When quantitative models are used for short-term multi-item sales forecasts it is possible that the managers who use such forecasts may disagree with at least some of the estimates obtained, and wish to change them so that they become more consistent with their own (subjective) evaluation of the marketplace. This study reports on an analysis of the effectiveness of judgemental revision of sales forecasts over six quarterly forecasting periods. The results give general support for the practice of forecast manipulation as a means of improving forecasting accuracy. It is also observed that the effectiveness of revision activity varies across different time periods.
KEY WORDS Forecast revision Judgemental forecasting Accuracy
Empirical study Extrapolation
This study presents a longitudinal analysis and evaluation of the practice of judgementally revising short-term sales forecasts initially produced by a rational quantitative forecasting model (in this case Holt's exponential smoothing method). The approach taken is similar to that used in an earlier paper by Mathews and Diamantopoulos (1986) which examined the same issue but on a single-period, cross-sectional basis only. Prior to the latter paper the only empirical works which addressed the issue of judgemental revision gave little support for a manager who subjectively revised the output of a formal model; on the contrary, results indicated that, at best, there would be no improvement in accuracy and possibly some degradation (Kelly and Fiske, 1950;Harris, 1963;Carbone et al., 1983). In an article aimed at managers, Armstrong (1981) urged that forecasts should be made using quantitative techniques and then accepted by management without modification. Seemingly, the only exponent of 'fine tuning' of forecasts by management has been Makridakis (1981a, b) although he did not provide any empirical support to back up his viewpoint. Within this background, the Mathews and Diamantopoulos (1986) study gave the rather 0277-6693/89/020129-12SO6.00