The Effect of Behavioral Factors on Stock Price Prediction using Generalized Regression and Backpropagation Neural Networks Models

2. Literature Review

As mentioned above behavioral finance tries to explain how unreasonable reactions can be influential in markets. Shefrin (2000) organized psychological phenomena pervading the landscape of finance into three themes: 1) heuristic-driven bias; 2) frame dependence; and 3) inefficient markets. A heuristic is the process by which people seek things out for themselves, usually by trial and error; while trial and error can develop into general rules of thumb, it can also culminate in further errors (Shefrin, 2000). In financial markets, these biases include representativeness, overreaction, underreaction and overconfidence, as well as availability bias, anchoring-and adjustment, aversion to ambiguity and gambler's fallacy, and stock market prediction. The second theme of behavioral finance presents, frame dependence, distinguishes between form and substance in decision-making. It reflects a mix of cognitive and emotional elements. The third one is about common question in finance, whether the market is efficient or inefficient (Lee, 2009). This study focuses on four biases: representativeness, overreaction, underreaction and overconfidence. We want to evaluate the impact of thesis biases on prediction accuracy.