of a
recession, in contrast, the index usually dips below 80. This is illustrated in FIGURE 9-2, which shows the confidence index overlaid by
highlighted bars representing recessions, as identified by the National Bureau of Economic Research (NBER), the official arbiter of
U.S. business cycles. Note that during the 1981–82 recession, the
index was below 80 for fifteen of the sixteen months, and during the
1990–91 recession for five of the eight months.
Figures 9-1 and 9-2 also illustrate one drawback of the two indices as economic indicators: Because of their strong links to consumer activity, they may fail to identify turning points that are not
consumer-driven. Figure 9-1, for example, shows that the Michigan
sentiment index badly lagged the 1991 economic recovery, remaining in the 60s and 70s even as the economy was steadily improving.
This “hangover” reflects the fact that the job market remained
stagnant for more than two years after the official end of the recession in March 1991. Labor conditions have a profound effect on
consumer attitudes.

182



The Trader’s Guide to Key Economic Indicators

The Conference Board’s confidence index has also missed the
mark from time to time, as witness its behavior during the 2001–02
recession, represented in Figure 9-2. During the entire seven
months from April 2001 to November 2001 that the economy was
in recession, the index never fell below 80. In fact, its lowest reading
during the recession was 84.9, recorded in November 2001. This
reflects the fact that unlike other post–World War II downturns,
the recession of 2001 was the result of a precipitous decline in business investment rather than a retrenchment in consumer spending.
In fact, spending was buoyed by low, 4.6 percent unemployment,
mild 2.5 percent inflation, and similarly low interest rates.
Another potential source of misleading sentiment readings is the
lag that occurs between conducting the surveys and releasing the
indices. The Conference Board indices reflect the conditions of the
first three weeks of the month. Any earth-shattering news occurring
after the twenty-first of the month will not be reflected in the confidence level recorded in the release. A dramatic example of this was
the stock market crash of 1987. Black Monday, when the Dow Jones
Industrial Average tumbled a record 508 points, was October 19.
Many responses to the NFO survey were elicited before this date.
As a result, the Confidence Index fell a mere 0.6 points in October,
to 115.1 from 115.7 in September. By November, everyone was
aware of the crash, and the index declined nearly fifteen points, to
100.8. Of course, 100 is still far from 80, the usual recession level.
The Confidence Index was thus one of the few indicators that accurately predicted economic prosperity after the stock market crash.

Although the overall confidence and sentiment indices possess
moderate predictive power, as illustrated in the preceding example,
they are more likely to move in synchrony with the business cycle.
The true powerhouses of predictability are the two expectations
indices constructed by the Conference Board and University of
Michigan. In fact, the University of Michigan’s Index of Consumer
Expectations (ICE) is included in the Conference Board’s leading

Source:

THE EXPECTATION INDICES

Source:

Consumer Confidence and Consumer Sentiment Indices



183

economic indicators index (LEI). The University of Michigan’s
Sentiment Index predates the Conference Board’s measure and was
used in the LEI prior to the Conference Board’s formal acceptance
of the calculation, publication, and caretaking responsibilities of the
Business Cycle Indicators in 1996.
Before a recession, the two expectations indices typically experience considerable declines for about five to six months. There is
no specific reading or percentage drop associated with either index
that signals a definitive downturn. However, momentous declines
in either index pretty much alert the analyst of a probable economic
downturn. For example, prior to the commencement of the August
1990 recession, the Conference Board’s Expectations Index tumbled from a reading of 100.3 in May to 74.2 in August. The Index
continued to decline into the 50s throughout the remainder of the
year. Declines in the Conference Board’s Expectations Index of this
magnitude have historically preceded downturns in the economy by
an average of about three months, and the University of Michigan’s
ICE by two months. The longest lead for both indices was six
months, before the beginning of the downturn in April 2001.
The indices are most accurate in predicting recessions when
their movements over an entire year are considered. The expectations indices of the Conference Board and the University of Michigan tumbled an average of 24 percent and 18 percent, respectively,
over the twelve months leading up to the 1980, 1990, and 2001
recessions. Interestingly, the total confidence and sentiment indices
are also strongly predictive when looked at in this way. In the twelve
months before those same recessions, the Conference Board’s index
declined 20.9 percent on average and the University of Michigan’s
index 14.4 percent.
The expectations indices tend to possess less predictive powers
when it comes to recoveries. For example, when the 1990–91 recession ended in March 1991, the Conference Board’s Expectations
Index bottomed in January 1991 with a reading of 5.3, and followed
with gains in February to 63.3 and a tremendous surge in March to
100.7. Similarly, the Conference Board’s Expectations Index bottomed in October 2001 with a reading of 70.7, a month before the

184



The Trader’s Guide to Key Economic Indicators

“official” end of the recession in November when the Expectations
Index had a reading of 92.4. The expectations indices clearly identify recoveries, but with little lead time.
CONFIDENCE AND DURABLES SPENDING
People generally buy big-ticket durable goods, such as stoves,
refrigerators, and autos, on credit, which means committing to
principal and interest payments over a long period. In making
such purchases, therefore, consumers need to take stock of their
financial future. That entails considering their employment status,
potential income growth—even the state of the economy. These are
the very considerations covered in the forward-looking questions
of the Conference Board and University of Michigan surveys. It’s
hardly surprising, then, that the two expectations measures possess a strong parallel relationship with durable goods purchases, as
illustrated in FIGURE 9-3.
Recognizing this relationship, the big automakers, producers
of some of the priciest durable goods, keep a watchful eye on the
consumer indices. Ford Motor Company’s sales analysis manager

Expectations

Durable goods (YOY%)

130

120
20

110
100

90
80

–10

Durable goods
Expectations

50
40

–20
1979

1982

1985

1988

1991

1994

1997

2000

2003

Source: The Conference Board; U.S. Department of Commerce,
Sources:
Bureau of Economic Analysis

Figure 9-3 Conference Board’s Expectations Index Versus Consumer Spending:
Durable Goods

Consumer Confidence and Consumer Sentiment Indices



185

Sources: University of Michigan; Board of Governors of the Federal
Reserve System

Figure 9-4 University of Michigan Index of Consumer Sentiment and Federal
Reserve’s Industrial Production Index of Auto Production
Sentiment

Autos

120

130

110

120
110

100

80
70

Sentiment
Autos

40
30

40
1978

1981

1984

1987

1990

1993

1996

1999

2002

underlined the importance given these numbers in the industry when
he told The Wall Street Journal in January 2001 that the drop in the
University of Michigan’s sentiment index in that month’s report to
93.6 from 98.4 the previous month had “played an important role” in
the company’s decision to cut back on North American production.
The automakers’ close reading of the indices is reflected in
FIGURE 9-4, which shows that for much of the 1980s and especially in
the 1990s, automobile production paralleled the sentiment index.
This relationship broke down a bit in the 2000s because of the
zero percent financing deals that automakers offered in fall 2001.
Consumers simply could not ignore such incredible deals. Still, this
relationship is steady over time and all analysts are encouraged to
keep this index in their toolkits.

Source:

HOW TO USE WHAT YOU SEE
Any indicator becomes a more useful tool when you understand
the factors that affect it. Such knowledge can aid in both interpreting and anticipating the indicator’s readings. The task is different
with measures of consumer attitude than it is with other indicators,

186



The Trader’s Guide to Key Economic Indicators

because the former deal not with something tangible—such as the
number of computers produced, the level of aircraft orders, or the
value of construction put in place—but with psychology. Wielding
consumer sentiment measures thus entails knowing how consumers’ emotions are brought into play and whether these emotions are
sufficiently strong to change spending habits and so have an impact
on the general economy.
EMPLOYMENT AND SENTIMENT
In the United States, employment status has deep psychological
resonance. You are what you do. When people meet at a cocktail
party or at a bar, their first question after exchanging names is usually, “Where do you work?” Not having a proper response can be
upsetting.
People who are uncertain about their job security or hear of relatives or neighbors being laid off become pessimistic, the depth of
pessimism depending on the breadth of job losses and the length of
unemployment. During a recession, jobless claims and widespread
layoff announcements usually garner headlines, producing anxiety
even among those that still have jobs. All this is reflected in the
consumer confidence and sentiment indices. As FIGURE 9-5 illustrates,
confidence falls steeply when nonfarm payroll growth slows and
plummets when that measure moves into negative territory. Watching the employment numbers can thus help you anticipate the next
month’s sentiment numbers or, just as important, interpret current
readings. As the confidence index’s behavior during the 1991 recovery illustrates, high unemployment can keep confidence low even
when other fundamentals are picking up.

Consumers react to many things besides employment and perceptions of general economic health. War, peace, and politics can
all shake or bolster consumer confidence. The emotional ride is
particularly turbulent today, because of the tremendous growth

Source:

NONECONOMIC INFLUENCES ON SENTIMENT

Consumer Confidence and Consumer Sentiment Indices



187

Source:

Sources: U.S. Department of Labor, Bureau of Labor Statistics;
The Conference Board

Figure 9-5 Consumer Confidence Index and Nonfarm Payroll Growth
Payroll YOY%

Confidence

160

Confidence
Payrolls

140

120

3
2

100

0
–1

–2

40
1978

–3
1981

1984

1987

1990

1993

1996

1999

2002

in news outlets on the Internet and cable and the almost instantaneous dissemination of information over these channels. The result
is often a false economic signal. Economists, analysts, and traders
must learn to differentiate between news that will stimulate or stop
spending and information that consumers will simply find uplifting
or irritating.
Consider the 1990s. The flood of investment advice, technical
and fundamental information, and live securities prices flowing
through financial websites and news channels fed stock market
fever in the general populace. Not surprisingly, consumer attitudes
closely paralleled monthly changes in stock prices. While consumers frequently misassociate stock market activity with economic
activity, differentiating between the two can be difficult for those
analysts interpreting the confidence measures. Rather than look at
the mounting excesses in the underlying economy and the irrational stock market valuations, consumers clung to the false sense of
enthusiasm generated by business journalists that were attempting
to scoop one another with respect to new investment trends. It was
this subsequent rising stock market wealth that was reflected in the
rising confidence measures of the 1990s, not the heavily burdened
economic foundations.



The Trader’s Guide to Key Economic Indicators

News, of course, affects confidence in both directions. Just as the
dot-com news festival propelled sentiment to some of the highest
levels ever recorded, so too did reports of President Bill Clinton’s
possible impeachment, the “hanging chad” presidential election of
2000, and the anthrax terrorist scares of 2001 drive the confidence
numbers down into the cellar.
When President Clinton was threatened with impeachment in
December 1998, Americans were gravely concerned about the leadership of the nation, its direction, and the possible consequences
this action would have on their personal situations. The same holds
true for the 2000 presidential election, when Al Gore won the
popular vote and George W. Bush the electoral vote. There were
suspicions of widespread scandal in Florida’s ballot counting and
doubts about the legitimacy of the electoral process. Both events
affected consumer sentiment negatively, but because neither really
had anything to do with the economy, economic activity was generally unaffected.
Impeachment and election uncertainty are mild disturbances
compared with war and terrorism. These can have devastating effects on consumers’ confidence. The Iraqi invasion of Kuwait in
August 1990, which sparked the first Gulf war in January 1991, sent
fear of casualties throughout U.S. households. The Oklahoma City
bombing in 1995, with its unexpected threat of domestic terrorism,
similarly shocked the country. These episodes rattled confidence
but didn’t constrain spending.
Even more psychologically crushing were the attacks of September 11, 2001. Americans feared for their safety, their finances,
and their jobs. The U.S. financial markets and banking system were
closed for the greater part of a week, prompting fears about their
soundness. In the face of all this, consumer confidence slumped
to a six-year low. The drop was interpreted as a recession signal,
especially given the economic context, with the nation’s airports,
borders, ports, and terminals shut down; domestic and international
commerce brought to a screeching halt; and workers idled in hundreds of industries. Miraculously, however, recession was actually a
month away from ending, not beginning.

Source:

188

Consumer Confidence and Consumer Sentiment Indices



189

The moral is that not every crisis of confidence presages an economic downturn, nor every swell of optimism an economic expansion. This is not a problem for the indices. Rather it is a challenge
to the interpreter, who must learn to distinguish between changes
in confidence that affect consumer spending and those that leave it
untouched.

Source:

TRICKS FROM

THE

TRENCHES

One way economists use the sets of consumer sentiment indices to
predict economic peaks and troughs is to chart the spread, or difference, between the Conference Board’s Expectations and Present
Situation indices. The reasoning behind this strategy is simple: If
the Expectations Index is higher than the Present Situation Index,
generating a negative spread, the implication is that people are happier with where they are now than with where they see themselves
in the near future. That attitude is bound to constrain spending and
so dampen economic growth. Conversely, a positive spread implies
belief that greater prosperity lies just around the corner, a good
sign for spending and the economy. The wider the spread in either
direction, the drearier or dreamier future conditions are expected to
be relative to the present.
As FIGURE 9-6 shows, the Expectations–Present Situation spread
generally bottoms out just before a recession begins and peaks just
after it ends. In February 2001, the spread widened considerably,
putting it into record negative territory. Apparently consumers
were spot-on with their concerns, because a recession began two
months later and lingered eight months, to November 2001. Consumers retrenched, as their incomes dwindled and stock markets
fizzled. In response, the Bush Administration instituted immediate
tax relief in the form of individual tax rebates ($300 per worker, up
to $600 per working family) and longer-term relief through reductions in the marginal tax rate.
Also in many economists’ bag of tricks is another index based on
the same survey as the University of Michigan’s sentiment index and
its associated sub-indices but constructed from a different subset of

190



The Trader’s Guide to Key Economic Indicators

Figure 9-6 The Expectations–Current Situation Spread and Recessions
100
75

Shaded areas = Recession

25
0
–25
–50
–75
–100
1978

1981

1984

1987

1990

1993

1996

1999

2002

Sources: The Conference Board; NBER

Source:

responses. The Price Expectations Index is based on two questions,
asking respondents whether they believe prices will rise, fall, or
remain stable in the next twelve months and by what percentage
they expect them to change. The end product, as FIGURE 9-7 demonstrates, is a highly accurate predictor of near-term inflation. Because
inflation influences both consumer spending and the fixed-income
market, this is a very useful and highly respected indicator.
Throughout the twenty-five years covered by the graph, growth
in the Price Expectations Index was almost identical to growth in
the Consumer Price Index, a proxy for inflation. Only in 1979–81
and 1990–91 did the actual inflation rate differ significantly from
consumers’ expectations. During both periods, the U.S. economy
was mired in recession. In the first, inflation unexpectedly topped
10 percent. At the time, no economist on Wall Street foresaw
price growth of that magnitude, so consumers’ predictive failure
shouldn’t be surprising.

Consumer Confidence and Consumer Sentiment Indices



191

Figure 9-7 Change in the Price Expectations and Consumer Price Indices

Source:

Sources: University of Michigan; U.S. Department of Labor,
Bureau of Labor Statistics

YOY%

16.0
14.0

U. of Michigan Inflation
CPI

12.0
10.0
8.0
6.0
4.0
2.0
0.0
1979

1982

1985

1988

1991

1994

1997

2000

2003

This page intentionally left blank

Advance Monthly Sales
for Retail Trade and
Food Services

he Census Bureau of the U.S. Department of Commerce releases the Advance Monthly Sales for Retail Trade and Food
Services report, known on the Street simply as the retail sales report,
about two weeks after the end of the record month, at 8:30 a.m. ET.
The report, which presents preliminary estimates of the nominal
(non-inflation-adjusted) dollar value of sales for the retail sector,
as well as the month-to-month change in that value, is available
on the Census Bureau’s website, at www.census.gov/svsd/www/
fullpub.html. These releases are avidly followed by economists and
analysts and have been known to generate serious jolts to the financial markets.
The reason for the intense interest in the retail sales report is that
retail spending provides a great deal of insight into personal consumption expenditures—the largest contributor to gross domestic
product (GDP)—both in the aggregate and with respect to several
industries and sectors. These data, moreover, are available up to two
weeks before the Bureau of Economic Analysis releases its monthly
Personal Income and Outlays report, the source for the consumption
expenditure statistics incorporated into the GDP report (see Chapter
11). The retail sales report, despite covering a narrower range of data
than the Personal Income and Outlays report, is therefore a timely
index of current and future economic health.
Not surprisingly, the report has a significant effect on the financial markets. Stocks react favorably when it shows an increase
in total retail sales, which generally equates to greater corporate
profitability. Higher sales numbers may also, however, imply higher
193

194



The Trader’s Guide to Key Economic Indicators

prices. Because inflation erodes the value of fixed-income coupon
and principal payments, a strong report depresses bond prices and
boosts bond yields, which move inversely to price. Sluggish retail
sales activity, on the other hand, portends weak profit performance
and few inflationary pressures. In this situation, stocks slump, bonds
rise, and yields fall.
If the total retail sales numbers are significant for the markets as
a whole, the figures for various components of the report can move
related sectors or even individual stocks. Noteworthy increases in
retail sales at electronics and appliances stores, for example, may lift
the shares of Best Buy and Circuit City, whereas lower postings at
clothing and accessory stores may hurt Ann Taylor and Talbots.

The Census Bureau began collecting data on retail sales in the early
1950s. The first Monthly Retail Trade Report, published in March
1952, contained estimates of sales at retail stores beginning with
January 1951.
In mid-1997, the Census Bureau significantly modified its collection procedures. Hoping to reduce the size of revisions necessary
between preliminary and final statistics, the bureau redefined the
sample used in its surveys based on the results of the 1992 Census of Retail Sales and instituted a system of polling all the 13,000
companies included each month. Previously, only the largest companies were surveyed monthly, with smaller companies divided into
three rotating panels, each of which was polled once every three
months. In mid-2001, the Census Bureau instituted another major
change, converting the data for this report, as it had for other of
its economic reports and surveys, from the old Standard Industrial
Classification (SIC) system to the North American Industry Classification System (NAICS). The goal was to facilitate comparisons
of retail numbers for the whole continent. Because of the size of
the task, however, data were converted to the new system back only
to 1992, thus reducing the scope of possible historical comparisons
and analyses.

