Indicator Qualities of the NAPM Report On Business(r).
The National Association of Purchasing Management (NAPM) has collected information on business conditions in the manufacturing sector of the economy since the early 1920s. Since 1931 NAPM has published business survey data on a regular basis (except for four years during World War II). The NAPM business survey, now known as the NAPM Report On Business(r), has become one of the most widely known and watched indicators of business activity in the United States. Its concept has also been applied In the United Kingdom and other countries by NAPM counterpart organizations. Because the NAPM Report On Business(r) spans numerous business cycles, economists find the NAPM data useful in studying long-term economic trends. 
In the post-World War II era, NAPM data from 1948 through May 1971 include series on Production, New Orders, Inventories, Employment, and Commodity Prices. Starting in 1971 several series have been added: Supplier Deliveries in June 1971, New Export Orders in January 1988, Imports in October 1989, and Backlog of Orders in January 1993. Also, in the early 1980s the Purchasing Manager's Index (PMI) was developed by the U.S. Department of Commerce and NAPM (Torda 1985). The PMI consists of a weighted average of the seasonally adjusted NAPM indexes for: (weights in parentheses) Production (0.25), New Orders (0.30), Employment (0.20), Supplier Deliveries (0.15), and Inventories (0.10). The weights were heuristically determined to reflect the maximum relationship between the indexes and GNP. After development, the PMI was back-calculated for prior years and has continued to be provided.
Much has been written over many years concerning use of the data in the NAPM Report On Business(r) as economic indicators but there has not been any recent attempt to review and evaluate the research. This article reviews the representation of business activity by economic indicators, compares qualities of NAPM Report On Business(r) data to those desirable in indicators, summarizes previously published research on the use of NAPM data as indicators, and presents some new and previously unpublished research on this subject. The use and value of NAPM data in the development of purchasing and supply strategy is also discussed. It is concluded that NAPM data possesses many desirable indicator qualities, are useful In the analysis of business and economic activity, and have value in the development of purchasing and supply strategy.
The NAPM business survey includes data from over 350 purchasing professionals in the 20 two-digit Standard Industrial Classification (SIC) manufacturing categories. To proportionally represent all 20 manufacturing categories, the sample is stratified according to the percentage of manufacturing GDP represented by each category. The sample is not randomly selected. Efforts are made to include a mix of company sizes and a broad geographic distribution of company locations in each category. To avoid inflationary influences, respondents are instructed to use unit measurements rather than dollars when responding to questions. Since the late 1950s, the survey has received increased publicity, study, and recognition (Hoagland 1959, 1960, 1964, 1969; Hoagland and Taylor 1987; Klein and Moore 1988; Raedels 1990; Dasgupta and Lahiri 1991; Harris 1991; Niemira 1991; Wall Street Journal, April 6, 1995).
The first objective of this article is to position the NAPM survey data in the context of indicators of business activity. This is accomplished by reviewing the use of economic indicators to represent business activity in general and manufacturing activity in particular and by comparing the attributes of the NAPM indicators to desirable indicator properties. The second objective is to review previous studies of the indicator qualities of the NAPM Report On Business(R) and, where possible, compare the results of these studies with the results from a new analysis. The final objective is to draw conclusions and provide examples of how the indicator qualities of NAPM data can be applied to the development of purchasing strategy.
Following a discussion of measures of business activity and of indicator qualities, the results of previous and new research are presented. An analysis examines the relationship of NAPM data to business cycles and gross national product and determines lead/lag relationships. The new analysis is referred to as "current study" in the Review and Discussion section and in the tables. When possible, current study results are compared to previous work. Four sets of research results are presented:
1. relationships between NAPM data and business cycle turning points,
2. relationships between NAPM data and U.S. economic activity as represented by real (uninflated) Gross National Product (GNP) and Gross Domestic Product (GDP),
3. correlation and lead/lag relationships of NAPM indexes with other indicators of economic activity, and
4. lead/lag relationships between the principal NAPM indexes themselves.
BUSINESS INDICATOR QUALITIES OF NAPM DATA
Measures of Business Activity
Business activity is often discussed by economists in terms of GNP or GDP.  These measures of economic output include only the "value added" at each step in the economic chain from raw material to consumer and exclude payments made by one company to another. What is overlooked by GNP and GDP is the fact that, to business people, "business" is the total amount of commerce or business transactions conducted. It has been estimated that total purchases in the U.S. economy by business, governments, and consumers (excepting financial businesses) are equal to approximately 200 percent of GNP (Hoagland 1993). Also, total economic activity in the economy, as represented by total purchases, wages and salaries, and other payments (again excepting financial businesses), is equal to approximately 300 percent of GNP (Hoagland 1993). Thus, if one is attempting to assess total "business" activity, GDP or GNP understate the total activity by a significant amount and are questionable as valid indicators of this activity.
