There is uncertainty in all predictions. This fact is evident from an unusual component in the time series. The presence of such an element, which indicates the existence of undefined or unpredictable fluctuations in the data, means that we should expect a prediction error if this unusual component has a similar effect. The capabilities of forecasting have been small, but if this unusual component is limited, we will be able to achieve a prediction that has a higher degree of accuracy by determining the trend of cyclical and seasonal patterns. The companies listed on the Tehran Stock Exchange during the years 1393 to 1397 with a sample of 163 companies. To test the hypothesis, multivariate linear regression and hybrid data were used using the generalized least squares model. The results of hypothesis testing showed that management forecasting error has an inverse and significant relationship with corporate investment efficiency, that is, the more management forecasting error increases, the company's investment efficiency decreases.