Source:

EVOLUTION OF AN INDICATOR

Advance Monthly Sales for Retail Trade and Food Services



195

Source:

DIGGING FOR THE DATA
The Census Bureau compiles the Advance Monthly Sales for Retail Trade and Food Services report from responses to a survey
it mails out to approximately 5,000 companies about five working days before the end of the reporting month. These 5,000 are
a subsample of the 13,000 or so companies polled for the later
Monthly Retail Trade report. The principal business of all the survey participants is selling goods that are intended for personal or
household consumption.
Replies, which are weighted and benchmarked to give an accurate representation of the more than 3 million retail and food
services companies in the United States, indicate what these
establishments earned during the record month from merchandise sales and for providing services that, as the Census Bureau
puts it, are “incidental to the sale of the merchandise.” In other
words, repairs offered at auto retailers are included but not life
insurance or taxi rides. In this respect, the retail sales figures
paint a less complete picture of consumer spending than the Personal Income and Outlays report, which does incorporate service
expenditures.
Included in the retail sales receipts are excise taxes—such as
those levied on alcohol, tobacco, and gasoline—that are paid by
the manufacturer or wholesaler, passed along to the retailer, and
bundled into the price of the good. Excluded are sales taxes paid by
customers.
The bureau aggregates the survey data into total sales figures
for the month. It also breaks down some of the numbers by type of
business, using the NAICS categories and subcategories, similar to
the table shown in FIGURE 10-1, and by two subgroupings: total sales
excluding motor vehicles and parts and GAFO (an acronym for
general merchandise, apparel, furniture, and other).
All the figures are given both unadjusted and adjusted for seasonal, holiday, and trading-day variations. The report consists of
a short summary of the survey findings followed by several charts
and tables. These present retail sales, adjusted and unadjusted,

196



The Trader’s Guide to Key Economic Indicators

Figure 10-1

Major Categories of Retail and Food Services Companies, With
Percentage Each Contributes to Total Sales

Major Category Percentages of Total Retail and Food Services Sales*
Total retail and food services sales

PERCENTAGE OF TOTAL RETAIL AND FOOD SERVICES SALES

24.51%
22.46%
2.06%
5.07%
2.60%
2.47%
0.67%
8.27%
7.27%
13.32%
12.00%
0.86%
4.97%
4.21%
6.95%
4.75%
3.42%
0.27%
0.96%
0.59%
0.70%
2.23%
12.57%
5.97%
6.59%
5.59%
1.01%
2.82%
5.16%
3.25%
0.85%
9.39%
100.00%

* percentages a/o December 2002
The percentage contributions are for September 2002. Note that those italicized subcategories do not add up to 100 percent because the entire detail for them is not provided
in this abbreviated table. The bold figures of the major subcategories, however, do total
100 percent.

Source: U.S. Department of Commerce, Bureau of the Census

Motor vehicle and parts dealers
Automobile and other motor vehicle dealers
Auto parts, accessories, and tire stores
Furniture, home furnishings, electronics and appliances
Furniture and home furnishings stores
Electronics and appliance stores
Computer and software stores
Building material and garden equipment and supplies
Building material and supplies dealers
Food and beverage stores
Grocery stores
Beer, wine, and liquor stores
Health and personal care stores
Pharmacies and drug stores
Gasoline stations
Clothing and clothing accessory stores
Clothing
Men’s clothing stores
Women’s clothing stores
Shoe stores
Jewelry stores
Sporting goods, hobby, book and music stores
General merchandise stores
Department stores (excluding leased departments)
Other general merchandise stores
Warehouse clubs and superstores
All other general merchandise stores
Miscellaneous store retailers
Nonstore retailers
Electronic shopping and mail-order houses
Fuel dealers
Food services and drinking places

Source:

Advance Monthly Sales for Retail Trade and Food Services



197

both in nominal dollars and as percent changes from the previous
month and the previous year. In addition to total sales, figures for
all the categories, subcategories, and subgroupings are given—
advance figures for the record month, preliminary ones for the
previous month, and final ones for the month before that. The
report also contains revised year-earlier numbers for the record
and previous months.
In its breakdown of sales numbers according to NAICS business classification, the retail sales report differs from the Personal
Income and Outlays report, which categorizes spending activity not
by where it was done but for what it was used—whether a service, or
durable or nondurable good. To illustrate, the purchase of a refrigerator from Best Buy would be included in durable goods spending
in the income and spending report but recorded as a sale at an electronics and appliance store in the retail sales report.
Organizing the retail data according to the NAICS business
classification simplifies the survey process for the respondents. A
reporting company simply records its total sales for the period and
sends the form back; no need to break receipts down by type of
merchandise or service involved. Unfortunately, the easiest collection method doesn’t necessarily produce the most informative data.
Consider that refrigerator purchase again. It happened to be made
at an electronics and appliance store. But a consumer could just as
well buy a refrigerator at a furniture store, a building material and
supplies center, a warehouse club, a department store, or an Internet site or mail-order house. Because the report organizes sales data
by point of purchase, it reveals nothing about the level of refrigerator sales each month.
That said, analyzing sales numbers by business classification can
be very informative for retail equity analysts. Economists, moreover, can glean important insights into consumer attitudes by looking at the numbers for the retail subgroupings. In the immediate
aftermath of the 2001 recession, for instance, consumer confidence
surveys slipped, suggesting economic frailty. At the same time,
however, sales at food services and drinking places remained strong.
Seeing this, savvy economists concluded that confidence was not

198



The Trader’s Guide to Key Economic Indicators

really that low: If consumers are truly upset about the economic
future and their personal financial situations, they don’t head out to
restaurants and bars.

In just a few years, new business models can significantly alter the
retail landscape. Two such models are represented in a couple of
retail subclassifications: warehouse clubs and superstores, and electronic shopping and mail-order houses.
In the past decade, discount wholesalers have had a profound
impact on the economy as a whole, and on retail prices in particular.
The percentage that warehouse clubs and superstores contribute to
total retail sales excluding motor vehicles and parts nearly tripled
from 2.6 percent in 1992 to 8.8 percent in 2003. This is testament
to the growing popularity of stores such as Costco, BJ’s Warehouse,
and Sam’s Club.
According to Leonard Nakamura, a research adviser at the
Federal Reserve Bank of Philadelphia, conventional supermarkets
accounted for 73 percent of supermarket sales in 1980. By 1994,
this share had tumbled to a mere 28 percent. Dominance was lost
to superstores (stores with in-store butchers, bakeries, pharmacies,
and so forth) and warehouses (large discount supermarkets).
The Internet and Web shopping sites have also upended traditional business models, with significant effects on both the retail
sector and on the broad economy. Census Bureau statistics show
that e-commerce made up about 1.3 percent of total retail sales in
2002. This may seem a small contribution, but it is still significant
and it is growing.
According to the Census Bureau, the number of American households with access to the Internet has soared in recent
years—from 26 percent of all households in 1998 to approximately
44 million (42 percent of all households) in August 2000. There’s
no doubt that these numbers have climbed considerably in the four
years since the survey was conducted, given the increased affordability of personal computers. Many schools and libraries provide

Source:

SURGING SUBCATEGORIES: SUPERSTORES AND E-COMMERCE

Advance Monthly Sales for Retail Trade and Food Services



199

free access. With a few keystrokes on their computers, consumers
are now able to locate hard-to-find merchandise, search thousands
of stores, and compare prices, any time of day and any day of the
week. All retailers want to make their merchandise known to as
many potential buyers as possible, and e-commerce allows them to
do just that: They can post their entire universe of products on their
websites, together with the latest prices and availability, inform
consumers of sales and new arrivals, and maintain up-to-the second
inventory levels, all at an incredibly low cost. The result: a severalfold increase in profitability.
Because of the growing importance of this sales channel, the
Census Bureau in 1999 began issuing the quarterly Retail E-Commerce report. This enumerates all sales of goods and services that
are negotiated over an online system (whether they’re paid for
online or through traditional channels). The methodology for the
Retail E-Commerce report is basically the same as that used for
the monthly retail sales report, except that only electronically sold
merchandise is counted. As you can see in FIGURE 10-2, the total estimated value of quarterly e-commerce retail sales has rocketed from
approximately $5.3 billion in the last quarter of 1999 to about $12.5
billion in the second quarter of 2003.

Source: U.S. Department of Commerce, Bureau of the Census

Figure 10-2 Retail E-Commerce Sales
$ in billions

14
12
10
8
6
4
2
0
2000

2001

2002

2003

200



The Trader’s Guide to Key Economic Indicators

E-commerce has transformed retailers’ way of doing business. It
has also had beneficial effects on inflation. Using the Internet, consumers can search for the least expensive set of tires, shortwave radio,
or hockey stick in a matter of minutes. Companies from Bangkok
to Boston have been compelled to either reduce prices or risk losing business. This has undoubtedly played a vital role in suppressing
price rises during the later half of the 1990s and the early 2000s.

WHAT DOES IT ALL MEAN?
As was noted, the sales figures for individual business classifications
are a rich lode of information for those conducting retail equity
analysis. Economists mine the whole report for precious insights.
The Street, however, focuses on two numbers: the monthly percentage changes in total retail and food service sales and the change
in total sales excluding motor vehicles and parts.

As noted earlier, the advance retail sales total creates a stir in the
markets because of the insight it provides into consumer spending,
one of the major forces driving the U.S. economy. Traders focus
on the month-to-month percentage change in the total, and in the
total ex-auto retail sales, rather than the monthly change in the
dollar amount. The primary reason is that a $10 million monthly
increase in 1993 isn’t the same as a $10 million advance in 2003.
Analysis of monthly percent change eliminates this distortion. For
the determination of longer trends, and a further refinement of
month-to-month swings, analysts look at the year-over-year percentage change.
The picture it paints of general economic activity is a bit distorted, however. Beyond the fact that retail sales don’t include expenditures on certain services, they are also given in nominal terms.
Because the numbers are unadjusted for inflation, it is impossible to
determine if growth is the result of larger sales volumes or of price

Source:

TOTAL RETAIL AND FOOD SERVICE SALES, NOMINAL AND REAL
FIGURES

Advance Monthly Sales for Retail Trade and Food Services



201

Sources: U.S. Department of Commerce, Bureau of the Census;
Federal Reserve Bank of St. Louis

Figure 10-3 Growth in Real and Nominal Retail Sales
YOY%

12
10

Real
Nominal

8
6
4
2
0
–2
–4

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

hikes. For this reason, some economists adjust the retail sales data
for inflation by subtracting the year-over-year percentage change
in the Consumer Price Index (CPI; see Chapter 12). As shown by
FIGURE 10-3, the real growth rate in retail sales is generally lower than
the nominal one, even dropping into negative territory during the
economically difficult early 2000s.
The Consumer Price Index has a significant flaw as an inflation deflator for retail sales—namely, it includes expenditures for
services such as health care, education, transportation, and housing
(the largest component of the CPI), all of which are largely absent
from the retail sales data. Still, CPI-adjusted sales figures produce
economically sensible estimates of real growth.

Source:

TOTAL SALES EXCLUDING MOTOR VEHICLES AND PARTS
Motor vehicle and parts dealers are responsible for a very large portion of retail sales, ranging from 20 to 27 percent since 1992. The
advance retail sales total is thus heavily influenced by the numbers
generated at auto shops. Because motor vehicle sales, like those of
all expensive goods, can vary considerably from month to month,
the total retail numbers including these sales are also highly vola-

202



The Trader’s Guide to Key Economic Indicators

YOY%

11
Total
Ex-autos

9
7
5
3
1
–1

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

Source: U.S. Department of Commerce, Bureau of the Census

Figure 10-4 Growth in Total Retail Sales and in Total Sales Excluding Autos

tile. This extreme volatility makes it difficult to discern long-term
trends in the retail numbers. Accordingly, economists often focus
on total sales excluding motor vehicles. As FIGURE 10-4 illustrates, the
total “ex-autos” year-over-year growth trend is relatively smooth,
eliminating, for instance, the extreme spikes and dips manifested by
the total sales numbers in late 2001 and late 2002.

The motor vehicle business is not the only one represented in the
retail sales figures that is subject to wide price swings. Similar volatility afflicts building materials dealers, health and personal care
stores, food and beverage purveyors, and gasoline stations, whose
dollar sales figures rise and fall with the highly mutable markets
for lumber, pharmaceuticals, food, and petroleum. As an example,
consider petroleum. Should the government decide to raise the
gasoline tax or require additives in the summer months to reduce
pollution, the dollar value of service station sales will rise. The
cause of the increase is higher prices, not greater demand, but this
would be hard to price out from the numbers. Because the components of the consumer price index and its associated weights differ

Source:

GAFO

Advance Monthly Sales for Retail Trade and Food Services



203

greatly from those of the retail sales report, “deflating” the retail
sales report with the CPI is not a perfect process. The same issue
arises with sales of building materials and health and personal care
products. In addition, as with auto sales, the volatility of these
components is transferred to the sales total, making trends hard
to read.
The solution: GAFO. GAFO excludes the volatile sectors just
discussed, in addition to motor vehicle sales. Also omitted are
food services and drinking establishments, because they are considered for this purpose services rather than goods. Sales growth
in the GAFO businesses—which include furniture and home furnishings stores; electronics and appliances shops; clothing and
clothing accessories purveyors; sporting goods, hobby, book, and
music stores; general merchandise stores; and office supplies,
stationary, and gift stores—is considered the “core” retail growth
rate, similar to the core inflation rate (see Chapter 12). As shown
by FIGURE 10-5, the trend line of year-over-year growth in GAFO
sales is noticeably smoother than the jittery movements in the
total retail and food service sales and the retail sales excluding
motor vehicles and parts.

Source: U.S. Department of Commerce, Bureau of the Census

Figure 10-5 Growth in GAFO Sales
YOY%

12
GAFO
Total
Ex-auto

10
8
6
4
2
0

1993

1995

1997

1999

2001

2003

204



The Trader’s Guide to Key Economic Indicators

As with the indicators discussed in other chapters, analysts and
traders strive to get an early lock on the retail sales numbers. Forecasting these figures is difficult, largely because of a paucity of information. Several sources do exist, however, that may give insight
into what the report will show and definitely help in interpreting
the figures when they do appear.
To gain a more accurate perspective on industry-specific activity
and trends, economists and retail equity analysts often supplement
the data in the monthly retail sales report with anecdotal evidence.
As baseball legend Yogi Berra once said, “You can see a lot just by
observing.” Many analysts and traders head out to malls and shopping centers every Saturday and Sunday during the crucial Christmas holiday season to get an idea of how strong, or weak, the pace
of spending is. Some count cars or empty parking spaces. That’s
obviously a very crude measure, however. People frequent malls
for reasons other than shopping—teenagers go to hang out, for
instance, and elderly “mall walkers” go to get some exercise protected from extreme heat or cold. Many shoppers, moreover, take
mass transportation rather than private cars. Even if the car count
reflected the real number of shoppers, moreover, it wouldn’t indicate which types of stores were being patronized. Short of actually
consulting with store managers and asking what’s selling and what’s
not, the most effective mall sales-estimate method is to count the
number of bags that consumers are carrying, noting the store logos
on them. People don’t generally carry unnecessary baggage, so the
presence of a bag implies that a purchase has been made.
Analysts and economists also consult chain store announcements, looking for advance insights into consumer activity. Early
each month—often as much as two weeks before the retail sales
report—and in some cases every week, retail chain stores report on
their sales activity. In addition to indications of individual retailers’
strengths or weaknesses, perspectives on total economic activity can
be gleaned from the comments of giant retailers such as Wal-Mart.
Because of their growing popularity, wholesale discounters also pro-

Source:

HOW TO USE WHAT YOU SEE

Advance Monthly Sales for Retail Trade and Food Services



205

vide important indices of retail activity. Finally, retail trade groups
such as the National Retail Federation, the International Council
of Shopping Centers, and the International Mass Merchants Association issue informative reports on seasonal spending patterns and
trends, most of which are available on their websites.
In interpreting these data, climate should be taken into account.
Hazardous storms close stores, disrupt transportation routes, and
reduce hours worked. All of this means lost retail business, offset
somewhat by increased sales of shovels, generators, snow blowers,
and related merchandise. That said, the role of nature has been
reduced by the advent of the Internet, which enables consumers
who can’t get to the store to pursue their shopping online.

Source:

SAME-STORE SALES
Early each month, usually during the first week, the nation’s largest retailers announce their same-store sales, or “comps,” for the
preceding month. These announcements can be useful in forecasting what trends the advance retail sales report will reveal for major
business categories. For instance, it’s a fair bet that the building
materials and suppliers group will post activity in the advance
report similar to that announced earlier by Home Depot and
Lowe’s—not the same magnitude necessarily but the same vector,
or direction. In like manner, announcements made by companies
such as Ann Taylor, The Sports Authority, and CVS provide clues
to the retail sales numbers for the clothing, sporting goods, and
health and personal care categories, respectively.
Same-store sales figures do have some drawbacks as predictive
tools. For one thing, they don’t include sales by stores that have
been open less than a year. This is an important omission. New
outlets are carefully sited and heavily promoted by the parent
companies and consequently often show the strongest sales activity. Second, only year-over-year growth is reported, so the figures
provide little insight into the monthly activity reflected in the retail
sales report. To interpret same-store sales, moreover, you must
know what retail conditions prevailed in the comparison, or base,

206



The Trader’s Guide to Key Economic Indicators

month. For instance, if sales were at all-time highs in December
2002, lower numbers in December 2003 wouldn’t necessarily mean
that activity was sluggish in that month, just slower than the record
pace of a year earlier.

