Cost Accounting TopClass
Igotyou
BusinessCycleIndicators.pdf
Volume 8, Number 1
January 2003
Business Cycle Indicators
A monthly report from The Conference Board
After declining from May through September 2002, the leading index has now improved for three straight months, suggesting a stronger economic recovery in the first half of 2003.
The negative contribution from unemployment insurance claims in December was more than offset by the positive contributions from housing permits, average manufacturing workweek, and consumer expectations. The leading index has almost regained its level in May 2002, or when it started to decline.
The coincident index, a measure of current economic activity, was flat in each of the last three months of 2002. But it’s still 0.8 percent above its cyclical low reached in November 2001.
* This release incorporates annual benchmark revisions to the composite indexes which brings them up-to-date with revisions in component data in the past year.
For more information, visit our website at www.globalindicators.org
2 January 2003 The Conference Board
About Business Cycle Indicators This report is an exciting complement to the
Board’s economics program. To ensure that
Business Cycle Indicators is one of the most
useful resources for monitoring current economic
conditions, suggestions pertaining to content
are welcomed and should be addressed to
Michael Fort at The Conference Board, or sent
via e-mail at [email protected].
For inquiries about subscriptions, call Customer
Service at 212 339 0345.
On the Web
Please visit us at: www.globalindicators.org and www.conference-board.org
Business Cycle Indicators
Published by The Conference Board, Inc. 845 Third Avenue, New York, NY 10022-6679 212 759 0900
President & CEO of The Conference Board RRiicchhaarrdd EE.. CCaavvaannaagghh
Director of Economic Research RRoobbeerrtt HH.. MMccGGuucckkiinn
Senior Fellow EEddggaarr RR.. FFiieeddlleerr
Senior Fellow VViiccttoorr ZZaarrnnoowwiittzz
Economist AAttaammaann OOzzyyiillddiirriimm
Manager, Business Cycle Indicators Program MMiicchhaaeell FFoorrtt
Business Cycle Analyst JJaacciinnttoo TToorrrreess
Editor JJoohhnn RRaaddzziinn
Production Manager & Design PPeetteerr DDrruubbiinn
Production Editor JJooaannnn LLeemmoonn Business Cycle Indicators (ISSN 1088-7857) is published monthly by The Conference Board, Inc.,
845 Third Avenue, New York, NY 10022-6679. Printed
in Alpha, NJ. Periodicals postage paid at New York, NY,
and other mailing offices.
POSTMASTER: Send address changes to
Business Cycle Indicators, The Conference Board, Inc., 845 Third Avenue, New York, NY 10022-6679.
A subscription to the BCI monthly publication is
$130 for twelve issues. Go to www.globalindicators.org
to subscribe.).
Copyright © 2003 The Conference Board, Inc. (“TCB”)
All rights reserved. TCB does not claim copyright to any
original work prepared by a U.S. Government officer or
employee as part of that person’s official duties. TCB
acknowledges transfer of responsibility for the “Cyclical
Indicators” from the U.S. Department of Commerce,
pursuant to U.S. Contract No. 52–SABA–5–00128.
The Conference Board and the torch logo are registered
trademarks of The Conference Board, Inc.
2003 Annual Benchmark Revisions to the Composite Indexes by Jacinto L. Torres Jr., Business Cycle Analyst
The January 23, 2003 release of The Conference Board’s U.S. composite indexes of leading, coincident, and lagging indicators through December 2002 incorporated annual benchmark revisions. Benchmark revisions have long been part of the index methodology and were adopted to avoid numerous minor revisions to the index during the course of the year. The process essentially updated the composite indexes to include the revisions made to the history of the components in the past year. Throughout the year, monthly updates to the composite indexes only included revisions to the underlying component data going back six months. Since the com- position of the indexes was not altered, the changes were very minor and the cycli- cal performance of the indexes was not affected. Nonetheless, the new benchmarked indexes are not strictly comparable to those published previously.
The last scheduled benchmark was in January 2002. However, in August 2002, The Conference Board undertook a midyear benchmark revision to accommodate significant changes, especially in the National Income and Product Accounts (NIPA), by the Bureau of Economic Analysis (BEA). The BEA announced sub- stantial data revisions in July 2002, which in turn had significant effects on our composite indexes, most notably the coincident index. A midyear benchmark revi- sion in August 2001 was also necessary to accommodate the conversion of many of the BEA statistics from the Standard Industrial Classification (SIC) to the North American Industrial Classification System (NAICS). However, midyear benchmark revisions to the composite indexes are not commonplace.
Standardization Factors The Conference Board’s index methodology includes an adjustment, through the standardization factors, which equalizes the volatility of each component in the index. The standardization factors are calculated by inverting the standard devia- tion of the monthly symmetric changes in each component and then normalized so that they add up to one. Hence, these factors only take into account the rela- tive volatility of the component indicators. Thus, a volatile component would get a smaller resulting standardization factor but this would not necessarily mean that it has less significance to the index performance. Components that are wider in coverage, however, typically tend to be less volatile. This results in larger stan- dardization factors for those components.
