Study of Effects of Financial Information on Stock Price Momentum in Winning and Losing Portfolios Using Data Mining Methods: Neural Networks and Decision Trees

 This study addressed the effects of financial information on stock price momentum in winning and losing portfolios using data mining methods (neural network and decision trees) in firms listed in Tehran Stock Exchange. The financial analysis included three general categories: profitability variables (return on assets, cash flows, accruals, and profitability growth), leverage and liquidity variables (financial leverage, liquidity ratio, and share issue), and operating efficiency variables (profit margin and assets turnover). The study sample comprised firms listed in Tehran Stock Exchange in the 2008-2019 time period. The results of the data mining model showed that the profitability variables had the greatest effect on the stock future returns in both winning and losing portfolios. This result has an important implication for investors in stock exchanges: when selecting a share, the investor is advised to choose a firm with high profitability on the basis of simple indices the present study introduces.