It should come as no surprise that many retail sectors cycle with
the season, all of which are accounted for in the seasonally adjusted
data. Apparel and accessories sales, for instance, usually surge in
August and September, when young kids are heading back to school
and older ones off to college. College freshmen, in particular, need
to stock up on new clothes, personal computers, and other electronic must-haves. Similarly, February shows strong candy, cards,
flowers, and restaurant sales. The reason, of course, is Valentine’s
Day. In fact, a very large portion of retail activity is holiday-related.
A good or bad holiday season can make or break the entire year for
a retail sector. No wonder economists and retail analysts keep an
eye on where in the calendar movable fetes like Thanksgiving and
Easter are scheduled to occur.
Timing is important. Easter, for instance, may fall either in
March or in April, boosting retail sales that month, particularly at
apparel stores. This can make for misleading comparisons. Say Easter occurs in March one year and in April the next. By comparison
with year-earlier numbers, the sales for March in the second year
will appear weak, whereas those in April will seem strong.
The timing of Thanksgiving is also important, because the following Friday is the unofficial start of the Christmas shopping season. Because Thanksgiving always falls on the fourth Thursday of
November, this season ranges between twenty-six and thirty-two
days. The assumption is that the longer the stretch, the stronger the
holiday sales numbers will be. Of course, most people must buy a particular number of presents, whether they have twenty-seven or thirty
shopping days in which to do so. But the longer time span (together
with expanded holiday hours) does mean more spending on everyday
items, such as gasoline, groceries, and bar and restaurant visits.

Source:

SEASONALITY

Advance Monthly Sales for Retail Trade and Food Services

Source:

TRICKS FROM

THE



207

TRENCHES

Holiday sales are good gauges of economic health and consumer
well-being. If the numbers are robust, people are probably feeling
secure both financially and in their jobs. A pullback, on the other
hand, may reflect consumers’ fears about finances and the labor
market. The sales numbers for Thanksgiving, Easter, and even Halloween and Valentine’s Day are all significant in this regard. But not
surprisingly, the real make-or-break shopping season is Christmas.
Because of the importance of the Christmas sales, economists,
analysts, and traders have developed different ways of estimating
these numbers from the monthly retail sales reports. Some extrapolate holiday sales from the total for November and December. Others use the total excluding motor vehicles and parts—a reasonable
approach, because few autos appear under the Christmas tree (at
least outside Mercedes Benz commercials). Still others look only
at the December figures, reasoning that in many years November
contains only a few real holiday shopping days and that most Americans every year wait until the last moment to buy their presents.
The two most popular holiday-sales proxies, however, seem to be
the combined November and December totals excluding autos and
food services, and the two months’ combined GAFO sales.
Which is the better proxy? It depends on your definition of
“holiday sales.” If you believe that spending at motor vehicles and
parts shops is not holiday related but that sales of gasoline, food and
beverages, health and personal care items, and building materials
and garden equipment are, the first approach is for you. If you think
only purchases of general merchandise qualify, the GAFO method
is more realistic.
Most economists favor the GAFO proxy. After all, how many
Santa wish lists include oak paneling, carpeting, or prescription
drugs? On the other hand, the hot housing and refinancing markets
of the late 1990s and early 2000s did create a ravenous demand for
home-improvement equipment, tools, and do-it-yourself merchandise, and many of these items may have been Christmas gifts. All
the parties given in November and December undoubtedly beef

208



The Trader’s Guide to Key Economic Indicators

YOY%

10
8
6
4
2
0

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

Source: U.S. Department of Commerce, Bureau of the Census

Figure 10-6 Holiday Retail Sales Estimates Using GAFO Approach

Source:

up food and beverage store sales as well as restaurant and bar sales.
Finally, if not for stores like Walgreens, CVS, Rite-Aid, Brooks
Pharmacy, The Body Shop, and Sephora, where would people get
their perfume and cosmetics stocking stuffers?
Whatever method they choose, economists estimate holiday sales
because of the relationship they bear to the broader economy. This
relationship is illustrated in FIGURE 10-6, which charts the year-overyear growth in holiday sales, estimated using the GAFO approach.
Note that sales increased steadily in the last half of the 1990s. In this
period, GDP was growing strongly and the stock market surging.
Sales then fell off sharply in 2000, at a time when consumers were
struggling with recession and a jobless recovery.

Personal Income and
Outlays

he monthly Personal Income and Outlays report, produced by the Bureau of Economic Analysis (BEA), contains
incredible detail on income-related measures, as well as spending
data for virtually every imaginable good and service. Commonly referred to on the Street as “income and spending,” it consists largely
of income statistics, but the underlying detail on spending may just
be the most comprehensive of all economics statistics.
What more could an economist—and those who trade on economic news—ask for in an economic indicator than timely detail on
what consumers earn and what they spend their earnings on? And
all on a monthly basis: The Personal Income and Outlays report
is released about four weeks after the record month, on the first
business day following the release of the Gross Domestic Product
(GDP) report, at 8:30 a.m. ET. It’s available on the BEA website
(www.bea.gov) within minutes after the formal release, and is extremely helpful in the analysis of macroeconomic trends.
Because spending and income data are coincident indicators, they
don’t rank high as market movers. Unexpected postings, however,
have occasionally given the financial markets a considerable jolt.
The BEA uses the spending data in the report in compiling the
consumption expenditures portion of the GDP report. Consumer
expenditures, as noted in Chapter 1, account for about 70 percent
of all economic activity in the United States. Strong spending is
a sign of an expansionary climate; slower spending signals softer
economic conditions. The income data, meanwhile, provide insight
into future spending and thus future economic activity.
209

210



The Trader’s Guide to Key Economic Indicators

While economists generally draw the inference that rising
incomes eventually result in greater spending, quantifying this
relationship is less clear-cut. Conceptually, the income-spending
relationship is somewhat leading, yet graphically that relationship
is more of a coincident association. Strong income growth usually means an expanding economy, whereas declining income may
signal weaker times ahead. Disposable income (what’s left of wages
and salaries after personal-tax and nontax payments) is particularly
important in identifying the likelihood of greater spending. Rising
income from transfer payments, such as unemployment benefit
insurance, can signal an economy that is spinning its wheels. (The
report also contains personal savings data, derived by subtracting
expenditures from income, but this is generally less revelatory than
the other information.)
The market reaction to the monthly percent change in the income and spending figures is generally subdued, unless of course
they deviate greatly from the Street expectations. Stronger than
expected increases in both the incomes or spending are a sign of
a strengthening economy, which would bode well for the general
economic climate, corporate profitability, and subsequently the
valuation of stock prices. Weaker spending and incomes generally
result in softer stock prices. The bond market usually reacts unfavorably to strong postings in income and consumption and favorably to sluggish income and spending growth. Considerably strong
releases spark unease in the stock and bond markets because of the
fear of a possible tightening in fed policy.

Because the GDP report draws on data in the monthly Personal
Income and Outlays report, you might expect the two to have similar origins. Well, they do. The birth of the income and spending
report, however, lies a little further back in time, in the 1920s. In
1921 Wesley Clair Mitchell, together with his staff at the National
Bureau of Economic Research (NBER), which he helped found
and for which he served as its first director of research, published

Source:

EVOLUTION OF AN INDICATOR

Personal Income and Outlays



211

Income in the United States: Its Amount and Distribution 1909–1919.
This two-volume work set out a framework for measuring national income and quantitatively describing its composition,
industrial sources, and distribution. The data used in the report
were collected from sources as obscure as American Telephone
and Telegraph town rent surveys and the Department of Agriculture’s annual crop estimates. It was an impressive achievement, especially because the NBER received its charter only in 1920. The
calculations, however, were relatively crude, and the reports available only on an annual basis. It was up to Simon Kuznets, a student of Mitchell’s, to bring the project to maturity. In the 1930s,
Kuznets, as described in Chapter 1, created the National Income
and Product Accounts, from which comes the GDP report, for the
Department of Commerce. The department published the first
national income statistics in 1934.
It wasn’t until the comprehensive revision of 1958 that quarterly
estimates began to appear and both income and outlays data were
formally presented. Over the next four decades there have been
several revisions, modifications, and definitional improvements that
have served to make the Personal Income and Outlays report a firstclass gauge of household activity.

Source:

DIGGING FOR THE DATA
Like the GDP report, the monthly Personal Income and Outlays
report contains data from both the income and the production sides
of the economy. Every month, these data are analyzed and displayed
in typically about eleven tables, one of which (from the May 2003
report) partially appears in FIGURE 11-1. The tables show personal
income and its disposition—that is, how it is distributed among tax
and nontax payments; personal outlays; and personal savings—in
terms of both dollar amounts and percent changes from previous
months, quarters, and years. Expenditures and disposable income
are expressed in current (nominal) and chained (real) dollars. (See
Chapter 1 for a discussion of chained, nominal, and real values.)

212



The Trader’s Guide to Key Economic Indicators

Figure 11-1

Personal Income and Outlays
($ in billions; months seasonally adjusted at annual rates)
Mar 2003
(r)

Personal income

Apr 2003
(r)

May 2003
(p)

$9,119.2
5,083.4
4,198.4
1,113.5
752.9
1,118.5
1,966.4
885.0
641.1

$9,136.6
5,083.4
4,195.7
1,110.9
748.5
1,118.7
1,966.0
887.7
642.9

$9,163.7
5,091.0
4,202.6
1,114.0
748.9
1,120.6
1,968.0
888.4
645.0

Proprietors’ income with inventory valuation
and capital consumption adjustments
Farm
Nonfarm

787.7
15.9
771.7

796.5
16.5
780.0

804.1
17.7
786.4

Rental income of persons with
capital consumption adjustment

126.2

120.5

113.6

Personal dividend income

453.7

456.4

459.1

Personal interest income

1,072.2

1,076.4

1,080.6

Transfer payments to persons
Old-age, survivors, disability,
and health insurance benefits
Government unemployment
insurance benefits
Other

1,348.9

1,354.6

1,365.1

727.1

728.8

735.7

63.6
558.1

64.8
561.1

65.8
563.6

394.0

394.1

394.8

Wage and salary disbursements
Private industries
Goods-producing industries
Manufacturing
Distributive industries
Service industries
Government
Other labor income

social insurance

Less: Personal tax and nontax
1,083.1

1,085.5

1,090.2

Equals: Disposable personal income

payments

8,036.1

8,051.1

8,073.6

Less: Personal outlays

7,769.8
7,553.9
871.7
2,223.3
4,458.9
182.5

7,779.6
7,564.2
895.2
2,194.6
4,474.4
182.0

7,790.2
7,575.2
889.1
2,191.4
4,494.7
181.6

33.4

266.4

271.5

283.4

Personal consumption expenditures
Durable goods
Nondurable goods
Services
Interest paid by persons
Personal transfer payments to the
rest of the world (net)

Equals: Personal saving
( p ) Preliminary, ( r ) Revised

Revisions include changes to series affected by the introduction of revised wage and salary
estimates for the fourth quarter of 2002.

Source: U.S. Department of Commerce, Bureau of Economic Analysis

Less: Personal contributions for

Personal Income and Outlays



213

PERSONAL INCOME
The BEA calculates personal income by adding together income
from seven major sources and then subtracting personal contributions for unemployment, disability, hospital, and old-age survivors’
insurance. The largest income source is wages and salaries, which
account for about 55 percent of the total. The BEA obtains data for
this category from Internal Revenue Service (IRS) reports. Transfer payments—government disbursements such as Social Security
payments, veteran’s benefits, and food stamps—usually constitute
about 15 percent of total income. The Social Security Administration and the Bureau of Labor Statistics supply the data for this
category. The remaining 30 percent or so of total monthly income
comes from personal interest income, which contributes 11 percent; proprietors’ income, 8 percent; other labor income (such as
group health insurance and pension and profit-sharing), 7 percent;
personal dividend income, 5 percent; and rental income, 1 percent.
(The actual percentages vary somewhat from month to month but
remain relatively close to the levels indicated here.)
By subtracting personal tax and nontax payments such as donations, fees, and fines from personal income, you arrive at disposable personal income. This figure is generally regarded as more
useful than personal income pure and simple, because it represents
the money that households have available to spend or to save.

Source:

PERSONAL CONSUMPTION EXPENDITURES
The BEA defines personal consumption expenditures as the
goods and services individuals buy, the operating expenses of nonprofit institutions serving individuals, and the value of food, fuel,
clothing, rentals, and financial services that individuals receive in
kind. The primary source for these data is the Census Bureau’s
monthly retail sales report.
The largest portion of consumer expenditures, accounting for
55 percent of the total, is for services. The U.S. economy is servicedominated. Approximately 80 percent of all workers in the United

214



The Trader’s Guide to Key Economic Indicators

Source:

States are employed in a service profession. U.S. consumers spend
incredible amounts on insurance, repair, transportation, investment
advice, and medical care. Legal services are involved in virtually
every aspect of America life, from buying a home to getting a divorce to writing a will. Entrepreneurial endeavors, such as setting
up a small business or writing a book, require accounting and legal
advice. Other service expenditures include school tuition, spending on hotel and motel accommodations, on sporting and theater
events, and on telephone and cable television service.
The next-largest category, representing 30 percent of total expenditures, is spending on nondurable goods. Nondurable goods
are products with relatively short life spans. They are divided into
four major groups: food; clothing and shoes; gasoline, fuel oil, and
other energy goods; and the catchall “other,” which encompasses
products such as perfumes, cleaning preparations, film, and greeting cards. Durable goods, which account for the remaining 15 percent of expenditures, are those intended to last a minimum of three
years. (Most do, and if they don’t, they contribute to services spending through the contracts for repairing them.) Durables include automobiles, refrigerators, washing machines, televisions, furniture,
and other big-ticket items, such as jewelry, sporting equipment, and
guns. Because durables are expensive and (because of their “durability”) are purchased infrequently, spending on these items as a
percentage of total expenditures can vary considerably from month
to month.
Personal outlays are one of the subcategories of the personal
income report. To compute personal outlays, the BEA adds net
transfers to the rest of the world and personal interest expense to
personal consumption expenditures. Net transfers include payments sent abroad by U.S. residents, such as remittances from foreign workers to their home countries. Interest expense comprises
what consumers pay on credit cards and on auto and personal loans
(but not mortgage interest, because housing is regarded as an investment).

Personal Income and Outlays



215

PERSONAL SAVINGS
The BEA includes personal savings and the personal savings rate in
Table 1 of the income and spending report. Personal savings are
defined as the difference between consumers’ disposable incomes—
the money they have available to spend—and what they actually
spend, their personal outlays. This figure expressed as a percentage
of disposable income is the personal savings rate. The May 2003
computation of the personal savings rate, for example, was:
Disposable personal income
less Personal outlays
equals Personal savings
expressed as % of disposable
personal income

$8,073.6 billion
– $7,790.2 billion
= $ 283.4 billion

$283.4 ÷ $8,073.6 x 100
equals Personal savings rate

approximately 3.5%

Source:

WHAT DOES IT ALL MEAN?
The two top attention-getters from the income and spending report are the monthly percent changes in nominal personal income
and nominal personal consumption expenditures. These month-tomonth changes receive the most attention of any number in the
report from the financial markets. Because the dollar values of these
two series are so large—in the trillions—they tend not to fluctuate
too greatly from month to month. In other words, the month-tomonth changes tend to be of the magnitude of 0.1, 0.2, or 0.3 percent. Therefore, monthly 0.2 percent announcements in personal
income or consumption can, as a result, be a bit of a wet firecracker.
The real story is in the detail, however, specifically that underlying
the income and consumption numbers. By analyzing those details
and the relationships among the personal income and expenditure
figures, economists and investors are able to identify possible turning points and developing trends in the economy.

216



The Trader’s Guide to Key Economic Indicators

The financial media tend to pay less attention to income than to
expense data. You are more likely to hear a business journalist comment on the monthly increase in services spending, for example,
than on an unexpected gain in dividend income. The reason for this
lack of interest is most likely due to the indirect effect of incomes
on the economy. Incomes don’t always have to be spent—they may
be saved. Conversely, the spending data are quite telling about what
consumers are actually doing with those incomes. Another reason
for the preference of expenditures data over incomes is that stock
market traders can directly determine what consumers are spending on. The income data don’t provide market traders with such
detailed information. That doesn’t mean personal income data are
less meaningful, however. On the contrary, they provide important
insights into the financial health of consumers, a group that, as we
have seen, has tremendous impact on all sectors of the economy.
Because some level of income is necessary for all economic activity, trends in income growth should theoretically permit inferences
about future spending patterns. Unfortunately, theory doesn’t always mesh with reality. As the chart in FIGURE 11-2 illustrates, personal
income tends to move in sync with, rather than lead, expenditures.
One reason for this synchronicity is that wages and salaries burn
a hole in most Americans’ pockets. Many people live paycheck to
paycheck, spending their earnings immediately and saving smaller
and smaller amounts. In addition, personal income includes not only
wages and salaries but also dividend and interest income and transfer payments, such as health insurance and unemployment benefits.
These crucial disbursements of unemployment benefits are generally
spent immediately on basic necessities, such as food or rent.
Because these payments are spent on necessities rather than
on durable goods and services, they have relatively little influence
on macroeconomic activity. Economists trying to judge economic
strength therefore focus on wage and salary income.
Still another explanation for the coincidence of income and
spending growth may lie in a source of income that is not included

Source:

PERSONAL INCOME

Personal Income and Outlays
Figure 11-2



217

12-Month Changes in Personal Incomes and Consumption
Expenditures

Source: U.S. Department of Commerce, Bureau of Economic
Analysis

YOY%

14
12

Spending
Income

10
8
6
4
2
0
1980

1983

1986

1989

1992

1995

1998

2001

in the monthly report: consumer credit. Consumer credit is a critical source of income in the United States, capable of altering the
amount of spending in the economy. It’s not exactly clear how wide
or narrow the gap between income and spending growth would be
without a formal credit channel. Perhaps income growth would
assume a more leading nature because consumers could not make
as many purchases with only wages and salaries. The existence of
the credit transmission mechanism permits greater access to funds,
enhancing spending without respect to wage and salary growth.
No doubt, consumer credit plays some role in the leading/lagging
quality of incomes. Because of this important economic role,
economists, retail analysts, and money managers keep a keen eye on
the section of the Federal Reserve’s Board of Governors’ monthly
release that shows the current amount of outstanding consumer
credit, including personal, auto, and education loans, as well as the
amount of revolving credit on credit cards. The Fed’s consumer
credit report doesn’t contain consumer loans secured by real estate,
such as mortgages or home equity lines of credit. All these data
and their histories are available on the Federal Reserve’s website at
www.federalreserve.gov.