Manufacturing Business Activity
When NAPM began collecting data in the early 1920s, the membership of NAPM was largely from manufacturing industries. As both the membership of NAPM and the U.S. economy have diversified over the years, to preserve the continuity of the series, the data for the NAPM Report On Business(R) have continued to be collected only from purchasing executives in manufacturing companies. In July 1997, NAPM began a separate survey of non-manufacturing industries and in June 1998 began monthly release of survey results. To differentiate the two surveys, the survey of manufacturing industries is now known as the Manufacturing NAPM Report On Business(R). Although manufacturing accounts for only 18 percent of GDP (Cahn 1994), it is estimated to account for as much as 33 percent of total business activity (Hoagland 1993). Also, the manufacturing part of GDP has been found to be more volatile than other sectors of the economy in relation to changes in total GDP. Harris (1991) reported that manufacturing accounted for two-thir ds of the variation of GNP around its trend. Another study found that as economic activity in general (as measured by total GDP) changes, manufacturing GDP tends to change by a factor exceeding 2.1 (Cahn 1994). This means that when the growth rate of total GDP has increased or decreased by 1 percent, the growth rate of the manufacturing part of GDP has historically increased or decreased by 2.1 percent. This is perhaps indicative of the "driver" effect of manufacturing (as a user of products and services from other sectors) on the rest of the economy. It also indicates that manufacturing should be a more sensitive indicator of what is happening in the overall economy than more general indicators would be. This also helps to explain why NAPM manufacturing data can be used as an indicator of change in the overall economy in addition to being an indicator of change in the manufacturing sector.
Interest in economic indicators comes from the desire to determine what is happening or likely to happen in the economy. Looking at one or a few statistic allows one to quickly assess economic activity. Because of the statistical nature of indicators, the following questions should be asked about each indicator: * Is the indicator valid -- does it represent or measure what it is intended to measure?
* Is the indicator reliable -- is it free from error and does it yield consistent results (Peter 1979)?
* What is the timing of the indicator -- does it lead, lag, or coincide with the activity it represents?
* Is the indicator timely -- is it available close to the time the activity it is measuring takes place?
* Is the indicator relatively stable or does it exhibit many random fluctuations?
* What information is provided by the indicator -- change, rate of change, level of activity?
* Is the indicator complete when released or is it subject to revisions?
Indicator Qualities of NAPM Data
Validity and Reliability. Unlike measurement standards that can be used to determine the validity and reliability of a yard stick, there are generally no "standards" of economic activity. In many cases, the data collected by the agencies of the U.S. federal government are used as standards for U.S. economic activity. These data are generally well-defined and collected using either sampling or a census approach. The series are often revised several times after initial issuance to improve their accuracy. The government series thus became de facto standards of validity and reliability for other series that attempt to measure the same or similar activity. 
Using government data as a comparison standard, the high correlation of NAPM indexes with government data that measure similar activity supports the validity and reliability of the NAPM data. For example, the correlation of the NAPM Production Index with the government index for Industrial Production in Manufacturing has averaged over 0.87 for the period 1993 to 1996 (McKittrick 1994, 1995, 1996, 1997). Additional examples are discussed later in this article.
Timing. The timing relationships of NAPM indexes can best be illustrated by examining the nature of the indexes. All the NAPM indexes are indicators of month-to-month change and are called "diffusion indexes." Diffusion indexes indicate the degree to which the indicated change is dispersed or "diffused" throughout the sample population (Valentine and Ellis 1991). Respondents to NAPM surveys indicate each month whether particular activities (e.g., new orders) for their companies have increased, decreased, or remain unchanged from the previous month. The NAPM indexes are calculated by taking the percentage of respondents that report that the activity has increased and adding It to one-half of the percentage that report the activity has not changed. This results in an index for which the value 50 indicates no change in the activity, 100 indicates all respondents are reporting increased activity, and 0 indicates that all respondents are reporting decreased activity. This method was adopted by NAPM because it par allels that used by the U.S. Department of Commerce for a number of diffusion indexes compiled from government surveys. 
When a diffusion index is rising in its increase range (50 to 100 for NAPM indexes), increased activity is becoming more dispersed and it implies that the activity is increasing at an increasing rate (Bowers 1985). The opposite is true when a diffusion index is decreasing in its decrease range (50 to 0 for NAPM indexes). Diffusion indexes tend to be leading measures of activity level because typically the rate of change of an activity will change direction before the level of the activity changes direction (Bowers 1985). For example, if the level of an activity is increasing over time, as it approaches a peak, the rate of increase of the activity will typically peak and decrease before the level itself peaks and begins to decrease. [Kacapyr (1996) presents a discussion of the use of diffusion indexes for forecasting.]
