This research looks at the inconsistency in accounting literature about the estimation of cost behavior. The main goal was to identify whether the Statistical Analysis of Cost Behavior theory, defended by accounting literature, is infringing modern econometric concepts used in finance, in terms of identifying and correcting the non-stationarity of the temporal series. Therefore, a deductive methodology was used through indirect research. To reinforce the concepts and analyses carried out in the development of this research, case study data were used and processed through the econometric software E-views. Finally, it is concluded that the theory used in accounting literature about the use of statistical techniques in cost behavior does not consider the stationarity analysis of the series and that, therefore, it generates statistical inconsistencies (spurious regressions) in the forecasting of this referred behavior. Thus, this research is relevant in scientific evolution by opening a broader discussion in an incipient area of Accountancy, i.e. econometrics applied to accounting.
Cost Accounting; Econometrics applied to Accounting; Stationarity