The Trader’s Guide to Key Economic Indicators

Another factor that influences personal income but is not included in the monthly reports is mortgage refinancing. During the
late 1990s and into the early 2000s, mortgage rates plunged to their
lowest levels in recorded history. Americans refinanced their existing mortgages—some doing so two or three times—to reduce their
interest burden. The sky-rocketing pace of mortgage rates affected
economic activity in two respects: first, it freed up income that had
previously been earmarked for interest payments. Second, several
homeowners took advantage of the equity that they had already paid
off in their original loans, permitting them to take out even bigger
loans—and purchase larger homes—yet make the same monthly
payment as the pre-refinancing loan. The savings and capital thus
created, which amounted to hundreds of billions of dollars, were
not recorded as income. But they did fuel spending. This, too, has
caused some discrepancy between the leading/lagging characteristic
of incomes.
Consumer credit and mortgage refinancings can thus increase
consumption expenditures. Nevertheless, the amount of income
from these two sources—and, consequently, the influence they
exert on spending—is tiny compared with total personal income as
measured in the BEA reports. To predict the economically crucial
consumption expenditures figure, it is necessary to understand the
factors that influence the level and growth rate of total income—
in particular, total disposable income. The most important of these
factors are employment, tax structure, and the general economic
climate.
As noted earlier, wage and salary disbursements are the largest
sources of personal income. To earn a salary, one must generally
perform some service. (OK, some people manage to receive compensation for doing absolutely nothing, but let’s just consider the
overwhelming majority of Americans.) So job growth should be an
important determinant of income growth. The associated chart in
FIGURE 11-3 highlights some incongruities, which may be credited to
a number of factors including changes in tax policy and the comeuppance of non-wage sources of income like stock distribution and
stock options.

Source:

218

Personal Income and Outlays

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; U.S. Department of Labor, Bureau of Labor Statistics

Figure 11-3



219

Disposable Personal Incomes and Employment

Disposable income, YOY%

Payrolls, YOY%

14
Disposable income
Payrolls

10
2

8
6

4
–2

–4

0
1980

1983

1986

1989

1992

1995

1998

2001

Growth in payrolls encourages spending not only by increasing
disposable income but also by lifting consumers’ spirits. Another
important influence on consumer expenditures is the taxation level.
When marginal tax rates are low, disposable personal incomes rise.
With more of their earned income left over, people have a greater
propensity to spend; and because consumer spending accounts for
almost 70 percent of all economic activity, and the United States is
traditionally a nation of spenders, economic growth will expand.
CONSUMER SPENDING
The connection between consumer expenditures and economic
growth has already been well established. But not all spending is
equally revealing of economic trends. Spending on nondurable
goods such as food and home-heating fuel tends to be fairly constant, remaining positive even in trying economic times. In contrast, spending on durable goods, which are relatively expensive
and long-lived, requires good economic conditions to flourish. In
less flush times, consumers aren’t going to head out to buy stereos,
furniture, or new china. Therefore, of all the subcomponents in
the Personal Incomes and Outlays report, durable goods spending



The Trader’s Guide to Key Economic Indicators

Figure 11-4

Spending on Consumer Durables and Recessions

YOY%

25
20
15
10
5
0
–5
–10
Shaded areas = Recession

–15
–20
1980

1990

2000

might be considered the most effective in calling turning points in
the economy.
The chart in FIGURE 11-4 shows that three of the last four recessions identified by the National Bureau of Economic Research—
1981–82, 1990–91, and 2001—were accompanied by simultaneous
declines in the growth rate in consumer spending on durable goods.
The 2001 recession broke this pattern, as automakers and certain
other retail giants kept consumers buying their products by offering
zero percent financing, hefty discounts, and other incentives.
PERSONAL SAVINGS RATE
Americans spend. The nation as a whole just can’t seem to “save for
a rainy day,” despite the warnings of previous generations. As the
chart in FIGURE 11-5 illustrates, this propensity has worsened in the
past five years.
Americans’ declining savings rate can be explained in part by demographics. Baby boomers—people born between 1946 and 1964—
are the first generation that stands to inherit a significant amount
of wealth. The baby boomers’ grandparents, born in the late nineteenth century, lost most of their accumulated assets during the

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; NBER

220

Personal Income and Outlays
Figure 11-5



221

Personal Savings as a Percentage of Disposable Personal Income

Source: U.S. Department of Commerce, Bureau of Economic
Analysis

Percent

16
14

Shaded areas = Recession

12
10
8
6
4
2
0
1959

1963

1967

1971 1975

1979

1983

1987

1991

1995

1999

2003

Great Depression. Even when the economy picked up, they retained memories of the hardscrabble times and raised their children
in households of frugality, saving everything they could. As the
Depression survivors began to die off in the mid-1990s, their children, the boomers, inherited the homes, investments, jewelry, cars,
autographed baseballs, and other assets they had amassed. This was
occurring during some of the best economic conditions in about
five decades—rock-bottom unemployment of around 3.9 percent,
virtually nonexistent inflation, and a skyrocketing stock market. No
wonder the boomers didn’t find savings crucial.
Economists generally worry when the personal savings rate
slows. This usually signals that consumers are dipping into their
savings to make ends meet. Depleted savings are most disturbing
during soft economic times, such as those of the early 2000s when
unemployment was on the rise. The hardships posed by the loss
of a job—and of its associated income—are exacerbated when the
worker is already overextended. That is why it is important to keep
an eye on the pace of consumer credit. If it is rising during a weak
economy, a very dangerous situation may be developing, which
could result in a double-dip recession as the consumer spending
that initially brought the economy out of the recession disappears.

222



The Trader’s Guide to Key Economic Indicators

The data in the monthly Personal Income and Outlays reports supply the raw material for the analyses in the quarterly GDP report.
To come up with the quarterly figures, the BEA simply averages
the monthly numbers recorded for each data category. Because the
GDP report isn’t published until about a month after the end of
the record quarter, analysts, money managers, and economists (as
well as some analysts and money managers) keep a running tab of
the monthly consumer spending figures to approximate the value of
this important contributor to economic activity.
Because personal incomes and expenditures are so critical to the
overall pace of economic activity, economists attempt to predict what
the report will show before it is released. One way they do this is to
go straight to the data sources. The largest source for monthly expenditures information is the Advance Monthly Sales for Retail Trade
and Food Services report. Spending on some retail goods, though, is
more significant than that on others. Durable goods, as noted above,
are more economically sensitive than nondurables. Among durable
goods, some are better as predictors of macroeconomic conditions.
Wall Streeters, for instance, watch the “RV Indicator.” Recreational
vehicles, or RVs, are usually purchased out of discretionary or unessential income. When expenditures on these vehicles slump, it’s a
good bet that the economy will soon slow. Conversely, when sales
begin to accelerate, the economy is expected to expand.
RVs may still be regarded as luxuries by many Americans. Cars,
in general, are not. Every teenager dreams of owning a sports car,
and virtually every adult has fond memories of his or her first automobile. (For the record, my first was my grandfather’s 1965 forest
green Chevy Impala with a 283—the best engine GM ever made.)
It’s not surprising, then, that purchases of motor vehicles and parts
constitute one of the largest components of consumer spending. As
the chart in FIGURE 11-6 shows, motor vehicle spending accounts for
a great deal of durable goods spending. In fact, at least since 1997,
this component never accounted for anything less than 40 percent
of total durable goods spending.

Source:

HOW TO USE WHAT YOU SEE

Personal Income and Outlays

Source: U.S. Department of Commerce, Bureau of Economic Analysis

Figure 11-6



223

Spending on Motor Vehicles as a Percentage of Durable Goods
Expenditures

Percent

50
48
46
44
42
40
38
1997

1998

1999

2000

2001

2002

2003

Knowing this, economists like to look at the monthly pace of
auto sales. Several indicators of motor vehicle demand exist. The
most relevant are contained in the advance retail sales report, because they are a measure of spending, and the monthly sales reports
of individual car companies (available on their websites). By assessing the pace of automotive sales, investors are able to get a good
idea of activity in durable goods as a whole and thus of consumer
spending in general and of overall economic conditions.

Source:

TRICKS FROM

THE

TRENCHES

This chapter’s tricks involve factors that influence levels of spending and that economists and investors can use to foretell what
those levels will be. What determines how much consumers
spend? What doesn’t? Some people spend because it’s Saturday
night. Gloomy days often prompt spirit-boosting trips to malls.
But rain crimps sales at the beaches and shores, and blizzards keep
everyone inside, preventing even determined shoppers from getting to a store and spending. Meanwhile, closed stores prevent
hourly workers from earning income. Weather, time of week, time

224



The Trader’s Guide to Key Economic Indicators

of year, and many other factors affect spending. For economists,
however, three influences are predominant: wealth, prices, and
employment.

Many economists identify a wealth effect: As individuals’ wealth
rises, the reasoning goes, so does the level of spending. This seems
logical. But how do you measure individuals’ wealth? One way is
to look at the Federal Reserve’s quarterly Flow of Funds report
(available on the Fed’s website, www.federalreserve.gov/releases/
Z1/), which tracks financial and physical asset flows in the U.S.
economy. The Fed’s report contains detailed information regarding outstanding levels of household ownership of several types of
assets such as U.S. government securities, mutual fund shares, and
corporate equities. Unfortunately, these data are provided solely
on a quarterly basis and delayed for about three months, rendering them useless for analyzing monthly trends. First quarter data,
for example, are not released until the middle of June—and they
undergo wide-ranging revisions.
Given these drawbacks, economists have developed their own,
more timely way of measuring the consumer confidence implied
by the wealth effect. They divide the dollar value of the Wilshire
5000—a stock market index composed of the equities of all the
companies headquartered in the United States—as of the end of
the quarter by the level of disposable personal income. When
the ratio of the Wilshire 5000 to disposal personal income—our
proxy for wealth—rises, the stock market wealth is on the rise.
This increase makes consumers feel wealthier and so willing to
spend more. Through econometric analysis, some economists
have determined that for every $1.00 increase in the level of
stock market wealth, consumer spending increases by $0.03 to
$0.07 per year.
Some economists argue against the wealth effect, pointing out
that only about 50 percent of American households have a link
(direct or indirect) to the stock market. In other words, our proxy

Source:

The Wealth Effect

Personal Income and Outlays

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; Bloomberg LP

Figure 11-7



225

Wealth Ratio Versus Consumer Spending

Wealth/Dis Income Ratio

Spending YOY%

2.5

2.0

6
1.5
4
1.0

Wealth ratio
Spending

0.5
1991

1993

1995

1997

1999

2001

2003

Source:

would only explain the gain or loss in wealth for about 50 percent
of all households. This might explain the disconnect between the
two series prior to 1997 and after 2002 in FIGURE 11-7. Because our
wealth-effect ratio does parallel trends in year-over-year growth
rate of consumer spending for the better portion of 1997 through
2001, when stock prices sky-rocketed, and then tumbled, it might
be the case that the wealth effect only works during periods of
high stock market participation. It seems reasonable, though, that
even people who don’t invest in equities feed off of the positive
atmosphere of a rising market and pick up the pace of spending.
Newspapers, evening news programs, and radio stations broadcast
the daily stock market gains and the reasons for them—usually
upbeat signals about the economy given by the indicators described in this book. Higher stock market valuations result from
higher earnings expectations, which in turn imply increased business spending and the increased likelihood of hiring. When job
creation increases, consumers are more upbeat, and incomes and
spending rise.

226



The Trader’s Guide to Key Economic Indicators

Prices
Another obvious relationship exists between spending and prices.
This is summed up in one of the first laws that every economics
student learns: the law of demand. In rough terms, this states that
the higher the price of a good is, the lower the demand for it will be,
and conversely, the lower the price, the greater the demand. Consider
gasoline. When the price of a gallon rises to $2.00 or more, consumers demand less of it. Commuters who usually drive to work may carpool or use alternative means of transportation, like the bus, train, or
ferry. Families might postpone vacation plans until fuel prices recede
to more affordable levels. FIGURE 11-8 illustrates the inverse relationship
between price and demand. The time span covered includes three
periods of high inflation in gasoline prices: 1990, 2000, and 2003.
During all three of these episodes, the growth rate of consumer spending on the fuel declined. Readers can obtain these data—in amazing
detail—from the BEA’s website, http://www.bea.gov/bea/dn1.htm.
Investors should get into the habit of knowing the underlying
price trends for several major spending categories such as healthcare, medicine, apparel, food, housing, tuition, and transportation.
Growth in Real Consumer Spending on Gasoline Versus Growth in
Consumer Price of Gasoline

Price, YOY%

Spending, YOY%

50
40

30
20

10
0

0
–10

–4

Prices
Spending

–20

–8

–30
1988

1991

1994

1997

2000

2003

Source: U.S. Department of Commerce, Bureau of Economic
Analysis

Figure 11-8

Personal Income and Outlays



227

Obviously the higher the price of these goods, in many cases necessities, the less money there will be available for spending on other
things. The detail regarding all of these consumer prices of goods
and services—and thousands more—are available at the Bureau of
Labor Statistics website (www.bls.gov). The price measures will be
discussed in Chapter 12.
Employment
Stock market losses and rising prices may dampen consumers’ enthusiasm, but they tend to slow spending growth rather than stop it
altogether. There’s one influence that will cause consumers to virtually cease all nonessential spending. That’s the loss of a job.
FIGURE 11-9 illustrates the incredibly tight relationship between
the growth rate of nonfarm jobs and the rate of spending. Nothing is as economically depressing as the loss of employment or the
fear of losing a job. It isn’t just that unemployed people don’t have
earned income to spend. They may also despair of finding a job
any time in the near future. This is particularly true during weak
economic times, when several hundred people might be applying
for the same advertised position. When consumer confidence mea-

Sources: U.S. Department of Commerce, Bureau of Economic
Analysis; U.S. Department of Labor, Bureau of Labor Statistics

Figure 11-9

Growth in Real Consumer Spending Versus Payroll Growth

Consumer Spending, YOY%

Payrolls, YOY%

9
Disposable Income
Payrolls

3
1

–1

–3

–3
1980

1983

1986

1989

1992

1995

1998

2001

2002

228



The Trader’s Guide to Key Economic Indicators

Source:

sures tumble before and during recessions, they are capturing this
discouragement. The feeling can paralyze spending.
Figure 11-9 shows a bit of disconnect in 2001 to 2002 when
payroll growth was contracting, yet spending registered some reasonable solid gains. This is a rarity; the 2001 recession was the first
in post–World War II history during which consumer spending did
not decline. This anomaly can be credited to the sound implementation of fiscal and monetary policy. Expeditious tax cuts by the Bush
administration and the pre-emptive lowering of interest rates by the
Greenspan Fed fueled incomes and kept spending on the rise.
Knowing that the level of employment is critical in the determination of spending, investors should always keep an eye on the pace
of job creation. The stronger the rate of growth in employment,
the stronger the pace of spending will be. Employment data and its
detail can also be found at the Bureau of Labor Statistics website.
The employment situation is discussed in detail in Chapter 3.

Consumer and Producer
Price Indices

hen the prices of goods and services rise, it is called inflation. A certain level of inflation in the economy is normal,
even healthy. Accelerating inflation, however, can cause severe
problems, sometimes sparking recession. No wonder the financial
markets keep a close eye on price measures and their growth rates.
For this purpose, many traders and economists, including those at
the Federal Reserve, favor the implicit price deflators contained in
the Gross Domestic Product (GDP) report (see Chapter 1). That
report appears only quarterly, however. For more timely—and
detailed—inflation indicators, most market participants turn to
the reports on the Consumer Price Index (CPI) and the Producer
Price Index (PPI).
The Bureau of Labor Statistics (BLS) calculates, maintains, and
reports on the CPI and the PPI. (The bureau also produces a third
set of indices in the international import and export price report,
but the market doesn’t react to these, so they will not be discussed
in this book.) The CPI and PPI reports are released around the
middle of the month following the record month, the PPI usually
at least one business day before the CPI. The releases—which hit
the newswires at 8:30 a.m. ET and are available on the BLS website,
www.bls.gov—often create quite a stir in the financial world, especially the fixed-income market.
The CPI tracks the change in price, at the consumer level, of a
weighted basket of a few hundred goods and services. The composition of this basket reflects households’ typical monthly purchases,
as revealed in the Consumer Expenditure Survey (CEX), which the
229



The Trader’s Guide to Key Economic Indicators

Census Bureau conducts for the Bureau of Labor Statistics. The
weight given each item is determined by its percentage of total
household expenditures. The index reading represents how much the
basket has increased in value since 1984, the base year. A reading of
130, for instance, denotes that the current average cost of the goods
and services is 30 percent greater than it was twenty years ago.
The CPI has two basic versions: the CPI-U, which reflects the
buying habits of all urban consumers, and the older CPI-W, which
relates only to urban households that include a wage earner or clerical worker. The two versions employ data from the same survey and
are constructed using the same methodology. They differ only in
the weight given certain basket components. The CPI-W is used
by the private sector in contract price-escalation clauses and by the
government in computing cost-of-living adjustments, or COLAs,
for Social Security. The Street and the media focus on the CPI-U,
because it represents roughly 87 percent of the noninstitutionalized population, against the CPI-W’s mere 32 percent. National
and local governments, businesses, and organizations employ the
CPI-U in forming and implementing policies. Economists use it to
adjust nominal-based indicators, such as retail sales, for inflation.
All of the discussions of CPI in this book refer to the CPI-U.
The PPI, also known as the wholesale price index, tracks
changes in the selling prices of some 3,450 items, at various stages
of manufacture, that are received by the producers of those items.
Price figures are collected monthly and, for the most part, are those
recorded on the Tuesday of the week containing the thirteenth day
of the record month. Components are weighted according to their
contribution to the GDP. As for the CPI, readings represent price
changes from the base year 1984.
The Producer Price Index incorporates data about prices before
the retail level is determined. It covers items not in the CPI, such
as raw materials and intermediate goods. Economists looking at the
PPI data can thus see how far in the production process inflation
pressures have traveled and how close they are to emerging in the
retail or consumer sector. They can also get a feel for whether any
rise in business costs is driven by demand or by supply.

