The data for the index are collected through a survey of purchasing managers in the manufacturing sector on five different fields, namely, new orders from customers, speed of supplier deliveries, inventories, order backlogs and employment level. Respondents can report either better, same or worse business conditions than previous months. For all these fields the percentage of respondents that reported better conditions than the previous months is calculated. The five percentages are multiplied by a weighting factor (the factors adding to 1) and are added.
Survey panels Purchasing managers form a near ideal survey sample base, having access to information often denied to many other managers. Due to the nature of their job function, it is important that purchasing managers are among the first to know when trading conditions, and therefore company performance, change for the better or worse. S&P Global therefore uses such executives to produce data on business conditions. In each country, a panel of purchasing managers is carefully selected by S&P Global, designed to accurately represent the true structure of the chosen sector of the economy as determined by official data. Generally, value added data are used at two-digit SIC level, with a further breakdown by company size analysis where possible. The survey panels therefore replicate the actual economy in miniature. A weighting system is also incorporated into the survey database that weights each response by company size and the relative importance of the sector in which that company operates. Particular effort is made to achieve monthly survey response rates of around 80%, ensuring that an accurate picture of business conditions is recorded over time. Data are collected in the second half of each month via mail, email, web, fax and phone.
Questionnaires A key feature of the PMI surveys is that they ask only for factual information. They are not surveys of opinions, intentions or expectations and the data therefore represent the closest one can get to “hard data” without asking for actual figures from companies. Questions asked relate to key variables such as output, new orders, prices and employment. Questions take the form of up/down/same replies. For example, “Is your company’s output higher, the same or lower than one month ago?” Respondents are asked to take expected seasonal influences into account when considering their replies. For each main survey question, respondents are asked to provide a reason for any change on the previous month, if known. This assists not only the understanding of variable movement but also in the seasonal process when X12 cannot be used.
Seasonal adjustment The seasonal adjustment of PMI survey data is usually calculated using a combination of the
X12 statistical programme of adjustment, as used by governmental statistical bodies in many developed countries, and alternative methods such as using the frequency of mentions by responding survey panel member companies for changes in variables where that mention is considered indicative of a seasonal variation. ==Other PMI surveys==