Many years ago, at the National Bureau of Economic Research and the U.S. Department of Commerce, the index methodology included component weighting. This was in addition to the volatility adjustment mentioned previously. The weights used then were derived from a complex scoring process performed on hundreds of individual economic indicators. A better performing indicator would get a larger score on the 0-100 scale. Performance was measured in terms of economic signifi- cance, statistical adequacy, timing, conformity, smoothness, and currency. The indi- cators with the highest scores were then selected to be included in one of the three indexes, depending on their cyclical classification. After these same scores were nor- malized to sum to one, they were used to act as the weights of the components of the associated index. Accordingly, the component with a higher score had more sig-
Business Cycle Indicators January 2003 3
nificance to the index. Because the high- est-scored indicators were selected, first to be included in the publication and then in the composite indexes, their scores were very narrowly dispersed around one. As a result, the weights were all almost equal and therefore, had little effect on the per- formance of the composite indexes. This process was dropped late in the 1980s and equal weights were given to each of the chosen component of the indexes.
Many users and followers of the composite indexes often wrongly refer to the standard- ization factors as weights. The current index methodology, when it adopted a consistent weighting system, effectively assumed that the components had equivalent importance to the performance of the indexes.
The standardization factors are updated at the same time annual benchmark revi- sions are undertaken. Last year’s stan- dardization factors were based on the period from 1959-2000. This year’s new standardization factors are based on the period from 1959-2001. Table 1 shows the standardization factors used in 2002 and the factors for 2003. As you will see in the table, the standardization factors do not change very much from year to year.
Component Revisions and Cyclical Performance of the Composite Indexes The BEA revisions in July 2002 affected manufacturers’ new orders for consumer goods and materials, and manufacturers’ new orders for nondefense capital goods. None of the other eight leading compo- nent series had any major revision in 2002.
Among the components of the coincident index, personal income less transfer payments and manufacturing and trade sales were the series mostly affected by the July 2002 BEA revisions. In December 2002, the industrial production index was reclassified from the SIC to the NAICS by the Federal Reserve. Moreover, the Federal Reserve also rebased this series as a percentage of output in 1997, changed from 1992.
Among the lagging index components, the July 2002 BEA revisions only affected the ratio of manufacturing and trade inventories to sales. Following the revisions by the Federal Reserve last December, one lagging index component, change in labor cost per unit of output manufacturing, was also revised since this particular series is calculated using indus- trial production for manufacturing.
One of the strengths of the index methodology is its ability to keep the indexes very stable. While revisions to the compo- nents may be substantial at times, these revisions are usually not reflected to the same extent in the index. Component revisions tend to offset each other, thereby diminishing the impact of data revisions to the index.
The levels of the new composite indexes differ from their levels prior to the revision and therefore must not be spliced. Their cyclical patterns, however, remain unchanged. The peak and trough dates of the leading, coincident, and lagging indexes, and the coincident-to-lagging ratio after the bench- mark process and the adoption of the new standardization factors are the same as peak and trough dates prior to the benchmark revision. The accompanying charts show the overall effect of the benchmark revisions on the indexes.
TTaabbllee 11 UU..SS.. CCoommppoossiittee IInnddeexxeess:: CCoommppoonneennttss aanndd SSttaannddaarrddiizzaattiioonn FFaaccttoorrss
August January 2003 2002 2002
Leading Index 1. Average weekly hours, manufacturing .1946 .1812 .1812
2. Average weekly initial claims for
unemployment insurance .0268 .0261 .0241
3. Manufacturers’ new orders, consumer
goods and materials .0504 .0496 .0456
4. Vendor performance, slower deliveries
diffusion index .0296 .0276 .0277
5. Manufacturers’ new orders, nondefense
capital goods .0139 .0130 .0131
6. Building permits, new private housing units .0205 .0191 .0191
7. Stock prices, 500 common stocks .0309 .0308 .0310
8. Money supply, M2 .2775 .3038 .3068
9. Interest rate spread, 10-year Treasury bonds
less federal funds .3364 .3305 .3330
10. Index of consumer expectations .0193 .0183 .0185
Coincident Index 1. Employees on nonagricultural payrolls .5186 .5230 .4805
2. Personal income less transfer payments .2173 .2176 .2814
3. Industrial production .1470 .1407 .1292
4. Manufacturing and trade sales .1170 .1187 .1090
Lagging Index 1. Average duration of unemployment .0368 .0378 .0367
2. Inventories to sales ratio, manufacturing and trade .1206 .1257 .1225
3. Labor cost per unit of output, manufacturing .0693 .0624 .0611
4. Average prime rate .2692 .2521 .2454
5. Commercial and industrial loans .1204 .1300 .1265
6. Consumer installment credit to personal
income ratio .1951 .1992 .2209
7. Consumer price index for services .1886 .1929 .1869
4 January 2003 The Conference Board
Leading Index (1996=100)
96
100
104
108
112
116
95 96 97 98 99 00 01 02
March 2001 Peak
Before Benchmark Revision
After Benchmark Revision
Coincident Index (1996=100)
March 2001 Peak
96
100
104
108
112
116
120
95 96 97 98 99 00 01 02
After Benchmark Revision
Before Benchmark Revision
96
98
100
102
104
106
108
110
95 96 97 98 99 00 01 02
Lagging Index (1996=100)
March 2001 Peak
After Benchmark Revision
Before Benchmark Revision
0.96
1.00
1.04
1.08
1.12
1.16
1.20
95 96 97 98 99 00 01 02
Ratio of Coincident to Lagging Index
March 2001 Peak
Before Benchmark Revision
After Benchmark Revision
2003 Benchmark Revisions