Source:

230

Consumer and Producer Price Indices



231

Because of these characteristics, and because of its earlier release
date, the PPI is used by some analysts to predict CPI readings. This
can be misleading, however. The two indices are very different,
both in the way they are constructed and in the items they include.
The PPI, for instance, doesn’t contain any information on prices for
services, the largest part of the U.S. economy. On the other hand,
it does incorporate information about the prices of raw materials,
which are extremely sensitive to weather conditions. As a result, the
PPI’s monthly readings are extremely volatile and can be quite different from those of the CPI, although the two indices do show a
high degree of correlation over the longer term.

EVOLUTION OF AN INDICATOR
The origins of the PPI and CPI, unlike their release dates, are
widely separated in time. They were created not only at different
times but to serve different purposes, and so each index evolved
quite differently.

Source:

PRODUCER PRICE INDEX
In the late 19th century the United States Senate authorized the
Bureau of Labor Statistics to start collecting and reporting wholesale prices, so that it could assess the economic effects of tariff laws.
The first Wholesale Price Index (WPI), the index’s official name for
nearly eight decades, was published in 1902. It was an unweighted
index of about 250 commodity prices, covering the period from
1890 through 1901. A weighting scheme was adopted in 1914; it
was later refined in 1952 and 1967.
The index’s original purpose was to reveal price activity at
the earliest stage of production. It was believed that this was best
accomplished by compiling the prices that domestic producers
or importers of the goods and commodities received for them.
The original method, unfortunately, involved skewed sampling
techniques where responses from large companies dominated the
sample. Over time, this resulted in a misrepresentation of goods

232



The Trader’s Guide to Key Economic Indicators

prices. This initial method had an overemphasis on prices of goods
produced by the larger firms and under-representation of goods
produced by smaller firms. By the 1940s, moreover, the original
weighting schema was outdated, giving too little importance to
certain mining and manufactured products that by then accounted
for about half (in dollar value) of all goods used in the production
process. These wrinkles were eventually ironed out through the
recategorization of data and constant revision to sampling methods and weighing systems.
In 1978, the BLS overhauled the index again, emphasizing the
categorization of prices by stage of processing, rather than by commodity or industry, and stressing finished goods over those at other
stages. The bureau also changed the name of the index at this time,
from “wholesale” to “producer” price index.

Compared with the PPI, the CPI is a newcomer. Prices on the consumer level were first collected during World War I, to estimate
cost-of-living adjustments to wages. From 1917 through 1919, the
BLS collected data from ninety-two industrialized population centers and analyzed spending patterns to create weighted indices of
consumer expenditures.
Profound changes in consumer buying habits, particularly during the Roaring Twenties and the Great Depression in the 1930s,
led to a comprehensive overhaul of the weights and composition of
the indices in 1940. World War II–related rationing and shortages
necessitated similar revisions. In 1953, the CPI underwent its greatest makeover up to that time, including improvements in methodology and collection procedures, as well as new sources of data and a
more representative list of items. This process of refinement and
restructuring continued through the late 1990s.
In December 1996, a commission created by the Senate Finance
committee to study the CPI and its framework released a highly
publicized report concluding that the CPI, as currently constituted,
overstated the true level of inflation by about 1.1 percentage points

Source:

CONSUMER PRICE INDEX

Consumer and Producer Price Indices



233

and suggesting several remedies. The Boskin Report (so-called after
the commission chair, Michael Boskin) set the financial world abuzz:
Because cost of living adjustments in employee compensation and
pensions are linked to the growth rate of the CPI, a reduction in the
index’s growth rate would lower disbursements to retirees, Social
Security recipients, and civil service workers and so restrict their
spending power. Due to the political, economical, and social nature
of these findings, there hasn’t been any serious adjustment to the
consumer price measures.

DIGGING FOR THE DATA
As do the indices’ origins, the data sources of the CPI and PPI and
the methodologies used in compiling them differ considerably.

Source:

CONSUMER PRICE INDEX DATA SOURCES
The CPI represents prices on the retail, or demand side, of the
economy. To gather the data used to compose the index, field
economists from the BLS visit supermarkets, department stores,
gasoline filling stations, hospitals, and other establishments in
eighty-seven urban areas all around the nation, recording prices of
food, fuel, beverages, apparel, health care, and other goods and services. Additional prices are obtained via mail survey.
The prices gathered are organized into eight expenditure categories: housing, transportation, food and beverages, recreation,
medical care, education and communication, apparel, and “other,”
which includes such items as personal-care products and tobacco.
The goods and services included in the survey are determined by
the results of the Consumer Expenditure Survey, as are the weights
given to their categories in computing the index. These weights,
which reflect the portion of their incomes that consumers spend on
the items in the category, range from 40.8 percent for housing and
17.3 percent for transportation, through 5.7 percent for education
and communication to 4.3 percent for “other” and 4.2 percent for
apparel (see FIGURE 12-1).

The Trader’s Guide to Key Economic Indicators

Figure 12-1

Composition of the CPI-U

Education & Communication 6%

Other 4%

Apparel 4%

Housing 41%
Food & Beverages
16%

Recreation 6%

Investment 6%
Consumption 17%

The index’s basket of goods and services does not, of course,
capture every individual’s or every group’s consumption pattern.
The elderly, for instance, probably spend more of their monthly
allowances on medical costs, whereas the younger generation lays
out more on tuition and apparel. The categories and their weights,
however, present a fairly accurate picture of Americans’ average
monthly spending habits.
Calculating the average prices for items and categories isn’t as
simple as it may sound. For goods like toothpaste, alcoholic beverages, tires, or a ticket to a sporting event, the process is straightforward. Services are a different matter. Housing, the largest
component of the CPI, is particularly complex. The index measures
the cost of using services, not of obtaining assets, such as condos
or Cape Cods. The BLS accordingly recognizes two categories of
housing costs: residential rent and owners’ equivalent of residential
rent. The BLS defines the latter as “the cost of renting housing
services equivalent to those services provided by owner-occupied
housing.” This definition removes the investment component of
ownership. Price information for the housing category is obtained
through interviews with landlords, tenants, and owner-occupants.
From the monthly pricing data, the BLS calculates values for

Source: U.S. Department of Labor, Bureau of Labor Statistics



Source:

234

Consumer and Producer Price Indices



235

the headliner all-items index (covering the entire basket) and for
various subindices, including one for each of the eight expenditure
categories and several “special” subindices. Tables throughout the
report present values for the various indices and subindices, both
adjusted and unadjusted for seasonal variations. What draw the
most attention are the percentage changes—from month to month,
year over year, and over a three-month period—which represent
inflation rates for the relevant periods and categories of items.
The top three special indices are for energy, food, and all goods
and services except food and energy. The last of these, referred to as
core CPI, is particularly influential.
Energy and food prices are extremely volatile. Tensions in the
Middle East, unusually cold or hot weather, changes in production
schedules, particularly among OPEC countries, are just a few factors that can send oil and gas prices soaring or plummeting. Similarly, food prices can move violently on news of droughts, storms,
or late frosts that destroy crops. Removing these components and
their erratic movements makes it easier to discern longer-term
inflationary trends. The result is termed “core inflation.” (Economists at the Federal Reserve Bank of Cleveland have gone one
step further in reducing CPI “noise,” lopping off those components showing the biggest gains or declines in a given month; the
so-called Median Index—also referred to as the Cleveland Fed
Index, the Median CPI, or the Cleveland Fed’s Median CPI—is
available on the Cleveland Fed’s website, www.clevelandfed.org,
and has become a Wall Street favorite.)

Source:

PRODUCER PRICE INDEX DATA SOURCES
The PPI tracks price trends from a seller’s, or supply-side, perspective. Every month, the BLS collects prices for about 100,000 goods
at various stages of production from voluntary surveys completed
by some 30,000 businesses. Using these prices, it compiles around
10,000 indices, which fall into three major categories: commodity indices, which organize data according to end use or material
composition (farm products, textiles and apparel, transportation



The Trader’s Guide to Key Economic Indicators

equipment, for instance); industry indices, which are organized according to Standard Industrial Classification (SIC) and weight their
components by “net output,” or the value of shipments outside the
industry (railroads, the U.S. Postal Service, tour operators); and
stage-of-processing indices, which are grouped by the amount of
processing of the good and the purchaser’s class. The last category
is the focus of the PPI report and the one most often cited in the
business press and on trading floors.
The stage-of-processing system classifies items as crude materials for further processing, intermediate components, or finished
goods. Crude materials are commodities that have not been refined
or processed, such as raw cotton, hides and skins, and copper and
aluminum base scrap. Intermediate materials have undergone some
processing but have not completed the fabrication process. Popular
goods in this category include industrial textile products, leather,
glass containers, and synthetic rubber. Finished goods are ready to
be sold to the final user (consumers and businesses) without further
refinement.
The stage of processing category is divided into two major
groups, with weights in line with respect to the contributions to
total economic activity in the national income and product accounts:
consumer goods, which are weighted in line with the composition of
personal consumer expenditures (accounting for approximately
three-quarters of all finished goods) and are themselves divided into
food and nonfood items; and capital equipment, a representation of
the value of business purchases. (FIGURE 12-2 shows the finished goods
components and subcomponents.) Finished goods include apparel,
roasted coffee, textile machinery, commercial furniture, and railroad equipment. Because these items are the closest to the retail
level, the Street focuses on their price indices. When the media,
economists, or analysts refer to producer prices, wholesale prices,
or PPI, they are referring to the inflation rate or percentage change
in the finished goods indices.
The monthly PPI news release contains only the key aggregate
indices of about two hundred or so seasonally adjusted and unadjusted indices. BLS publishes more than five hundred industry price

Source:

236

Consumer and Producer Price Indices
Figure 12-2

237

Finished Goods Components

Crude Foods
Processed Foods
Source: U.S. Department of Labor, Bureau of Labor Statistics



1.333 %*
19.339

Finished Consumer Foods

20.672

Nondurable Goods less Foods

36.364

Durable Goods

16.303

Nonfoods

52.667

Finished Consumer Goods

73.339

Manufacturing Industries

7.839

Nonmanufacturing Industries

18.822

Capital Equipment

26.661

FINISHED GOODS

100.000

*All percentages as of July 2003.

indices, ten thousand specific product line and product category
subindices, and 3,200 commodity price indices. The complete series
of indices, as well as their histories, are on the BLS website and in
the BLS’s monthly PPI Detailed Report, available by subscription.
CALCULATING THE INFLATION RATE
You can use the following formula to determine the rate of inflation
between two periods implied by any of the index values in the CPI
or PPI reports:

Source:

RInf = 100 x (ICP – IPP) ÷ IPP, where RInf is the rate of inflation, ICP is the
current index value, and IPP is the previous index value

To illustrate, say you wanted to figure out the twelve-month
inflation rate for copper base as of July 2003. For that month, the
unadjusted copper base index value was 123.1, versus 114.2 in July
2002. Plugging those values into the formula, you get

238



The Trader’s Guide to Key Economic Indicators

RInf = 100 x (123.1 – 114.2) ÷ 114.2 = 890 ÷ 114.2 = 7.79 percent,
or approximately 7.8 percent inflation, year over year.

WHAT DOES IT ALL MEAN?
The inflation rate can tell us a great deal about economic conditions. When the economy is strengthening, companies experience
increased demand for their products and so can charge higher
prices for them. As a result, revenues increase, lifting profits and
permitting companies to boost capital investment and create new
jobs. At the same time, however, the higher prices squeeze consumers, who may have to choose where to allocate limited funds: The
more they pay for one good or service, the less they have for others. It’s a delicate balance. Too little inflation, and corporate profits
tumble, curtailing capital spending and causing unemployment; too
much, and consumers can’t afford to buy. The results are the same,
only the course is different.
That said, the PPI and CPI indices aren’t generally considered
leading indicators. Changes in the general price level aren’t as predictive of business cycle turning points as are many of the indicators
discussed in previous chapters. They do tell a great deal about the
microeconomic conditions of individual commodities or industries,
however. Just don’t read too much into a single month’s activity.
Price indices, even those excluding energy and food, can be affected
by any number of influences. Legislation and taxes, for example,
can push up prices on items like liquor and tobacco quite dramatically from one month to the next.

Prices can display three trends: inflation, or a sustained increase in
prices; disinflation, a slowing of the rate of increase; and deflation,
a sustained decrease. FIGURE 12-3 illustrates all three phenomena.
First, inflation. The line graphing year-over-year changes in consumer prices remains above zero for the entire chart. That means
the inflation rate was positive for the whole period. In other words,

Source:

PRICE TRENDS

Consumer and Producer Price Indices
Figure 12-3



239

Consumer and Producer Price Indices

Source:

Source: U.S. Department of Labor, Bureau of Labor Statistics

YOY%

20
PPI

CPI

10
5
0
–5
1970

1980

1990

2000

1970 through 2003 saw rising prices for consumer goods. The
increases have not been uniform, however. The price-growth line
falls sharply from 1972 to 1974, 1975 to 1977, 1980 to mid-1983,
and then again from 1991 to 1992. Those drops indicate slower
rates of price growth—that is, disinflation—in consumer goods. To
see a graphic representation of deflation, you have to turn to the
more volatile producer prices (the only serious bout of consumer
price deflation that ever occurred in the United States was during
the Great Depression). As the PPI-growth graph shows, since 1975
the twelve-month inflation rate in producer prices has fallen into
negative—deflationary—territory several times during 1986–87,
1992, 1994, 1997–99, and 2001–03.
Deflation is as damaging to economic health as high inflation.
When prices are falling, consumers postpone purchases in anticipation of even lower prices in the future. Without the engine of
consumer expenditures (the largest component of GDP), economic
growth slows and may even contract if the situation continues.
Deflation also hurts corporate profits, causing companies to cut
production and reduce staff.
In recent years, economists have fretted about the possibility
of deflation in the United States. Growing globalization has sent

240



The Trader’s Guide to Key Economic Indicators

production facilities to low-wage nations such as China and India,
which send incredibly low-priced toys, textiles, computer parts, and
foods back to the United States. Prices of nonimported services,
such as tuition, medical care, rent, and electricity, however, have
been rising. So, although certain industries have experienced deflation, the economy as a whole has not. In a true deflationary period,
all prices decline, on both the consumer and the producer level.
The only hint of deflation has been in the Producer Price Index.
One reason for this is that no services are included in the PPI.
Moreover, the core rate of producer price growth—which excludes
the volatile food and energy components—has fallen into negative
territory only once, and that barely. (Core PPI deflation exists only
during prolonged periods of manufacturing weakness.)

To bondholders, inflation is public enemy number one. Bond buyers are actually lending the security’s purchase price to the issuer; in
return for their loans, they get coupon payments at regular intervals
for the life of the bond (unless they sell or it is called). Inflation rates
erode the purchasing power of future payments. Say a 10-year bond
pays a 6 percent coupon. If inflation rises to 4.5 percent, the investors’ real (inflation-adjusted) rate of return is only 1.5 percent—not
very good over a ten year period. No wonder that, at the slightest
whiff of inflation, investors sell their fixed-income securities, sending prices down and yields (which are inversely related to price) up.
Equity investors generally react very little to the inflation reports.
Even stockholders, however, can get exercised when the monthly
postings differ greatly from expectations or suggest an inflation
rate that could impede consumer spending and disrupt economic
growth. A series of high inflation numbers—say, three consecutive
monthly increases of 0.7 percent in the CPI or PPI—will have both
bond and stock investors anticipating a possible tightening by the
Federal Reserve.
To cool the economy down and dampen inflationary pressures,
the Fed may raise its target for the Fed funds rate (the rate banks

Source:

PRICE INDICES AND THE MARKETS

Consumer and Producer Price Indices



241

charge each other for overnight loans used to meet reserve requirements; see Chapter 1). Longer-maturity interest rates usually follow suit. High rates discourage consumers from buying assets, such
as houses and motor vehicles, whose purchases are financed with
loans. Companies may also put off construction and other projects
that would necessitate forays into the debt markets. If rates rise to
truly restrictive levels, they may be forced to eliminate workers.
PRICE INDICES AND THE BUSINESS CYCLE
As with other price measures, the “core” rate of inflation, computed
by excluding the volatile food and energy components, is often
used for tracing inflation trends. The BLS calculates core rate for
the PPI: The crude nonfood materials less energy index. The core
crude PPI, as it is known, is more obscure than the other core indices but very useful in tracking the business cycle.
Raw-material prices historically indicate turning points in the
business cycle. Early in a recovery, companies prepare for the anticipated pickup in demand for their products by speeding up their
own purchases of commodities they need to begin manufacturing.
Conversely, at the first sign of a downturn, they protect against
Figure 12-4

Core Crude PPI Versus S&P 500 Operating EPS

Sources: U.S. Department of Labor, Bureau of Labor Statistics;
S&P Corp.; Federal Reserve Bank of St. Louis

YOY%

40
30
20
10
0
–10
–20

S&P 500 EPS
Core Crude PPI

–30
–40
1989

1991

1993

1995

1997

1999

2001

2003

242



The Trader’s Guide to Key Economic Indicators

slowing demand by reducing their consumption and purchases of
crude materials, depressing these items’ prices. This relationship is
manifested in the correlation, shown in FIGURE 12-4, between yearover-year changes in the core crude PPI and the twelve-month
growth rate of S&P 500 operating earnings per share.