The leading nature of the NAPM indexes was first recognized in the late 1950s and has been documented numerous times since then (Hoagland 1959, 1964, 1969; Hoagland and Taylor 1987; Bretz 1990). It is also noteworthy that the NAPM Supplier Deliveries Index has been a component of The Conference Board Index of Leading Indicators (formerly issued by the U. S. Department of Commerce) for many years.
It is important to note that the NAPM indexes do not indicate the volume or level of the activity represented. The indexes only represent the change in the volume or level.
A good leading indicator will consistently turn, or reverse direction, ahead of a reversal of direction in the activity it represents. The ability of NAPM indexes to do this has been researched and documented (Niemira 1991).
Timeliness. An indicator that is unavailable for months after the period for which it is indicating is not very useful. NAPM indicators are among the most timely available. They are released on the first business day after the end of the month for which they are indicating.
Stability. Another desirable characteristic of an indicator is the lack of random fluctuations or a small random component compared to the trend and cyclical components of the activity being measured. This quality of NAPM indexes has also been documented (Hoagland and Taylor 1987).
Revisions. Finally, NAPM indexes are never revised except for small annual revisions in seasonal adjustment factors. Once the monthly data are released, no revisions are ever made to the raw data.
REVIEW AND ANALYSIS
The results in this section and in the accompanying tables referred to as "current study" are from an analysis that was undertaken to: (1) examine the relationships among the NAPM indexes that relate to overall economic activity and business cycles and GNP, and (2) examine lead/lag relationships among the overall NAPM indexes (including the NAPM Price index). For this reason, the NAPM indexes for New Export Orders and Imports were not tested. The NAPM index for Backlog of Orders was not included because it has not been in existence through an entire business cycle.
To conduct the analysis, correlation and linear regression techniques were used to determine the linear association between the indexes and to determine the lead/lag relationships. For the NAPM data, 12-month centered moving averages of indexes were used. The indexes were obtained by taking the percent increase minus the percent decrease for each month. For correlation and lead/lag comparisons to GNP data, the NAPM data were averaged to convert from monthly to quarterly data.
Relationships Between NAPM Indexes and Turning Paints in U. S. Economic Activity
Turning points are defined as the time when an activity that has been increasing starts to decrease (referred to as a "peak") or where one that has been decreasing starts to increase (referred to as a "trough"). Virtually all the studies that have looked at the relationship between NAPM data and turning points in economic activity have found that the NAPM data leads peaks and troughs in the measured activity, often by many months. Table I summarizes the results of studies concerning the behavior of NAPM indexes at business cycle turning points, as established by the National Bureau of Economic Research. Previous research found that for the NAPM general series studied, the range of average leadtimes for economic peaks was 8.0 to 14.3 months, and for troughs, 2.0 to 5.0 months. Also as indicated in Table I, the current study generally confirms the results of the previous research.
Relationships Between NAPM Indexes and Changes in General Economic Activity
Correlation analysis from previous research indicates that a strong relationship (R in the range of 0.76 to 0.91) exists between the PMI and changes in Real GNP and GDP. These results and the results of the current study are presented in Table II.
Previous research on lead/lag relationships between the PMI and measures of general economic activity found that the PMI lags changes in both GNP and industrial production. The results from this research and the current study results are shown in Table III. In the current study, It was found that the PMI lags changes in GNP by two months but that the NAPM Production and New Orders components of the PMI both coincide with GNP. Thus, as shown in Table III, it appears that in general NAPM indexes appear to coincide with or lag GNP changes.
However, as also shown in Table III, the current study found that the Forecasting Index, computed as the NAPM New Orders index minus the NAPM Inventories index (Hoagland and Taylor 1987), leads changes in real GNP by three months. 
Several other researchers have also noted a confirmatory (as opposed to leading) relationship of NAPM indexes to the manufacturing and general economy (Harris 1991, Niemira 1991). However, the NAPM indexes lead the business cycle at most turning points as shown in Table I and have the advantage of early availability. Government GDP statistics are computed only quarterly, are not released until about one month after the end of the quarter, and are subject to later revision. The early availability of NAPM data and the strong relationship between the PMI and GDP enables one to infer, as early as the beginning of the second month of a quarter, what GDP may have done in the first month of a quarter.
