Design of a Financial Decision Support System based on Artificial Neural Networks for Stock Price Prediction

Stock markets are highly volatile by nature and difficult to predict due to the non-linear and complex nature of the market. A system that can forecast and predict the stock prices is of great value to individual investors who do not have sufficient knowledge to understand the complex dynamics involved in evaluating and predicting stock prices. Machine learning focuses on the development of computer programs that can access data and use it to learn for themselves. Machine learning is widely being used in the financial domain including prediction of stock prices. Based on the extensive literature review in this domain, traditional methods of using Machine Learning techniques including Artificial Neural Networks (ANN) for stock price prediction have taken in to account only the Technical Features. The current machine learning models do not take in to account the Intrinsic or fundamental features of the stock; the results of such prediction models are not accurate and at best could predict an intraday price of stocks with high levels of Variance. Literature review in the domain of stock predictions has shown that future stock prices are seldom dependent on the past performance and technical indicators and they invariably depend on the fundamental value and macro-economic factors.In this paper, we propose development of an Artificial Intelligence based decision support system (DSS) for guiding individual investors to buy and sell stocks. The Financial decision support shall be based on mathematical modeling of the various financial parameters to predict stock prices on a long term basis with a reasonable degree of accuracy and eliminate the behavioral biases of human decisions. The ANNs in this study were trained using open source financial data of select stocks listed on the BSE/NSE. The results of this study are quite encouraging as the stock prices can be predicted at least one month in advance and are closer to the real-time market prices. This DSS has the potential to help millions of Individual Investors who can make their financial decisions on stocks using this system for a fraction of cost paid to corporate financial consultants and value eventually may contribute to a more efficient financial system.