As with other indicators discussed in this book, many of the strategies associated with the inflation reports are aimed at getting a
jump on the data in the reports. Economists wishing to predict the
CPI and PPI numbers keep a close eye on material and commodity
prices. Increases at that stage can lead to price hikes farther down
the production pipeline. This is known as cost-push inflation—an
industry experiencing rising costs for materials, capital, labor, or
land passes the increase on to another sector of the economy by
charging higher prices for its own goods or services. When copper
prices rise, for instance, homebuilders and buyers feel the pinch.
The Copper Development Association, a trade council, estimates
that the average single-family home uses about 439 pounds of the
metal, in roofing, flanging, gutters, plumbing, circuitry, wire, fillings, valves, appliances, hardware, and lighting fixtures. Builders
pass their increased cost for these items along to the buyers. So,
when copper prices are on the rise, it’s a safe bet that new-housing
prices will be rising as well.
The story is similar with natural gas and aluminum. Natural
gas not only heats millions of homes in the winter but also is the
second-largest resource, behind coal, used in electricity production.
One of the most energy-dependent operations is aluminum manufacturing. Higher prices for natural gas can thus push up the price
of aluminum. That in turn boosts the prices of goods such as cars,
which use aluminum or aluminum derivatives in fenders, motors,
axles, bodies, wheels, and other components.
Just about every serious business periodical contains some measure of commodity and raw-materials prices. The Wall Street Journal,
the Financial Times, Investor’s Business Daily, The Economist, Barron’s,

Source:

HOW TO USE WHAT YOU SEE

Consumer and Producer Price Indices



243

and Business Week have detailed listings and usually publish graphs of
the most pronounced movements in a select group of goods.
One word of caution: It is not enough to discern movements
in commodity and raw-material prices; you must also identify the
causes of those movements. Price increases that are due to heightened demand (so-called demand-pull inflation) are more likely to be
long term and passed on to end users than those caused by supplyrelated factors such as strikes, bad weather, factory explosions, and
other production disruptions.

Source:

TRICKS FROM

THE

TRENCHES

The price indices are extremely versatile and are employed in a
wide array of circumstances. Economists, for instance, use the core
rate of CPI as a deflator—subtracting the year-over-year percentage change in core CPI from the twelve-month growth rate in
the nominal values of the indicators they track so they can discern
trends without the distorting effects of inflation. Sometimes analysts tweak, combine, or compare the inflation indices to produce
other findings. The Misery Index is one product of such tweaking.
Aficionados of soprano recitals are familiar with the aria from
Antonio Vivaldi’s opera Griselda—“Agitata Da Due Venti,” or
“Battered by Two Winds.” Economics has its own battering winds:
inflation and unemployment. Acknowledging the buffeting these
ill winds can give consumers, businesses, and investors, economists
have combined the twelve-month growth rate of consumer price
inflation with the Bureau of Labor Statistics’ unemployment rate to
form the Misery Index, shown in FIGURE 12-5.
When the Misery Index rises above 13 percent, economic conditions are, well, miserable. During the seven recessions occurring in
the forty-plus years covered by the chart, the average index value
was 13.25 percent; from mid-1973 through 1984, it was 15.40 percent, with May 1980 recording a dismal 21.9 percent. Expansions,
in contrast, are characterized by average Misery Index values of
around 9.7 percent. In recent years, the Index has remained mostly
in the 6 to 7 percent range.

244



The Trader’s Guide to Key Economic Indicators

Figure 12-5

The Misery Index

25
20
15
10
5
0
1960

1963 1966 1969 1972 1975 1978 1981 1984

1987

1990

1993 1996 1999 2002

Source: U.S. Department of Labor, Bureau of Labor Statistics

YOY%

Source:

Yale University economist Ray Fair has shown in several working papers and his book, Predicting Presidential Elections and Other
Things (Stanford University Press, 2002), that changes in economic
conditions, including inflation and unemployment rates, have an
effect on voting outcomes. The high levels of the Misery Index
registered in May 1980, within months of the presidential election, pretty much guaranteed a loss for incumbent President Jimmy
Carter. Americans vote with respect to the economy, and economic
conditions during this period were the worst in decades.

References
Introduction
Burns, Arthur F. and Wesley Clair Mitchell. Measuring Business Cycles. New York:
National Bureau of Economic Research, 1946.
Cottle, Sidney, Roger F. Murray, and Frank E. Block. Graham and Dodd’s Security
Analysis. New York: McGraw-Hill, 1988.
Crescenzi, Anthony. The Strategic Bond Investor. New York: McGraw-Hill, 2002.
Goldberg, Linda, and Deborah Leonard. “What Moves Sovereign Bond Markets?
The Effects of Economic News on U.S. and German Yields.” Federal Reserve Bank
of New York Current Issues in Economics and Finance, 9, no. 9 (September 2003).
Hagenbaugh, Barbara. “Hints of Optimism Point to Rebound.” USA Today, 21
August 2003.
Hager, George. “Economists Seek Clues in Daily Life.” USA Today, 12 February
2001.
Mitchell, Wesley Clair. Business Cycles: The Problem and Its Setting. New York:
National Bureau of Economic Research, 1927.
National Bureau of Economic Research. U.S. Business Cycle Expansions and Contractions. 2003. http://www.nber.org.
Vance, Julia. “Titanium Dioxide’s Message.” The Dismal Scientist, Economy.com, 12
April 2001. http://www.dismal.com/todays_econ/te_041201_2.asp.
Yamarone, Richard. “The Business Economist at Work: Argus Research Corporation.” Business Economics, July 1999, 65–70.
Chapter 1: Gross Domestic Product
Baumol, William J., and Alan S. Blinder. Macroeconomics Principles and Policy. 5th
ed. New York: Harcourt Brace Jovanovich, 1991.

245

246



References

Bernanke, Ben S., and Andrew B. Abel. Macroeconomics. 3d ed. New York: Addison-Wesley, 1998.
Congressional Budget Office. The Budget and Economic Outlook: An Update. July
2000.
———. CBO’s Method for Estimating Potential Output: An Update. August 2001.
Jones, Charles I. “Using Chain-Weighted NIPA Data.” Federal Reserve Bank of
San Francisco Economic Letter, no. 2002–22 (2 August 2002).
Landefeld, J. Steven. “BEA’s Featured Measure of Output and Prices.” National
Association of Business Economists: NABE News, no. 113 (September 1995).
Mankiw, N. Gregory. Principles of Macroeconomics. Fort Worth, TX: The Dryden
Press, 1998.
Mitchell, Wesley C., ed. Income in the United States: Its Amount and Distribution
1909–1919. New York: Harcourt, Brace and Company, 1921.
Okun, Arthur M. “Potential GNP: Its Measurement and Significance.” Proceedings
of the Business and Economic Statistics Section of the American Statistical Association (1962): 98–104.
Popkin, Joel. “The U.S. National Income and Product Accounts.” Journal of Economic Perspectives 14, no. 2 (spring 2000): 214–224.
Prakken, Joel L., and Lisa T. Guirl. “Macro Modeling & Forecasting with ChainType Measures of GDP.” National Association of Business Economists News, no.
113 (September 1995).
Ruggles, Nancy D., and Richard Ruggles. National Accounting and Economic Policy.
Cheltenham, UK: Edward Elgar, 1999.
Steindel, Charles. “Chain-Weighting: The New Approach to Measuring GDP.”
Federal Reserve Bank of New York: Current Issues in Economics and Finance 1,
no. 9 (December 1995).
United States Department of Commerce, Bureau of Economic Analysis. An Introduction to National Economic Accounting, Methodology Papers: U.S. National
Income and Product Accounts. Washington, DC, March 1985.
———. National Income and Product Accounts of the United States, 1929–94: Volume 1, 2. Washington, DC, April 1998.
———. “GDP: One of the Great Inventions of the 20th Century.” Survey of Current Business ( January 2000): 6–14.
———.Corporate Profits, Profits Before Tax, Profits Tax Liability and Dividends
Methodology Paper. Washington, DC, September 2002.

References



247

———. Gross Domestic Product Report (various issues).
———. Survey of Current Business (various issues).
———. Website: http://www.bea.doc.gov.
Whelan, Karl. A Guide to the Use of Chain Aggregated NIPA Data. Washington,
DC: Federal Reserve Board of Governors, June 2000.
Chapter 2: Indices of Leading, Lagging, and Coincident Indicators
Anderson, Gerald H., and John J. Erceg. “Forecasting Turning Points with Leading Indicators.” Federal Reserve Bank of Cleveland Economic Commentary
(1 October 1989).
Burns, Arthur F., and Wesley Clair Mitchell. “Statistical Indicators of Cyclical Revivals.” NBER Bulletin 69 (1938).
———. Measuring Business Cycles. New York: National Bureau of Economic Research, 1946.
Diebold, Francis X., and Glenn D. Rudebusch. “Forecasting Output with the
Composite Leading Index: A Real Time Analysis.” Journal of the American
Statistical Association 86 (1991): 603–610.
Dueker, Michael J. “Strengthening the Case for the Yield Curve as a Predictor of
U.S. Recessions.” Federal Reserve Bank of St. Louis Economic Review 79, no. 2
(March/April 1997): 41–51.
Duesenberry, James S. Business Cycles and Economic Growth. New York: McGrawHill, 1958.
Estrella, A., and F. S. Mishkin. “The Yield Curve as a Predictor of U.S. Recessions.” Federal Reserve Bank of New York, Current Issues in Economics and
Finance 2, no. 7 (June 1996).
Gordon, Robert J., ed. The American Business Cycle. Chicago: University of Chicago Press, 1986.
Haubrich, Joseph G., and Ann M. Dombrosky. “Predicting Real Growth Using
the Yield Curve.” Federal Reserve Bank of Cleveland Economic Review 32, no. 1
(First Quarter 1996): 26–35.
Kozicki, Sharon. “Predicting Real Growth and Inflation with the Yield Spread.”
Federal Reserve Bank of Kansas City Economic Review (Fourth Quarter 1997):
39–57.
Mankiw, N. Gregory. Principles of Economics. Fort Worth, TX: The Dryden Press,
1998.

248



References

Mitchell, Wesley C. Business Cycles, the Problem and Its Setting. New York:
National Bureau of Economic Research, 1954.
Moore, Geoffrey H. Business Cycles, Inflation and Forecasting. 2nd ed. Cambridge,
MA: Ballinger Publishing, 1983.
———. Leading Indicators for the 1990s. Homewood, IL: Dow Jones-Irwin, 1990.
National Bureau of Economic Research. Conference on Business Cycles. New York:
National Bureau of Economic Research, 1951.
———. U.S. Business Cycle Expansions and Contractions. www.nber.orgcycles
.html.
Sherman, Howard J. The Business Cycle: Growth and Crisis Under Capitalism.
Princeton, NJ: Princeton University Press, 1991.
Stock, James H., and Mark W. Watson. Business Cycles, Indicators, and Forecasting.
Chicago: University of Chicago Press, 1993.
The Conference Board. Business Cycle Indicators, New York: The Conference
Board, several issues.
———. U.S. Leading Economic Indicators and Related Composite Indexes. New
York: The Conference Board, several issues.
Zarnowitz, Victor. Business Cycles. Chicago: University of Chicago Press, 1992.
Chapter 3: The Employment Situation
Baumol, William J., and Alan S. Blinder. Macroeconomics Principles and Policy. 5th
ed. New York: Harcourt Brace Jovanovich, 1991.
Dwyer, Jr., Gerald P. and R. W. Hafer. “Interest Rates and Economic Announcements.” Federal Reserve Bank of St. Louis Economic Review 71, no.2 (March/
April 1989): 34–46.
Fleming, Michael J., and Eli M. Remolona. “What Moves the Bond Market?” Federal Reserve Bank of New York Research Paper, no. 9706 (February 1997).
Getz, Patricia M., and Mark G. Ulmer. “Diffusion Indexes: A Barometer of the
Economy.” Monthly Labor Review. Washington, DC: U.S. Department of
Labor, April 1990.
Hagenbaugh, Barbara. “Part-Time Workforce Grows, Which Is Good and Bad.”
USA Today, 17 June 2003.
Hager, George. “Why Hasn’t Today’s Jobless Rate Skyrocketed?” USA Today, 20
August 2001.

References



249

Jossi, Frank. “Temporary Canary.” Federal Reserve Bank of Minneapolis fedgazette
(March 2003).
Leonhardt, David. “A Tale of 2 Totals In the Jobs Report.” New York Times, 28
January 2001.
Mankiw, N. Gregory. Principles of Macroeconomics. Fort Worth, TX: The Dryden
Press, 1998.
Miller, Rich. “The Labor Pool: Why Greenspan Is Worried.” Business Week, 6
December 1999.
Schweitzer, Mark, and Jennifer Ransom. “Measuring Total Employment: Are a
Few Million Workers Important?” Federal Reserve Bank of Cleveland Economic
Commentary (June 1999).
U.S. Department of Labor, Bureau of Labor Statistics. “Labor Force Data Derived
from the Current Population Survey.” Handbook of Methods (April 1997).
———. “Employment, Hours, and Earnings from the Establishment Survey.”
Handbook of Methods. http://www.bls.gov/opub/hom/pdf/homch2.pdf.
———. “How the Government Measures Unemployment.” July 2001. www.bls
.gov/cps/cps_htgm.htm.
———. The Employment Situation (various issues).
Chapter 4: Industrial Production and Capacity Utilization
Board of Governors of the Federal Reserve System. Industrial Production—1986
Edition with a Description of the Methodology. Washington, DC, 1986.
———. Federal Reserve Statistical Release, G.17. Industrial Production and Capacity Utilization (various issues).
Corrado, Carol, Charles Gilbert, and Norman Morin. “Industrial Production and
Capacity Utilization, The 2001 Annual Revision.” Federal Reserve Bulletin
(March 2002): 173–187.
Corrado, Carol, and Kristen Hamden. “Industrial Production and Capacity Utilization, The 2002 Historical and Annual Revision.” Federal Reserve Bulletin
(April 2003): 151–176.
Emery, Kenneth M., and Chih-Ping Chang. “Is There a Stable Relationship
Between Capacity Utilization and Inflation?” Federal Reserve Bank of Dallas
Economic Review (First Quarter 1997): 14–20.
Finn, Mary G. “Is ‘High’ Capacity Utilization Inflationary?” Federal Reserve Bank
of Richmond Economic Quarterly 81 (winter 1995): 1–16.

250



References

———. “A Theory of the Capacity Utilization/Inflation Relationship.” Federal
Reserve Bank of Richmond Economic Quarterly 82/3 (summer 1996): 67–86.
Garner, C. Alan. “Capacity Utilization and U.S. Inflation.” Federal Reserve Bank of
Kansas City Economic Review (Fourth Quarter 1994): 1–21.
Gittings, Thomas A. “Capacity Utilization and Inflation.” Federal Reserve Bank of
Chicago Economic Perspectives 13, no. 3 (May/June 1989): 2–9.
Hall, Robert E. Recessions. Cambridge, MA: National Bureau of Economic Research, 30 January 2001.
Koenig, Evan F. “Capacity Utilization as a Real-Time Predictor of Manufacturing Output.” Federal Reserve Bank of Dallas Economic Review (Third Quarter
1996): 16–23.
Rogers, R. Mark. “Forecasting Industrial Production: Purchasing Managers’ versus Production-Worker Hours Data.” Federal Reserve Bank of Atlanta Economic
Review ( January/February 1992): 25–36.
Shapiro, Matthew D. “Assessing the Federal Reserve’s Measures of Capacity and
Utilization.” Brookings Papers on Economic Activity, no. 1 (1989): 181–225.
Chapter 5: Institute for Supply Management Indices
Economic Cycle Research Institute. 2003. Website: www.businesscycle.com.
Harris, Ethan S. “Tracking the Economy with the Purchasing Managers’ Index.”
Federal Reserve Bank of New York Quarterly Review (fall 1991): 61, 69.
Institute for Supply Management. Institute for Supply Management’s Manufacturing Report on Business Information Kit (as of September 2002).
———. Institute for Supply Management Manufacturing Report on Business (various
issues).
———. Institute for Supply Management Non-Manufacturing Report on Business
(various issues).
Klein, Philip A., and Geoffrey H. Moore. “N.A.P.M. Business Survey Data: Their
Value as a Leading Indicator.” Journal of Purchasing and Materials Management
(winter 1988).
Raedels, Alan. “Forecasting the NAPM Purchasing Managers’ Index.” Journal of
Purchasing and Materials Management (fall 1990).

References



251

Chapter 6: Manufacturers’ Shipments, Inventories, and Orders
Gordon, Robert J. The Measurement of Durable Goods Prices. Chicago: The University of Chicago Press, for NBER, 1990.
Harberger, Arnold C., ed. The Demand for Durable Goods. Chicago: The University of Chicago Press, 1960.
Klein, L. R. Studies in Investment Behavior. New York: Conference on Business
Cycles (Universities-National Bureau Committee for Economic Research),
1951.
Mayer, Thomas, and S. Sonenblum. “Lead Times for Fixed Investment.” Review of
Economics and Statistics 37 (August 1955): 300–304.
Roos, C. F. “The Demand for Investment Goods.” American Economic Review 38
(1948): 311–20.
Stone, J. R. N., and D. A. Rowe. “The Market Demand for Durable Goods.”
Econometrica 25, no. 3 ( July 1957): 423–43.
U.S. Department of Commerce, Bureau of the Census. Manufacturers’ Shipments,
Inventories, and Orders: 1947–1963 Revised, Series M3-1. Washington, DC:
GPO (October 1963).
———. Manufacturers’ Shipments, Inventories, and Orders: 1992–2001, M3-1
(01), Washington, DC: GPO (June 2002). Website: www.census.gov/indicator/
www/m3/bench/m3-01.pdf.
———. Instruction Manual for Reporting in Monthly Survey M3, M3-SD(I). Washington, DC: GPO (21 January 2003). Website: www.census.gov/indicator/
www/m3/contactinfo/instructionman.pdf.
———. Advance Report on Durable Goods Manufacturers’ Shipments, Inventories,
and Orders (various issues).
———. Manufacturers’ Shipments, Inventories, and Orders (various issues).
Zarnowitz, Victor. Orders, Production, and Investment. New York Columbia University Press for the National Bureau of Economic Research, 1973.
Chapter 7: Manufacturing and Trade Inventories and Sales
Abramovitz, Moses. Inventories and Business Cycles: With Special Reference to Manufacturers’ Inventories. New York: National Bureau of Economic Research, 1950.
Allen, Donald S. “Changes in Inventory Management and the Business Cycle.”
Federal Reserve Bank of St. Louis Economic Review 77, no. 4 ( July/August 1995):
17–33.

