A Conceptual Framework for Computing U.S. Non-Manufacturing PMI Indexes
Cho, Danny I., Ogwang, Tomson, Journal of Supply Chain Management
Beginning July of 1997, the Institute for Supply Management[TM] (ISM) (1), formerly known as the National Association of Purchasing Managers (NAPM) in the United States, began to release the Non-Manufacturing ISM Report on Business[R] on a monthly basis. The report is based on data compiled from monthly responses to survey questions asked of more than 370 purchasing/supply executives in over 62 different industries in the non-manufacturing sector covering nine divisions from the Standard Industrial Code classification system, including Agriculture, Forestry and Fisheries; Mining; Construction; Transportation, Communications, Electric, Gas and Sanitary Services; Wholesale Trade; Retail Trade; Finance, Insurance and Real Estate; Services; and Public Administration. These survey responses reflect the change in the current month over the previous one for each of the following 10 business indicators--Business activity, New orders, Employment, Supply deliveries, Inventories, Prices, Backlog of orders, New export orders, Imports and Inventory sentiment.
The ISM has a long and successful history of releasing important business indicators for the manufacturing sector. For over 70 years, it has released the PMI Index and 10 other business indicators in its monthly Manufacturing ISM Report on Business[R]. The PMI, which ranges from 0 to 100, with 50 as the critical reference, is a composite index comprising five seasonally adjusted diffusion indexes (Production, New orders, Employment, Supplier deliveries and Inventories) of economic activity for the manufacturing sector (2): a PMI score greater than 50 indicates growth, while a score below 50 forecasts a sluggish manufacturing economy. The index, which is released at 10:00 a.m. EST on the first business day of each month, is widely recognized by many economists and business practitioners to be a reliable and valuable near-term indicator of economic activity in spite of the decline in the relative importance of manufacturing. The popularity of the PMI partly stems from the earlier time of its release than those of most other monthly economic reports (e.g., Employment and GDP reports (3)). The ability of the PMI to signal changes in economic trends and to provide early clues on the turning points of business cycles has been demonstrated by Harris (1991), Dasgupta and Lahiri (1992, 1993), Kauffman (1999), Koenig (2002) and Lindsey and Pavur (2005). Klein and Moore (1988), Harris (1991) and Kauffman (1999) also discussed the strengths and weaknesses of the PMI and its components as leading indicators. Recent papers by Pelaez (2003a, b) have documented the effects of unexpected changes in the PMI on domestic and international financial markets, which may react partly from expectations that the U.S. Federal Reserve may change its policy stance in response to information about the direction of PMI.
The ISM originally constructed the PMI as an equally weighted index of the five aforementioned diffusion indexes. This equal weighting scheme was abandoned in 1982 in favor of the following weighting scheme: New orders (0.3), Production (0.25), Employment (0.2), Supplier deliveries (0.15) and Inventories (0.1), which is still in use today with no changes for more than 2 decades. The current PMI weighting scheme, which assigns the highest weight of 0.3 to "New orders" and the lowest weight of 0.1 to "Inventories," was chosen so as to maximize the correlation of the PMI with real Gross Domestic Product (GDP)/GNP growth (Torda, 1985; Bretz, 1990; Lindsey and Pavur, 2005). However, a number of empirical researchers have recently begun to question this weighting scheme, leading to the search for alternative weights. For example, Pelaez (2003a) proposed an alternative to ISM's current PMI, which dominates the PMI as an indicator of the growth rates of real GDP and of industrial production. Pelaez's (2003a) weighting scheme proposals were based on regressions of the quarterly growth rate of GDP and the growth rate of the industrial production index on current and/or lagged values of the ISM's five diffusion indexes. …