The Feasibility of Reporting Forecasted Information.

This article suggests a procedure for evaluating the accuracy of forecasted information in accounting and presents the results of a limited empirical investigation of forecasting accuracy. Two aspects are important in the evaluation of forecasting accuracy: (1) differences between the forecasted amount and the actual amount for a particular year and (2) trends in differences over time. An analysis of differences arising in a particular year provides an indication of a firm's ability to meet its goals and would reveal substantial differences that would merit investigation into a management's ability to react to unanticipated events. At the outset of this project, a request to participate in research into forecasting accuracy was extended to more than 50 firms. Interest in the project was evidenced by eighteen firms. Of these, twelve firms provided usable empirical data while the other six firms limited their participation to an interview with a corporate executive. The type of data obtained were forecasts and actual results for various income statement classifications for as many years as was practicable, usually five years. Forecasted balance sheet data were excluded because executives generally stated that such data were very inaccurate or that the firm prepared a forecasted income statement only. Within this article, analysis is limited to forecasts of sales and net income because of their dominant role.