252



References

Bils, Mark, and James A. Kahn. “What Inventory Behavior Tells Us about Business Cycles.” Rochester Center for Economic Research Working Paper, no. 428
(September 1996).
Blanchard, Oliver J. “The Production and Inventory Behavior of the American Automobile Industry.” Journal of Political Economy 91 ( June 1983):
365–400.
Blinder, Alan S. “Retail Inventory Behavior and Business Fluctuations.” Brookings
Papers on Economic Activity 2 (1981): 443–505.
Blinder, Alan S., and Louis J. Maccini. “Taking Stock: A Critical Assessment of Recent Research on Inventories.” Journal of Economic Perspectives 5, no. 1 (winter
1991): 73–96.
Fair, Ray C. “Production Smoothing Model Is Alive and Well.” Journal of Monetary Economics 24 (1989): 353–70.
Fitzgerald, Terry J. “Inventories and the Business Cycle: An Overview.” Federal
Reserve Bank of Cleveland Economic Review 33, no. 3 (Third Quarter 1997).
Fitzgerald, Terry J., and Jennifer K. Ransom. “What Happened to the Inventory Overhang?” Federal Reserve Bank of Cleveland Economic Commentary 15
(March 1998).
Holmes, Thomas J. “Bar Codes Lead to Frequent Deliveries and Superstores.”
Federal Reserve Bank of Minneapolis Research Department Staff Report 261 (May
1999).
Holt, Charles C., Franco Modigliani, John F. Muth, and Herbert A. Simon. Planning Production, Inventories and Work Force, Englewood Cliffs, NJ: PrenticeHall, 1960.
Huh, Chan. “Just-in-Time Inventory Management: Has It Made a Difference?”
Federal Reserve Bank of San Francisco Weekly Letter, no. 94–18 (6 May 1994).
Kahn, James A. “Inventories and the Volatility of Production.” American Economic
Review 77 (1987): 667–79.
Little, Jane Sneddon. “Changes in Inventory Management: Implications for the
U.S. Recovery.” Federal Reserve Bank of Boston, New England Economic Review
(November/December 1992): 37–65.
McConnell, Margaret M., and Gabriel Perez Quiros. “Output Fluctuations in the
United States: What Has Changed Since the Early 1980s?” Federal Reserve
Bank of New York Staff Report, no. 41 ( June 1998).
Metzler, Lloyd A. “The Nature and Stability of Inventory Cycles.” Review of Economic and Statistics 23 (1941): 113–29.

References



253

Morgan, Donald P. “Will Just-in-Time Inventory Techniques Dampen Recessions?” Federal Reserve Bank of Kansas City Economic Review (March/April
1991): 21–33.
Ramey, Valerie A. “Inventories as Factors of Production and Economic Fluctuations.” American Economic Review 79, no. 3 ( June 1989): 338–354.
U.S. Department of Commerce, Bureau of the Census. Census of Retail Trade
(various issues).
———. Annual Benchmark Report for Wholesale Trade: January 1992 Through
February 2002 (May 2002).
———. Monthly Wholesale Trade: Sales and Inventories (various issues).
———. Manufacturing and Trade Inventories and Sales (various issues).
West, Kenneth D. “The Sources of Fluctuations in Aggregate Inventories and
GNP.” Quarterly Journal of Economics 105, no. 4 (November 1990): 939–971.
Worthington, Paula R. “The Increasing Importance of Retailers’ Inventories.”
Federal Reserve Bank of Chicago Economic Perspectives 22, no. 3 (Third Quarter
1998): 2–12.
Zarnowitz, Victor. Orders, Production, and Investment. New York: Columbia University Press, for NBER, 1973.
Chapter 8: New Residential Construction
Hagenbaugh, Barbara. “More People Refinance to Wring Cash Out of Their
Homes.” USA Today, Money Section, 28 October 2002.
Hancock, Jay. “Amazing Housing Market Bound to Help the Economy.” The Baltimore Sun, 27 February 2002.
Mortgage Bankers Association, www.mbaa.org.
National Association of Home Builders. “Housing the Key to Economic Recovery.” November 2002. www.nahb.com.
National Association of Realtors, nar.realtor.org.
Smart, Tim. “The Housing Boom Gives the Economy a Boost.” Washington Post,
9 December 1998.
U.S. Department of Commerce, Bureau of the Census. Expenditures for Improvements and Repairs of Residential Properties ( Third Quarter 2001).
———., New Residential Construction (various issues).

254



References

———. New Residential Construction Documentation. www.census.gov/const/www/
newresconsdoc.html.
Chapter 9: Conference Board Consumer Confidence and University of Michigan Consumer Sentiment Indices
Bram, Jason, and Sydney Ludvigson. “Does Consumer Confidence Forecast
Household Expenditure? A Sentiment Index Horse Race.” Federal Reserve
Bank of New York Economic Policy Review ( June 1998): 59–78.
Carroll, Christopher D., Jeffrey C. Fuhrer, and David W. Wilcox. “Does Consumer Sentiment Forecast Household Spending? If So, Why?” The American
Economic Review 84, no. 5 (December 1994): 1397–1408.
Conference Board. Consumer Confidence Survey. New York: Conference Board,
various issues.
Curtin, Richard T. “Indicators of Consumer Behavior: The University of Michigan Surveys of Consumers.” Public Opinion Quarterly 46 (1982): 340–352.
———. “Index Calculations.” University of Michigan. Mimeo.
———. “Surveys of Consumers.” University of Michigan. Mimeo.
Friend, Irvin, and F. Gerald Adams. “The Predictive Ability of Consumer Attitudes, Stock Prices, and Non-Attitudinal Variables.” Journal of the American
Statistical Association 89 (December 1964): 987–1005.
Fuhrer, Jeffrey C. “What Role Does Consumer Sentiment Play in the U.S. Macroeconomy?” Federal Reserve Bank of Boston, New England Economic Review
( January/February 1993): 32–44.
Garner, C. Alan. “Forecasting Consumer Spending: Should Economists Pay Attention to Consumer Confidence Surveys?” Federal Reserve Bank of Kansas City
Economic Review (May 1991): 57–71.
Graber, Doris A. “Reading Between the Lines of Consumer Confidence Measures.” Public Opinion Quarterly 46 (1982): 336–339.
Katona, George. The Powerful Consumer. New York: McGraw-Hill, 1960.
——— and Eva Mueller. Consumer Response to Income Increases. Washington, DC:
The Brookings Institution, 1968.
Keynes, John Maynard. The General Theory of Employment, Interest, and Money.
New York: Harcourt, Brace and Company, 1936.
Kulish, Nicholas. “Michigan’s Consumer Sentiment Index for January Fell More
Than Expected.” Wall Street Journal, 22 January 2001.

References



255

Lansing, John B., and James N. Morgan. Economic Survey Methods. Ann Arbor,
MI: The University of Michigan Press, 1980.
Leeper, Eric M. “Consumer Attitudes and Business Cycles.” Federal Reserve Bank
of Atlanta Working Paper 91-11 (August 1991).
———. “Consumer Attitudes: King for a Day.” Federal Reserve Bank of Atlanta
Economic Review 77, no. 4. (July 1992): 1–15.
Linden, Fabian. “The Consumer as Forecaster.” Public Opinion Quarterly 46
(1982): 353–360.
Matsusaka, John G., and Argia M. Sbordone. “Consumer Confidence and Economic Fluctuations.” Federal Reserve Bank of Chicago Working Paper 93-13
(November 1993).
Mehra, Yash P. and Elliot W. Martin. “Why Does Consumer Sentiment Predict
Household Spending?” Federal Reserve Bank of Richmond Economic Quarterly
89, no. 4 (fall 2003): 51–67.
Mishkin, Frederic S. “Consumer Sentiment and Spending on Durable Goods.”
Brookings Papers on Economic Activity 1 (1978): 217–232.
Roper, Burns S. “The Predictive Value of Consumer Confidence Measures.” Public
Opinion Quarterly 46 (1982): 361–367.
Throop, Adrian W. “Consumer Confidence and the Outlook for Consumer
Spending.” Federal Reserve Bank of San Francisco Economic Letter, no. 91–26
(19 July 1991).
Tobin, James. “On the Predictive Value of Consumer Intentions and Attitudes.”
The Review of Economics and Statistics 41 (February 1959): 1–11.
University of Michigan. Surveys of Consumers. Ann Arbor, MI: University of
Michigan Survey Research Center, various issues.
Chapter 10: Advance Monthly Sales for Retail Trade and Food Services
Coleman, Calmetta. “Retailers Suffer Sales Slowdown for December.” Wall Street
Journal, 5 January 2001.
Driscoll, Marie. “Monthly Retail Review.” Argus Research Investment Analysis 68,
no. 284 (November 2001).
Harris, Ethan S., and Clara Vega. “What Do Chain Store Sales Tell Us about Consumer
Spending?” Federal Reserve Bank of New York Research Paper, no. 9614 ( June 1996).
Lazear, Edward P. “Retail Pricing and Clearance Sales.” American Economic Review
76 (March 1986): 14–32.

256



References

Mencke, Claire. “Hot Retail Seems to Cinch Rate Hike Even as Producer Prices
Stay Tame.” Investor’s Business Daily, April 14, 2000.
Nakamura, Leonard I. “The Retail Revolution and Food-Price Mismeasurement.”
Federal Reserve Bank of Philadelphia Business Review (May/June 1998).
Newberger, Eric C. Home Computers and Internet Use in the United States: August
2000. U.S. Department of Commerce, Bureau of the Census, Current Population Reports, P23-207 (September 2001).
U.S. Department of Commerce, Bureau of the Census. Annual Benchmark Report
for Retail Trade and Food Services: January 1992 Through March 2002. Washington, DC, May 2002.
———. Current Business Reports, Retail e-Commerce Reports, Washington, DC,
various issues 1990–2002.
———. Monthly Retail Services Branch. E-Commerce, Frequently Asked Questions (FAQs). http://www.census.gov/mrts/www/efaq.html.
Chapter 11: Personal Income and Outlays
Chow, G. C. 1957. Demand for Automobiles in the United States (‘Contributions to
Economic Analysis,’) Vol. XIII, Amsterdam: North-Holland Publishing Co.
Larkins, Daniel. “Note on the Personal Saving Rate.” Survey of Current Business
(February 1999): 8–9.
Leonhardt, David. “Belt Tightening Is Called Threat to the Economy.” New York
Times, 15 July 2001.
Madigan, Kathleen, and David Lindorff. “Consumers Have Money to Burn.” Business Week, 20 April 1998.
Mitchell, Wesley C., ed. Income in the United States, Its Amount and Distribution
1909–1919. New York: Harcourt, Brace and Company, 1921.
Moran, Larry R., and Clinton P. McCully. “Trends in Consumer Spending, 1959–
2000.” Survey of Current Business (March 2001): 15–21.
U.S. Department of Commerce, Bureau of Economic Analysis. Personal Consumption Expenditures, Methodology Paper Series MP-6. Washington, DC, June
1990.
———. National Income and Product Accounts of the United States, 1929–94: Volume 1, 2. Washington, DC, April 1998.
———. “Personal Income and Outlays.” Various issues, 2003. www.bea.gov/bea/
rels.htm.

References



257

Chapter 12: Consumer and Producer Price Indices
Aluminum Association, Inc. Industry Overview: From Alumina to Automobiles.
2003. http://www.aluminum.org.
Bryan, Michael F., and Stephen G. Cecchetti. “The Consumer Price Index as
a Measure of Inflation.” Federal Reserve Bank of Cleveland Economic Review
(December 1993): 15–24.
Bryan, Michael F., Stephen G. Cecchetti, and Rodney L. Wiggins II. “Efficient
Inflation Estimation.” National Bureau of Economic Research Working Paper
Series, 6183, Cambridge, MA, September 1997.
Bullard, James B., and Charles M. Hokayem. “Deflation, Corrosive and Otherwise.” Federal Reserve Bank of St. Louis National Economic Trends ( July
2003).
Bureau of Labor Statistics. “The Consumer Price Index.” Chapter 17 in BLS
Handbook of Methods, April 1997.
———. “The Producer Price Index.” Chapter 16 in BLS Handbook of Methods,
April 1997.
———. “Measurement Issues in the Consumer Price Index.” June 1997. Website:
www.bls.gov/cpi/cpigm697.htm.
———. “How Does the Producer Price Index Differ from the Consumer Price
Index?” PPI Program Spotlight, No. 98-3 (1998).
Cagan, Phillip, and Geoffrey H. Moore. The Consumer Price Index, Issues and Alternatives. Washington, DC: The American Enterprise Institute for Public Policy
Research, 1981.
Clark, Todd E. “Do Producer Prices Help Predict Consumer Prices?” Federal
Reserve Bank of Kansas City Research Working Paper, 97-09 (December 1997).
Copper Development Association. Copper Facts. 2003. http://www.copper.org.
Fair, Ray C. Predicting Presidential Elections and Other Things. Stanford, CA: Stanford University Press, 2002.
———. “Econometrics and Presidential Elections.” The Journal of Economic
Perspectives (Summer 2003): 89–102.
Fuerbringer, Jonathan. “Commodity Costs Soar, But Factories Don’t Bustle.” New
York Times, 31 July 2003.

258



References

Groshen, Erica L., and Mark E. Schweitzer. “Identifying Inflation’s Grease and
Sand Effects in the Labor Market,” Federal Reserve Bank of Cleveland Working
Paper, no. 9705 (July 1997).
Hagenbaugh, Barbara. “Fuel Prices ‘Cream’ Firms.” USA Today, 6 August 2003.
Hess, Gregory D., and Mark E. Schweitzer. “Does Wage Inflation Cause Price
Inflation?” Federal Reserve Bank of Cleveland Policy Discussion Paper, no. 1
(April 2000).
Madrick, Jeff. “Economic Scene.” New York Times, 27 December 2001.
Motley, Brian, and Allison Wallace. “A Better CPI.” Federal Reserve Bank of San
Francisco Economic Letter, no. 99-05 (5 February 1999).
Platt, Gordon. “No Pickup in Inflationary Pressures.” The Journal of Commerce,
30 July 1998.
Solomon, Jolie. “An Economic Speedometer Gets an Overhaul.” New York Times,
23 December 2001.
United States Advisory Commission to Study the Consumer Price Index. Toward a
More Accurate Measure of the Cost of Living, Final Report. 4 December 1996.
Wu, Tao. “Improving the Way We Measure Consumer Prices.” Federal Reserve
Bank of San Francisco Economic Letter, no. 2003–24 (22 August 2003).

Index

Adecco, 85
Advance Monthly Sales for Retail
Trade and Food Services.
See Retail Trade and Food
Services report
after-tax profits, 29
aggregate expenditure equation,
18–22
American Forest and Paper
Association, 92
American Staffing Association, 85
Ann Taylor, 205
base year, problem with, 23–24
Beazer Homes USA, 171
Beige Book, 129
BJ’s Warehouse, 198
Boeing, 39
book profits, 29
Boskin, Michael, 233
Boskin Report, 233
Briggs & Stratton, 38
building permits for new private
housing, 57
Bureau of Economic Analysis
(BEA), 12
how data is collected, 15–16
inventories and the aggregate
expenditure formula, 147
Personal Income and Outlays
report, 21, 54, 209–228
Burns, Arthur, 3, 53

Bush, George W., 188
business cycle, 3–5
housing and, 167–169
inventories and, 153–154
price indices and, 241–242
Business Cycle Indicators, 51
how to use, 66–68
Capacity Utilization: Manufacturing,
Mining, Utilities and
Industrial Materials, 91
capacity utilization production, as an
indicator
description of, 94–95, 100–104
evolution of, 90–91
how to use, 104–105
capital consumption adjustment
(CCAdj), 28
capital goods, nondefense, 56–57
capital spending, 21, 36–38
Carlyle, Thomas, 11
Carter, Jimmy, 244
Caterpillar, 114
chain weighting, 24
Challenger, Gray and Christmas, 85
Chamber of Commerce, 108
Cleveland Fed Index, 235
Clinton, Bill, 188
coincident economic indicators
(index), 5, 51–52, 54–55, 60,
64–65
coincident-to-lagging index, 68
259

260



Index

commodity prices, 242–243
Conference Board, 53
Business Cycle Indicators, 51,
66–68
Consumer Confidence Survey,
176–191
Help-Wanted Advertising Index,
85–86
Index of Coincidental Indicators,
12
Congressional Budget Office (CBO),
46
constant dollars, 22–25
consumer confidence and sentiment
Conference Board’s Consumer
Confidence Survey, 176–191
durable goods spending and,
184–185
employment and, 186
noneconomic influences on,
186–189
University of Michigan’s Surveys
of Consumers, 176–191
Consumer Confidence Survey, 176
collection of data, 178–179
evolution of, 177
how to use, 185–191
meaning of, 179–185
consumer credit, ratio of personal
income to, 63–64, 217
Consumer Expenditure Survey
(CEX), 229–230, 233
consumer goods, manufacturers’ new
orders for, 56–57
Consumer Price Index (CPI), 34, 35,
64, 201
business cycle and, 241–242
CPI-U, 230
CPI-W, 230
data sources, 233–235

evolution of, 232–233
how to use, 242–244
inflation rate calculation, 237–
238
markets and, 240–241
meaning of, 238–242
role of, 229–230
Consumer Sentiment and Current
Economic Conditions, 61
Consumer Sentiment Index, 82
consumer spending, 19–21, 219–220
consumption
See also Retail Trade and Food
Services report
government, 21–22, 39
personal, 19–21, 35–36, 213–214
Copper Development Association,
242
Core PCED, 35
corporate profits, 29–30, 43–45
Costco, 198
Cummins Inc., 38
current dollars, 22–25
Current Employment Statistics
(CES) (establishment/payroll
survey), 70, 71, 74–77
Current Industrial Report (CIR), 132
Current Population Survey (CPS)
(household survey), 70, 71,
73–74
Curtiss-Wright, 39
CVS, 205
cyclical unemployment, 73
Danaher Corp., 38
Deere & Co., 38
deflation, 33, 238, 239–240
deflators, 25–26, 34–35
diffusion indices, 86–87
vendor performance, 57

