Evidence for the Ability of the Regression Model and Particle Swarm Optimization Algorithm in Predicting Future Cash Flows

This study predicts future cash flows using a regression model and a particle swarm optimization algorithm (PSO). The variables of accruals components and operating cash flows were used, and the data of 137 listed companies on the Tehran Stock Exchange during 2009-2017 were studied. Eviews9 software for the regression model and Matlab13 software for the Particle swarm optimization algorithm was used to test the hypotheses. The results indicate that the regression model's variables and the Particle swarm optimization Algorithm in this study can predict future cash flows. Furthermore, the results of the fitting Particle swarm optimization Algorithm show that a structure with eight hidden neurons is the best model for predicting future cash flows, and the proposed neural network model compared with the regression model has higher prediction accuracy in predicting future cash flows. This study shows that the classification of assets and liabilities provides useful information from future operating cash flows.