Quantitative and qualitative approaches for stock movement prediction

Knowing about future values and trend of stock market has attracted a lot of attention by researchers, investors, financial experts and brokers. The benefits of having a good estimation of the stock market behavior are well-known, minimizing the risk of the investment and maximizing the profits. In recent years, mathematical and computational models from artificial intelligence have been used for such purpose. This research studies quantitative and qualitative modeling approaches to forecast four indices of the Indian stock market. As quantitative methodologies we use time delay feed forward neural networks, auto regressive integrated moving average and their combination. As a qualitative counterpart we use the fuzzy inductive reasoning methodology. 10-fold cross validation has been used to evaluate the generalization capacity of each predictive model developed. The best results are obtained with the time delay feed forward neural networks models and the worst with the fuzzy inductive reasoning models. No significant enhancement is obtained with the approaches proposed when compared with the simple random walk method.