Index

discouraged workers, 73
disinflation, 238, 239
disposable personal income, 213
Dodd, David, 1
Dover Corp., 38
durable goods, 20, 21
consumer confidence and
spending on, 184–185
Durable Goods report, 134–138,
139–142
Eagle Homes, 171
Eaton Corp., 38, 114
e-commerce, 198–200
Economic Cycle Research Institute,
123
economic indicators
classification of, 5
distribution of reports, 5–6
evolution of, 13–14, 53
employed, use of term, 72
employee(s)
compensation, 27–28
number of, on nonagricultural
payrolls, 54
employment, as an indicator
aggregate hours worked index,
83
consumer confidence and, 186
Current Employment Statistics
(CES) (establishment/payroll
survey), 70, 71, 74–77
Current Population Survey (CPS)
(household survey), 70, 71,
73–74
definitions, 72–73
impact of, 69–70, 77
relationship of unemployment,
business cycle and, 78–80
spending and, 227–228



261

state information, 71
temporary workers, 84–85
Employment Index (ISM), 117–118
employment-population ratio, 73
Employment Situation report (BLS),
54, 69–70
collection of data, 72–77
evolution of, 71–72
how to use, 85–87
meaning of, 77–85
establishment/payroll survey, 70, 71,
74–77
Factory Orders report, 138–139,
142–143
Fair, Ray, 244
Federal National Mortgage
Association (FNMA), 166
Federal Reserve
Flow of Funds, 224
Industrial Production and
Capacity Utilization, 55,
89–105
monetary aggregates, 58
wealth effect, 224–225
Federal Reserve Bank of New York,
108
final goods, 16–17
final sales, 41–43
of domestic product, 42
to domestic purchasers, 42–43
fixed capital, consumption of, 30
fixed investment, 21, 36–38
fixed-weight approach, 24
Ford Motor Co., 184–185
Franco, Lynn, 180–181
frictional unemployment, 73
GAFO (general merchandise,
apparel, furniture and other),

262



Index

195, 202–203
General Dynamics, 39
General Theory of Employment,
Interest, and Money, The
(Keynes), 175
Global Crossing, 99
goods-producing category, 75
goods-providing category, 75
Gore, Al, 188
government consumption
expenditures and gross
investment, 21–22, 39
Graham, Benjamin, 1
Greenspan, Alan, 98, 107
gross domestic product (GDP)
adjustments, 30–32
advance report, 15
benchmark revisions, 16
business cycles and, 5
calculating, 12, 18–22
collection of data, 14–16
composition of, 19
content of reports, 12
defined, 11–12, 16–17
deflators, 25–26, 34–35
evolution of, 13–14
final report, 15
growth, 32–34
how to use, 46–49
meaning of, 32–45
national income, 26–27
nominal (current dollars), 22–25
output gap, 46–49
preliminary report, 15
real (constant dollars), 22–25
versus gross national product,
17–18
gross domestic purchases, 42
gross national product (GNP)
adjustments, 30–32

defined, 11
gross domestic product versus,
17–18
gross private domestic investment,
21
G17 report. See Industrial
Production and Capacity
Utilization
Help-Wanted Advertising Index,
85–86
Home Depot, 205
hoot-and-holler task, 6
Hoover, Herbert, 108
household survey, 70, 71, 73–74
housing. See New Residential
Construction
Hovnanian Enterprises, 171
Illinois Tool Works, 38
implicit price deflator, 25
Income in the United States: Its
Amount and Distribution
1909–1919 (Mitchell), 211
Indexes of Domestic Business, 90
Index of Coincidental Indicators, 5,
12, 51–52, 54–55, 60, 65
Index of Consumer Expectations
(ICE), 61, 178, 182–183
Index of Consumer Sentiment (ICS),
178, 183
Index of Current Economic
Conditions (ICC), 178
Index of Production in Selected
Basic Industries, 90
indirect business taxes, 30
Industrial Materials Price Index, 123
industrial production, as an indicator
description of, 92–94, 95–99
evolution of, 90–91

Index

how to use, 104–105
Industrial Production and Capacity
Utilization (Federal Reserve),
55
collection of data, 91–95
evolution of, 90–91
how to use, 104–105
meaning of, 95–104
role of, 89–90
industrial production index, total, 55
inflation indicators, 80–81, 238–239
inflation measure, 35
inflation rate, calculating annualized,
25–26, 237–238
Ingersoll-Rand, 38, 114
Institute for Supply Management
(ISM)
Employment Index, 117–118
Inventories Index, 126
Manufacturing Report on
Business, 108
New Orders Index, 126, 127, 128
New Orders minus Inventories
Index, 126–127
Non-Manufacturing Report on
Business, 123–125
Price Index, 119–121, 127–128
Purchasing Managers’ Index
(PMI), 57, 97–98, 107–109,
112–117
Supplier Deliveries Index, 121–
123
2003 Survey, 110–111
interest expense, 214
interest rate spread, 59–61
Internal Revenue Service, 92
International Council of Shopping
Centers, 205
International Mass Merchants
Association, 205



263

International Paper, 114
inventories, 21
See also Manufacturing and
Trade Inventories and Sales
(MTIS)
business cycle and, 153–154
importance of, 147–148
ratio of manufacturing and trade,
to sales, 62
-to-sales ratios, 154–155
inventory valuation adjustment
(IVA), 28
job leavers, 72
job losers, 72
Journal of Commerce, 123
judgment trend, 30
K&B Homes, 171
Katona, George, 177
Kelly Services, 85
Keynes, John Maynard, 175
Kuznets, Simon, 13, 211
labor force participation rate, 73
labor pool, available, 81
lagging economic indicators/index,
5, 52, 61–64, 66
layoff announcements, 85
leading economic indicators/index
(LEI), 5, 51, 55–61, 65–66,
67
Leggett & Platt, 114
Lennar Corp., 171
loan levels, outstanding commercial
and industrial, 63
Lockheed Martin, 39
Lowe’s, 205
macroeconomy, 14

264



Index

Manpower, 85
Manufacturers’ Shipments,
Inventories, and Orders (M3)
survey, 62, 149
Durable Goods report, 134–138,
139–142
evolution of, 132–133
Factory Orders report, 138–139,
142–143
how to use, 143–146
manufacturing and retail trade sales,
55
Manufacturing and Trade
Inventories and Sales (MTIS)
report, 55, 62
business cycle and, 153–154
collection of data, 149–151
evolution of, 149
how to use, 155–157
importance of, 147–148
inventories-to-sales ratios, 154–
155
meaning of, 151–155
manufacturing labor cost per unit of
output, 62–63
Manufacturing Report on Business,
108
markets, price indices and, 240–241
McGraw-Hill, 91
MDC Holdings, 171
Measuring Business Cycles (Burns
and Mitchell), 3
Median Index, 235
Misery Index, 243–244
Mitchell, Wesley C., 3, 53, 210–211
Money, 176
money supply, 58
Monthly Report of Unemployment,
71
Mortgage Bankers Association

(MBA), 171
motor vehicles, 222–223
Mueller, Eva, 177
NAICU (non-accelerating
inflationary rate of capacity
utilization), 102–103
NAIRU (non-accelerating inflation
rate of unemployment), 80,
102
Nakamura, Leonard, 198
National Association of Home
Builders (NAHB), 163, 164
National Association of Purchasing
Agents, 108
National Association of Purchasing
Management, 108
National Association of Realtors, 160
National Bureau of Economic
Research (NBER), 3, 13, 53,
89, 90, 95, 181, 210, 220
national defense spending, 39
national income
accounting, 13
adjustments, 30–32
categories of, 27–30
defined, 26–27
National Income and Product
Accounts (NIPAs), 12, 89
collection of data, 14–16
evolution of, 13–14, 211
National Retail Federation, 205
net exports, 22, 40–41
net interest, 28
net transfers, 214
new orders, 135
New Orders Index (ISM), 126, 127,
128
New Orders minus Inventories Index
(ISM), 126–127

Index

New Residential Construction, 57
business cycle and, 167–169
collection of data, 161–163
evolution of, 160–161
housing indicators, 160
how to use, 171–173
importance of, 159
influences on, 164–166
meaning of, 163–171
regional differences, 166–167
single-family housing starts 169–
171
NFO Research, 178
nominal (current dollars), 22–25
nondefense capital goods
spending excluding aircraft
(NDCGXA), 144–146
nondurable goods, 20, 21, 138–139
nonfarm payrolls, 54, 75–77
Non-Manufacturing Report on
Business (ISM), 123–125
nonresidential investment, 21, 36, 37
North American Industry
Classification System
(NAICS), 91, 93, 133, 149
Northrop Grumman, 39
operating profits, 29
Outlook, The, 57
output gap, 46–49
Paccar Inc., 38
Parker-Hannifin Corp., 38
PCEDXF&E, 34
personal consumption expenditures,
19–21, 35–36, 213–214
personal income
calculation of, 213
disposable, 213
less transfer payments, 54–55



265

ratio of consumer credit to, 63–
64, 217
use of, 216–219
wealth effect, 224–225
Personal Income and Outlays report
(BEA), 21, 54
collection of data, 211–215
evolution of, 210–211
how to use, 222–228
importance of, 209–210
meaning of, 215–221
personal outlays, 214
personal savings, 215
personal savings rate, 215, 220–221
Phillips, A. W., 80
Physical Volume of Trade, 90
Predicting Presidential Elections and
Other Things (Fair), 244
pretax profits, 29
Price Expectations Index, 190
Price Index (ISM), 119–121, 127–
128
price indices. See Consumer Price
Index (CPI); Producer Price
Index (PPI)
prices
role of spending and, 226–227
trends, 238–240
prime rate, 63
produced goods, 17
Producer Price Index (PPI), 34, 119,
120
business cycle and, 241–242
data sources, 235–237
evolution of, 231–232
how to use, 242–244
inflation rate calculation, 237–
238
markets and, 240–241
meaning of, 238–242

266



Index

role of, 229, 230–231
profits
after-tax, 29
from current production, 29
operating, 29
pretax, 29
proprietors’ income, 28–29
Pulte Homes, Inc., 171
Purchasing Managers’ Index (PMI),
57, 97–98, 107–108
collection of data, 109–113
description of, 113–117
evolution of, 108–109
industries cover, 109
quantity theory of money, 58
Qwest, 99
raw-materials prices, 242–243
Raytheon, 39
real (constant dollars), 22–25
reference week, 73
rental income, 29
residential construction. See New
Residential Construction
residential investment, 21, 36–37
Retail Trade and Food Services
report, 149
collection of data, 195–200
e-commerce and superstores,
198–200
evolution of, 194
GAFO (general merchandise,
apparel, furniture, and other),
195, 202–203
holiday sales, role of, 207–208
how to use, 204–208
impact of, 193–194
meaning of, 200–203
same-store sales, 205–206

seasonality issues, 206
Robert Half, 85
Ryland Group, 171
sales, ratio of manufacturing and
trade inventories to, 62,
154–155
Sam’s Club, 198
seasonal unemployment, 72–73
Security Analysis (Graham and
Dodd), 1
service payrolls, 75–76
services, 20–21
shipments, 135
single-family housing starts 169–171
smoothing techniques, 140–141
Sports Authority, 205
Standard Industrial Classification
(SIC), 91, 109, 133, 236
statistical discrepancy, 31
stock prices, 57–58
structural unemployment, 73
subsidies, 31
Summary of Commentary on Current
Economic Conditions by
Federal Reserve District, 129
Supplier Deliveries Index (ISM),
121–123
Survey of Construction (SOC), 161
Surveys of Consumers, 176
collection of data, 177–178
evolution of, 177
how to use, 185–191
meaning of, 179–185
survey week, 73
Tanner’s Council of America, 92
tape reading, 6
temporary workers, 84–85
Timken Co., 38

Index

Toll Brothers, 171
transfer payments, 30–31
personal income less, 54–55
underground economy, 17
unemployed, use of term, 72
unemployment
claims, average number of
weekly initial, 56
duration, average, 62
inflation indicators and, 80–81
as a lagging indicator, 78, 80
rate, calculation of, 74
sentiment and, 81–82
types of, 72–73
unfilled orders, 135
United Nations System of National
Accounts, 18
U.S. Department of Commerce,
Bureau of the Census, 53, 91
Manufacturers’ Shipments,
Inventories, and Orders (M3),
62, 131–146
Manufacturing and Trade
Inventories and Sales
(MTIS), 55, 62, 147–157
New Residential Construction,
57, 159–171
Retail Trade and Food Services
report, 149, 193–208
Value of Construction Put in
Place, 162
Wholesale Trade Survey, 62
U.S. Department of Housing and
Urban Development, New
Residential Construction, 57,
159–171
U.S. Department of Labor, Bureau of
Labor Statistics
Consumer Price Index (CPI), 34,



267

35, 64, 201, 229–244
Employment Situation report, 54,
69–87
Producer Price Index (PPI), 34,
119, 120, 229, 230–244
U.S. Geological Survey, 92
United Technologies, 114
University of Michigan, Survey
Research Center, 61
Survey of Consumers, 176–191
Value of Construction Put in Place,
162
vendor performance diffusion index,
57
vendor supplier index, 121–123
Wal-Mart, 204
warehouse clubs and superstores,
198–200
wealth effect, 224–225
weekly manufacturing hours,
average, 56
Whelan, Karl, 24
Wholesale Price Index (WPI), 230,
231
See also Producer Price Index
Wholesale Trade Survey, 62
Wilshire 5000, 224
Wolverine Tube Inc., 38
WorldCom, 99
yield curve plots, 59–61

About Bloomberg

Bloomberg L.P., founded in 1981, is a global information services,
news, and media company. Headquartered in New York, the company
has sales and news operations worldwide.
Bloomberg, serving customers on six continents, holds a unique position within the financial services industry by providing an unparalleled
range of features in a single package known as the BLOOMBERG PROFESSIONAL® service. By addressing the demand for investment performance and efficiency through an exceptional combination of information, analytic, electronic trading, and Straight Through Processing tools,
Bloomberg has built a worldwide customer base of corporations, issuers,
financial intermediaries, and institutional investors.
BLOOMBERG NEWS®, founded in 1990, provides stories and columns on business, general news, politics, and sports to leading newspapers and magazines throughout the world. BLOOMBERG TELEVISION®, a 24-hour business and financial news network, is produced
and distributed globally in seven different languages. BLOOMBERG
RADIOSM is an international radio network anchored by flagship station
BLOOMBERG® 1130 (WBBR-AM) in New York.
In addition to the BLOOMBERG PRESS ® line of books, Bloomberg
publishes BLOOMBERG MARKETS ® and BLOOMBERG WEALTH
MANAGER®. To learn more about Bloomberg, call a sales representative at:
London: . . . . . +44-20-7330-7500
New York: . . . +1-212-318-2000
Tokyo: . . . . . . +81-3-3201-8900

About the Author
Richard Yamarone, vice president and director of economic
research at Argus Research Corp., has more than fifteen years of
experience on Wall Street analyzing and researching domestic and
international economic trends, monitoring monetary and fiscal
policy developments, and forecasting the U.S. macroeconomy. He
has worked for several international banks and domestic money
centers and investment banks in a variety of senior positions. In
addition, he has taught numerous courses on macroeconomics and
economic indicators at several colleges and institutions. He has
served as president of the prestigious Downtown Economists Club
of New York City and is a member of the National Association
of Business Economists, the American Economic Association, the
New York State Economic Association, and the NYU Moneymarketeers. At Argus Research, Yamarone is responsible for
establishing the firm’s top-down economic and interest rate forecasts as well as its estimates for monthly economic indicators. He
makes frequent appearances on business television and radio shows,
including CNBC, CNNfn, Bloomberg TV and Bloomberg Radio,
and National Public Radio.

for in-depth market information and news, visit the Bloomberg
website at www.bloomberg.com, which draws from the news and
power of the BLOOMBERG PROFESSIONAL® service and
Bloomberg’s host of media products to provide high-quality news
and information in multiple languages on stocks, bonds, currencies,
and commodities.

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.6
Linearized                      : Yes
XMP Toolkit                     : Adobe XMP Core 4.0-c316 44.253921, Sun Oct 01 2006 17:14:39
Modify Date                     : 2009:08:27 11:25:46+02:00
Create Date                     : 2006:09:23 11:53:58-04:00
Metadata Date                   : 2009:08:27 11:25:46+02:00
Format                          : application/pdf
Title                           : The Trader's Guide to Key Economic Indicators
Creator                         : RICHARD YAMARONE
Document ID                     : uuid:4ca9f2e7-d31d-480b-8cc8-7eee054facc8
Instance ID                     : uuid:bfd54e47-f644-4f65-bb29-e86b1aaaca3c
Has XFA                         : No
Page Count                      : 284
Author                          : RICHARD YAMARONE
Producer                        : Adobe Acrobat 7.0

EXIF Metadata provided by EXIF.tools

The Trader's Guide To Key Economic Indicators Bloomberg.the Traders Indicators.2004.isb

Navigation menu

Versions of this User Manual:

Views

Navigation