Neural Network Forecasts of Stock Return Using Accounting Ratios

The aim of this paper is how to predict stock return by using accounting ratios and also by using the procedure of neural network. This paper has considered the prediction of stock return by using accounting ratios with two procedures, the artificial neural network and least square regression. The independent variables in this paper are accounting ratios and dependent variable of stock return, thus; the accounting ratios have collected for 8 years in the industries of cement and medicine. The hypothesis of the paper has two hypothesizes, that shows the survey ability of neural network procedure in predicting stock return in comparison with least square regression in the level of two active firms in the two industries