Development of a stock trading system based on a neural network using highly volatile stock price patterns

This paper proposes a pattern-based stock trading system using ANN-based deep learning and utilizing the results to analyze and forecast highly volatile stock price patterns. Three highly volatile price patterns containing at least a record of the price hitting the daily ceiling in the recent trading days are defined. The implications of each pattern are briefly analyzed using chart examples. The training of the neural network was conducted with stock data filtered in three patterns and trading signals were generated using the prediction results of those neural networks. Using data from the KOSPI and KOSDAQ markets, It was found that that the proposed pattern-based trading system can achieve better trading performances than domestic and overseas stock indices. The significance of this study is the development of a stock price prediction model that exceeds the market index to help overcome the continued freezing of interest rates in Korea. Also, the results of this study can help investors who fail to invest in stocks due to the information gap.