Stock Price Prediction Using Stepwise Regression and Improved with Factor Analysis

Owing to volatility in stock markets, it is quite elusive to forecast stock prices. Albeit, sometimes regular patterns are manifested in stock prices and a variety of factors are proved to be competent to determine stock prices partly. Hence, using stepwise regression on historical stock price data, this paper proposes determining similar patterns in stock prices and exploring potential rules to select the main factors that can affect stock prices significantly while taking all factors into account. Difference analysis is also employed to probe possible correlations in the data. Eventually, this paper tries to improve stock price prediction using factor analysis and manages to achieve higher accuracy.