The relationship between annual changes in the PMI and in real GDP was estimated as follows by McKittrick:
Real GDP = -15.1534 + 0.3476 X PMI (percent change 4 Q to 4 Q) (Annual Average)
Applying this relationship to a recent value of the PMI (49.4 percent for August 1998) yields 2.02 percent as a rough estimate of annual rate of change of real GDP based on the August PMI. Of course, the assumption that a monthly value of the PMI will approximate the annual average value may not be valid. Therefore, this approach gives only a rough estimate of the change in GDP.
Relationships of NAPM Indexes with Changes in Comparable Indicators of Economic Activity
Correlation. A number of researchers compared NAPM indexes with other measures of the same of similar activity and found a high degree of correlation between the two series. For example, as shown in Table IV, comparisons of changes in the NAPM inventories Index with changes in the Bureau of Economic Analysis' Manufacturing and Trade Inventories resulted in a correlation of 0.69 (Klein and Moore 1988). Similar comparisons between the NAPM New Orders Index and the Department of Commerce Manufacturers' Net New Orders resulted in an average R of 0.71 (McKittrick 1994, 1995, 1996, 1997). Other comparisons have produced similar results (Dasgupta and Lahiri 1991; Harris 1991; Hoagland and Taylor 1987; Cahn 1997). With a few exceptions, correlations are greater than 0.67. The generally high correlation between the NAPM indexes and comparable government indicators supports the use of the NAPM indexes for Production, New Orders, Inventories, Employment, and Prices as early or confirmatory indicators.
Two studies, as shown in Table IV, have compared NAPM Supplier Deliveries with the Federal Reserve Board Capacity Utilization Rate and found high degrees of correlation. However, these two indexes are not measuring the same thing. This historical relationship is perhaps due to a common underlying cause.
Finally, the relationships between the NAPM indexes for New Export Orders and Imports and their counterpart indicators were not as strong as for most of the other comparisons (Niemira and Hormozi 1994).
Previous research has found that in many, but not all, cases the NAPM data series lead changes in comparable indexes. The results are summarized in Table V. One study in particular found that NAPM New Orders leads changes in Manufacturers New Orders by an average of five months at peaks but lags by an average of one month at troughs; NAPM Inventories leads changes in Manufacturing and Trade Inventories by an average of five months at peaks but the two series coincide at troughs; and NAPM Prices leads peaks in the Consumer Price Index by an average of eight months and trough by seven months but that it lags peaks in Industrial Materials Prices by three months and troughs by two months (Klein and Moore 1988).
Lead/Lag Relationships Between the Principal NAPM Indexes
Most research on this topic has been concerned with the relationships between NAPM Indexes and government or other indicators. One study looked at the lead/lag relationships among some of the NAPM indexes (Raedels 1991). This study found that the PMI lags both NAPM New Orders and an index consisting of NAPM New Orders + NAPM Supplier Deliveries by one month. (See Table VI.)
The results of the current study, also shown in Table VI, found the PMI to lag both the NAPM New Orders and Production Indexes by two months. The current study also found that the PMI coincides with Supplier Deliveries and that it leads Employment by one month, Inventories by three months, and Prices by three months.
The Forecasting Index provides an earlier indication of possible turning points than is possible with the published NAPM Indexes. When the Forecasting Index is negative (Inventories index greater than the New Orders index), business is likely to decline and when it is positive (New Orders index greater than the Inventories index), business is likely to improve (Hoagland and Taylor 1987). As shown in Table VI, the Forecasting Index was found in the current study to lead the PMI by six months while Raedels (1991) found a seven-month lead using a different time period for analysis. This index appears useful in forecasting the future direction of the PMI and the direction of the general economy. However, Raedels, after evaluating the Forecasting Index and other indexes as predictors of the PMI, concluded that for looking ahead only one or two months, the NAPM New Orders Index is the most consistently reliable indicator of the PMI.
IMPLICATIONS FOR PURCHASING STRATEGY DEVELOPMENT
This section will briefly discuss the application of NAPM data to development of purchasing strategy. Three references for the use of NAPM Report On Business(r) data in purchasing planning and strategy development are Bretz (1990b), Kauffman (1994), and the NAPM CDROM, Using the NAPM Report On Business(r) to Forecast Purchase Trends.
Purchasing strategy has been characterized as including three challenges or types of issues: (1) Interpretation and support of corporate objectives, (2) Utilization of purchasing and supply functions and issues in corporate strategy, and (3) Selection of specific strategies to be employed (Ellram and Carr 1994; Leenders and Fearon 1997). NAPM index data on business conditions appear to have applications in all three areas.
In the area of interpretation and support of corporate objectives, purchasers can use NAPM data to assess potential effects of the current and near future economic climate on achievement of corporate objectives. For example, consider a situation where a corporate objective of a manufacturer is to reduce employment costs and the NAPM Employment Index and other indicators are reporting that manufacturing employment is increasing at a rapid rate and the current labor market is tight. In this situation, offering lower salaries to new hires would likely result in a failure to hire qualified personnel.
