Fuzzy Decision Mechanism for Stock Market Trading

Abstract

Investors utilize various methods to make buy/sell decisions depending on time-dependent stock market prices. In this study, a fuzzy decision mechanism that makes buy/sell decisions for stock market data is proposed. The proposed mechanism generates instant buy/sell decisions by evaluating three popular indicators which are the Moving Average Convergence/Divergence (MACD) Strategy, Chaikin Money Flow (CMF), and Stochastic Oscillator (SO). The fuzzy decision mechanism has three inputs and one output which are defined by using Gaussian membership functions. In the design of the decision mechanism, Mamdani inference method is used and the rule table is defined by nine rules. Therefore, the structure of the proposed fuzzy decision mechanism is simple and straightforward. The performance of the proposed fuzzy decision mechanism is compared with two classical decision mechanisms using MACD and CMF indicators separately. In the comparisons, the stock market data of Borsa Istanbul 100 Index (XU100), Dow Jones Industrial Average (^DJI), and S&P 500 (^GSPC) are used. The comparison results show that the proposed fuzzy decision mechanism provides significantly higher profit than the mechanisms using either MACD or CMF indicators for all stock market data.