Fundamental Analysis of Cost Behavior by Function and Nature and Its Impacts on Earnings Forecast Accuracy

This study will provide a new approach to cost behavior analysis. Also, by completing past models, examine the behavior of costs when sales decrease (cost stickiness) and sales increase (cost positive signal). In addition, in this study, costs are categorized in terms of function (COGS and SG&A) and nature (payroll, service and depreciation expenses), and examined the impact of them on the earnings forecast accuracy. Fixed effect data model and OLS regression were applied to test the research hypotheses. The sample of the research is 95 companies listed in Tehran Stock Exchange, during 2014-2019. The results show when sales decrease there is a negative and significant relationship between the stickiness of the total cost, COGS and services costs with earnings forecast accuracy. However, no significant relationship was found between the stickiness of operating and depreciation costs and earnings forecast accuracy. Additionally, the results show when sales increase there is a positive and significant relationship between the positive signal of COGS with earnings forecast accuracy. On the other hand, there is no significant relationship between the positive signal of total costs, SG&A, payrolls, services and depreciation costs with earnings forecast accuracy. When sales decrease, the behavior of total costs, COGS and services cause to reduce earnings forecast accuracy and when sales increase, the positive signal of COGS cause to reduce earnings forecast accuracy. Therefore managers and financial analysts need to carefully consider the behavior of them and apply cost stickiness and positive signals in their forecast in order to improve earnings forecast accuracy.