Thus, the NAPM data could be used to justify changes in the corporate strategy. This example also illustrates the importance of using NAPM data with other data when assessing the economic climate. The NAPM Employment Index may be indicating that manufacturing firms are hiring people at a rapid rate but other data such as unemployment rates and information specific to one's own industry are needed to fully assess the importance of that indication to an individual firm.
In the area of the utilization of purchasing and supply functions and issues In corporate strategy, purchasers can use NAPM data to assess general economic conditions and their potential impact on particular purchase markets. For example, consider a manufacturing company in a rapidly growing industry that needs additional production capacity to meet expected demand. The firm might add capacity by temporary means, such as additional shifts or subcontracting, or by permanent means, such as building additional factory space. The NAPM Purchasing Manager's Index, Production Index, and New Orders Index and other data on general economic conditions can be helpful in assessing whether rapid growth will continue (suggesting a permanent capacity increase) or not (suggesting a temporary capacity increase). Again, it is important to use other data in addition to the NAPM data to get as broad a perspective as possible.
In the area of selection of specific purchasing strategies to be employed, purchasers can use NAPM data to assess the economic climate. For example, consider a manufacturing company that is trying to determine a pricing strategy for a major raw material component. They are considering whether to lock in a price long-term or to make only a short-term price commitment. The NAPM Price Index along with other price information such as the federal government's Producer Price Index can indicate recent history and current trends in industrial prices. If the trends are up and general economic conditions are strong as indicated by the Purchasing Managers' Index and other general economic indicators, perhaps a lock-in strategy is best. On the other hand, if trends are down and economic conditions appear to be weakening, perhaps a short-term price would be best.
Research has shown that the NAPM indexes possess desirable indicator qualities and are useful indicators of what is happening in the manufacturing and general economies. Almost all of the results of prior research and the current study indicate that:
* NAPM indexes, as indicators of manufacturing business activity, are more representative than the manufacturing part of GDP.
* Because manufacturing activity is more sensitive to general economic conditions than overall measures, NAPM indexes are good indicators of change in general economic conditions.
* NAPM indexes have advantages of timeliness, high proportions of trend and cyclical components, non-revision of data, and leading indicator and single index number properties of diffusion indexes.
* NAPM indexes on average generally lead business cycle turning points and with a greater lead at peaks than at troughs.
* NAPM PMI, Production, and New Orders Indexes correlate well with general economic indicators.
* NAPM indexes, while they often lead at turning points, usually coincide or lag general economic indicators, but have the advantage of earlier availability.
* NAPM indexes on average lead most of the peaks and troughs of most comparable indicator series.
* The timing of NAPM indexes increases from the Production and New Orders indexes, which generally coincide with general economic activity, to the PMI, and Supplier Deliveries indexes, which lag by two months, Employment index, which lags three months, and Prices and Inventories indexes, which lag five months.
It should be emphasized that the NAPM indexes should be used along with other indicators of the same or similar activity to obtain as broad an indication as possible.
Ralph C. Kauffman is Assistant Professor Management at the University of Houston-Downtown. He earned his Ph.D. degree from the University of Texas at Dallas. Dr. Kauffman's research interests include choice processes in purchasing and supply management, value and application of the NAPM Report On Business(r), and organizational marketing strategy.
The author wishes to thank Dr. John H. Hoagland, Professor Emeritus, Michigan State University, for providing the current analysis data in this article and for comments and suggestions on an earlier version of this article.
(1.) NAPM has recently made the post-World War II data even more accessible by publishing a CD-ROM, Using the NAPM Report On Business(r) to Forecast Purchase Trends, that explains the data and provides examples of how to use them for purchase planning. The CD-ROM also includes a complete set of data from 1948 to 1996 that is transferable to common spreadsheet analysis programs.
(2.) Gross Domestic Product (GDP) is equal to Gross National Product (GNP) less receipts of, and plus payments of, factor income from the rest of the world. Because some of the research included in this article was done using either measure, and the difference between them is small (in 1996, for example, the difference between GNP and GDP was about 0.1 percent), either measure is reported, depending on which was used for the particular research being discussed.
(3.) This is not to say that government measures always exist, are always the best indicators, or are necessarily comparable to NAPM indicators. In some cases, Industry groups collect their own data because government data does not exist or because they feel they can collect better data. Hoagland and Taylor (1987) discuss some of these concerns.
