Designing a Model for Predicting the Probability of Stock Prices Crash in the Stock Market: Artificial Neural Network Approach

The probability of stock prices crash has great importance in portfolio analysis and pricing of capital assets. Therefore, one of the major issues that investors face in the capital markets is predicting the fall of stocks. Given this necessity, the purpose of this study is to provide an approach to estimate the risk of stock price crashes. Recently, methods called "artificial neural networks" have been used to predict monetary and financial variables in parallel with structural models and time series. These models, which are actually derived from the brain learning process, use computer computational speed to learn complex relationships between variables and use them to predict the future. Using the data of 20 companies listed in the Tehran Stock Exchange, the present study presents models to estimate the probability of stock prices crash in the Iranian stock market using artificial neural networks. The results indicate that artificial neural networks have good performance in estimating the probability of stock prices crash in the Iranian stock market.