Agent Application in the Stock Market

Abstract

The stock market is a key market in any economy and financial forecast such as stock price prediction is a field receiving much attention both for research studies and commercial applications. Stock market forecasters are keen on developing a successful approach to predict
stock prices even more accurately since there is motivation of gaining massive profits from trading shares by using well defined attractive strategies. This research project develops a stock price prediction model built using JADE environment. It is based on multi-agent architecture in order to harness the power of agents. This provides investors with predicted trend of share price by incorporating various correlated factors like economic, political, company outlook to traditional price over time, demand and supply in order to accurately forecast the stock price trend and thus, provide a buying or selling signal to traders. The Trend is determined by incorporating text processing in agents from live news sources. The model was tested and has proved to be a key tool for stockbrokers, novice traders and investment bankers since it is automated and more robust than traditional methods of price prediction.