(4.) Examples Include the diffusion indexes of leading, coincident, and lagging Indicators formerly issued by the U.S. Department of Commerce and currently published by The Conference Board. Another equally valid method of computing a diffusion index is to subtract the percent of responses indicating decreased activity from the percent indicating increased activity. In that case, zero becomes the index value that indicates no change, positive index values indicate increased activity, and negative index values indicate decreased activity.
(5.) The coefficient of correlation between the Forecasting Index and GNP was the greatest (0.7425) for a one-quarter (3 month) lead of the Forecasting Index. The coefficient of correlation for a two-quarter (6 month) lead of the Forecasting Index was almost as great (0.7283).
Bowers, D.A., An Introduction to Business Cycles and Forecasting, (Addison-Wesley, 1985)
Bretz, R.J., 1990a, "Behind the Economic Indicators of the NAPM Report on Business(r)," Business Economics, (July 1990), pp. 42-48.
Bretz, R.J., 1990b, "Forecasting with the Report on Business," NAPM Insights (August 1990), pp. 22-25.
Cahn, R.M., "Industry Sensitivity to Economic Activity," CS First Boston Weekly Insights, CS First Boston Corp. (February 7, 1994).
Cahn, R.M., "Analyzing NAPM," eCAHNonomics, Credit Suisse First Boston Corp. (June 3, 1997).
Dasgupta, S. and K. Lahiri, "On the Use of Dispersion Measures from NAPM Surveys in Business Cycle Forecasting," Albany Discussion Papers, Paper No. 91-05, Department of Economics, State University of New York at Albany (July 1991).
Ellram, L.M. and A. Carr, "Strategic Purchasing: A History and Review of the Literature," International Journal of Purchasing and Materials Management, (Spring 1994), pp. 9-18.
Harris, E.S., "Tracking the Economy with the Purchasing Managers' Index," FRBNY Quarterly Review (Autumn 1991), pp. 61-69.
Hoagland, J.H., "Monthly Change Indices and Other New Measures of Business Fluctuations," presented at the American Statistical Association Conferences, Washington, D.C. (December 1959).
Hoagland, J.H., "Purchasing and Inventory Forecasting," presented at the American Statistical Association Conferences, Washington, D.C. (August 1960).
Hoagland, J.H., "Forecasting Purchasing and Business Trends," presented at the 49th Annual International Conference of the National Association of Purchasing Agents, Dallas, Texas (May 1964).
Hoagland, J.H., "Keys to Business Forecasting," presented at the Twelfth Annual International Conference of the American Production and Inventory Control Society, New York, New York (November 1969).
Hoagland, J.H., "Forecasting Business Trends," 78th Annual International Purchasing Conference Proceedings, National Association of Purchasing Management (May 1993).
Hoagland, J.H., and B.E. Taylor, "Purchasing Business Surveys: Uses and Improvements," 72nd Annual International Purchasing Conference Proceedings, National Association of Purchasing Management (May 1987).
Kacapyr, E., Economic Forecasting, The State of the Art, (Armonk, New York: M. E. Sharpe, 1996).
Kauffman, R.G., "Making the NAPM Report on Business(r)" Work for You," NAPM Insights, (October 1994), pp. 62-64.
Klein, P.A. and G.H. Moore, "N.A.P.M. Business Survey Data: Their Value as Leading Indicators," Journal of Purchasing and Materials Management (Winter 1988), pp. 32-39.
Leenders, M.R. and H.E. Fearon, Purchasing and Supply Management, 11th ed. (Irwin, 1997).
McKittrick, G., unpublished analyses of selected NAPM indexes versus other economic Indicators, U. S. Department of Commerce, Office of Economic Conditions (1994, 1995, 1996, 1997).
Niemira, M.P., A Purchasers' Guide to file NAPM Business Survey, research report, Bank of Tokyo-Mitsubishi, New York, New York (October 1991).
Niemira, M.P. and M. Hormozi, Forecasting Foreign Trade Flows With Surveys, unpublished research report, Bank of Tokyo-Mitsubishi, New York, New York (1994).
Peter, J.P., "Reliability: A Review of Psychometric Basics and Recent Marketing Practices," Journal of Marketing Research (February 1979), pp. 6-17.
Raedels, A., "Forecasting the NAPM Purchasing Managers' Index," Journal of Purchasing and Materials Management (Fall 1990), pp. 34-39.
The Conference Board, Inc., Economics Group, 845 Third Avenue, New York, NY 10022.
Torda, T.S., "Early Clue on Turning Points," Business America (June 4, 1985).
Valentine, L.M. and D.F. Ellis, Business Cycles and Forecasting; South-Western (1991).
Wall Street Journal, "Fed's Key Sign? Supplier-Delivery Time" (April 6, 1995), p. A2.
