Forecasting

Forecasting

Article Type: Abstracts From: Strategic Direction, Volume 24, Issue 5.

Fildes R., Goodwin P. Financial Management (UK), December 2007-January 2008, Start page: 42, No. of pages: 2

Purpose To investigate why managers may override their software’s forecasts; and the effect on accuracy. Design/methodology/approach Examines 60,000 forecasts made by four UK businesses, interviews forecasters and observes forecast review meetings. Findings Three-quarters of the forecasts were changed to some extent although this varied between firms. Adjustments sometimes did more to foster an illusion of control than to improve accuracy. On average forecast errors were reduced by 3.6 percent but adjustments at the retail firm increased error. Large changes were more likely than small ones to increase accuracy and negative changes more likely to be right than positive ones (over a third of positive changes were in the wrong direction). Adjustments were often based on recent data and ignored longer-term trends. Research limitations/implications None stated. Practical implications Implies that firms should clarify what they are forecasting, restrict forecasters to making larger adjustments with documented reasons for doing so and use long term data. Originality/value Shows that some forecasting work is a hindrance rather than a help to planning.ISSN: 1471-9185Reference: 37AC407

Keywords: Accounting research, Accuracy, Computer software, Forecasting, Managers