Predicting Stock Returns

Abstract

In this article, we use the artificial neural network in the forecasting of daily Bombay Stock Exchange (BSE) Sensitive Index (Sensex) returns. We compare the performance of the neural network with performances of random walk and linear autoregressive models by using six performance measures. The major findings are that neural network out-performs linear autoregressive and random walk models by all performance measures in both in-sample and out-of-sample forecasting of daily BSE Sensex returns. The findings suggest that stock markets do not follow a random walk and there exists a possibility of predicting stock returns. The superiority of the neural network model over linear autoregressive and random walk models in forecasting daily BSE Sensex returns indicates that neural network is able to capture non-linearities contained in stock returns.