NAPM DATA LEAD/LAG RELATIONSHIPS WITH BUSINESS CYCLE TURNING POINTS [a] Average Lead/Lag NAPM index (months) Study/Study Period Previous Research PMI Leads peaks 14.3, Niemira, (1991) Leads troughs 4.0 6/48-6/91 PMI Leads peaks 2.7, Harris, (1991) Lags troughs 2.0 11/48-7/90 New Orders Leads peaks 16.0, Klein-Moore, (1988) Leads troughs 5.0 1947-1986 Inventories Leads peaks 8.0, Klein-Moore, (1988) Leads troughs 2.0 1947-1986 Supplier Deliveries (percent reporting Leads peaks 10.0, Klein-Moore, (1988) slower deliveries) Leads troughs 5.0 1947-1986 [b] Current Study PMI Leads peaks 14.9, 1/75-9/94 Leads troughs 3.1 Production Leads peaks 15.4, 1/75-9/94 Leads troughs 3.6 New Orders Leads peaks 15.9, 1/75-9/94 Leads troughs 4.0 Inventories Leads peaks 14.1, 1/75-9/94 Leads troughs 1.1 Supplier Deliveries Leads peaks 12.3, 1/75-9/94 Leads troughs 8.7 Employment Leads peaks 15.1, 1/75-9/94 Leads troughs 2.1 (a.)Business cycle turning points as determined by the National Bureau of Economic Research. Note that these are particular dates and not a series of monthly or quarterly data. (b.)Purchasing Association of Chicago series 1947-1975. CORRELATION OF NAPM DATA WITH CHANGES IN U. S. ECONOMIC ACTIVITY Average Number and Indicator of Correlation Author of Studies/ NAPM Index Economic Activity (R) [a] Study Period Previous Research PMI Percent change in 0.84 [b] 2, Bretz, (1990) (annual average) Real GDP, 4 qtr 1953-1987 to 4 qtr PMI Percent change in 0.910 [b] 5, McKittrick, (annual average) Real GDP, 4 qtr (1994-1997) to 4 qtr 1976-1996 PMI Percent change in 0.76 [b] 1, Dasgupta-Lahiri, Real GNP, 4 qtr (1991) 1953-1986 to 4 qtr Current Study [b] PMI Quarterly percent 0.878 3/70-6/92 change in Real GNP Production Quarterly percent 0.900 3/70-6/92 change in Real GNP New Orders Quarterly percent 0.888 3/70-6/92 change in Real GNP Forecasting Quarterly percent 0.742 3/70-6/92 Index [c] change in Real GNP (a.)Approx. R calculated from adjusted [R.sup.2] (b.)GNP data from U. S. Business Cycle Indicator series A0Q050. (c.)NAPM New Orders minus NAPM Inventories (Hoagland and Taylor 1987). NAPM INDEX LEAD/LAG RELATIONSHIP WITH CHANGES IN GENERAL ECONOMIC ACTIVITY NAPM Lead/Lag NAPM Index Economic Indicator (months) Previous Studies PMI [a] Industrial Production Lags 2 (year-over-year changes) PMI [a] GNP Lags 3 (year-over-year changes) Current Study [b] Production Percent change in Real GNP Coincides New Orders Percent change in Real GNP Coincides PMI Percent change in Real GNP Lags 2 [c] Supplier Deliveries Percent change in Real GNP Lags 2 [c] Employment Percent change in Real GNP Lags 3 [c] Prices Percent change in Real GNP Lags 5 [c] Inventories Percent change in Real GNP Lags 5 [c] Forecasting Index [c] Percent change in Real GNP Leads 3 [c,d] (a.)Niemira, 1991, period studied: 1948-1991. (b.)Data period studied: 3/70-6/92. GNP data from U. S. Business Cycle Indicator Series A0Q050. (c.)Lag form GNP derived from the relationship between these indexes and the production and new orders indexes which coincide with GNP. (d.)NAPM New Orders minus NAPM Inventories (Hoagland and Taylor, 1987). Also refer to endnote 5. CORRELATION OF NAPM DATA WITH CHANGES IN COMPARABLE INDICATORS NAPM Index Correlated Index Previous Research PMI BCD CCI [a] PMI BCD CLI [c] PMI Comp. -4 [d] Production FRB [e] Industrial Production Production Industrial Production in Manufacturing (quarterly percent change) New Orders Manufacturers New Orders New Orders Manufacturers New Orders (deflated, quarterly percent change) New Orders Manufacturers Net New Orders (deflated, quarterly percent change) Inventories Smoothed Change in Manufacturing and Trade Inventories Inventories Change in Manufacturers