An Investigation of Performances of Technical Indicators in the Stock Exchanges Using Hybrid Methods of SRA, PCA and Neural Networks

Prediction price indicator movement has always been a challenging task in the exploitation of time series for forecasting. Exact prediction of price indicator movement may offer numerous privileges for investors. As a result of the complexity of stock market data, development of efficient models is often not simple. This research have combined a number of methods namely as Principal Component Analysis (PCA), Stepwise Regression Analysis (SRA) and Artificial Neural Networks (ANN) by technical analysis tools of financial markets. In proceeding, the efficiency of each set in predicting the indicator trend of stocks’ total price, have been compared. Data used in this research have been collected from cable companies in the stock exchange between ٢٠٠٧ and ٢٠١٣. Using empirical results, this research introduces an efficient set of technical indicators for forecasting total price indicator movement in cable companies in TSE. Other results of this research indicate more accuracy of SRA and neural networks in comparison with PCA and neural networks. This paper investigates the efficiency of indicators in Tehran’s Stock Exchange (TSE

کلیدواژه ها:

Technical indicators ، Stepwise regression analysis ، Principal component analysis ، Neural networks