Total Inventories Inventories Manufacturers Inventories (deflated, quarterly percent change) Inventories " " Supplier Deliveries Capacity Utilization, primary process manufacturing Supplier Deliveries FRB [e] Capacity Utilization Employment Manufacturing Employment (quarterly percent change) Employment " " Employment " " Prices BLS [f] Manufacturing Price Index (quarterly percent change) Prices PPI [g] for Intermediate Materials except Food and Energy (quarterly percent change) New Export Orders DOC [h] Real Merchandise Exports (quarterly percent change) New Export Orders Merchandise Exports (percent change from 4 qtrs. ago) Imports DOC Real Merchandise Imports (quarterly percent change) Average Number Correlation of NAPM Index (R) Studies Previous Research PMI 0.78 [b] 1 [j] PMI 0.28 [b] 1 [j] PMI 0.91 1 [k] Production 0.842 [b] 1 [l] Production 0.877 [b] 4 [m] New Orders 0.688 [b] 1 [l] New Orders 0.71 [b] 4 [m] New Orders 0.73 1 [n] Inventories 0.69 1 [o] Inventories 0.716 [b] 1 [l] Inventories 0.67 [c] 4 [m] Inventories 0.75 1 [n] Supplier Deliveries 0.87 1 [m] Supplier Deliveries 0.96 1 [n] Employment 0.96 1 [n] Employment 0.935 [b] 4 [m] Employment 0.867 [b] 1 [l] Prices 0.83 1 [n] Prices 0.894 [b] 3 [m] New Export Orders 0.247 1 [p] New Export Orders 0.82 [b] 1 [m] Imports 0.528 1 [p]
(a.)Business Conditions Digest composite coincident indicator
(b.)Approximate R calculated from adjusted [R.sup.2]
(c.)Business Conditions Digest composite leading indicator
(d.)Composite Index of 4 Roughly Coincident Indicators
(e.)Federal Reserve Board
(f.)Bureau of Labor Statistics
(g.)Producer Price Index
(h.)U. S. Department of Commerce
(j.)Dasgupta and Lahiri, 1991
(o.)Klein and Moore, 1988
(p.)Niemira and Hormozi, 1994
LEAD/LAG RELATIONSHIPS OF NAPM DATA WITH COMPARABLE INDICATORS Average NAPM Index Lead/Lag NAPM Index Compared Indicator (Months) Previous Research New Orders [a] Diffusion Index of Manufacturers New Orders Leads peaks 5.0 New Orders [a] " " Lags troughs 1.0 lnventories [a] Smoothed Change in Manufacturing and Trade Inventories (constant $) Leads peaks 5.0 Inventories [a] " " Coincides troughs Supplier Deliveries [a] Ratio of Manufacturing and Trade Sales to Inventories (constant $) Lags peaks 1.0 Supplier Deliveries [a] " " Leads troughs 3.0 Prices [a] Consumer Price Index (6 month smoothed change) Leads peaks 8.0 Prices [a] " " Leads troughs 7.0 Prices [a] Journal of Commerce Index of Industrial Materials Prices Lags peaks 3.0 Prices [a] " " Lags troughs 2.0 New Export Orders [b] Real Merchandise Exports (Dept. of Commerce) Leads series 3.0 lmports [b] Real Merchandise Imports (Dept. of Commerce) Leads series 2.0 (a.)Klein and Moore, 1988 (b.)Niemira and Hormozi, 1994 LEAD/LAG RELATIONSIP AMONG NAPM INDEXES NAMP Index NAPM Index Lead/Lag of Compared Compared to NAPM Index (months) Previous Research PMI [a] New Orders Lags 1.0 PMI [a] New Orders + Supplier Deliveries Lags 1.0 PMI [a] Forecasting Index [b] Lags 7.0 Current Study PMI New Orders Lags 2.0 PMI Production Lags 2.0 PMI Supplier Deliveries Coincides PMI Employment Leads 1.0 PMI Inventories Leads 3.0 PMI Prices Leads 3.0 PMI Forecasting lndex [b] Lags 3.0 [c] (a.)Raedels, 1991 (b.)NAPM New Orders minus NAPM Inventories (Hoagland and Taylor, 1987) (c.)Refer to endnote 5
|Printer friendly Cite/link Email Feedback|
|Author:||Kauffman, Ralph G.|
|Publication:||Journal of Supply Chain Management|
|Date:||Mar 22, 1999|
|Previous Article:||Measuring Contractors' Performance.|
|Next Article:||The Benefits of Multiple Boundary Spanning Roles in Purchasing.|
|TAKING STOCK OF DOW DOWNTURN : REPORTS CONFIRM STRONG ECONOMY; RATE HIKE POSSIBLE.|
|Let the numbers do the spinning. (At Large in the Shop).|