Prediction Of Collapse Probability And The Cross-section Of Stock Returns

In the post-epidemic era,the global economy is facing greater uncertainties,where Black Swans,Gray Rhinos and other events will exert significant impacts on the economy.In the past years,China economy has been developing rapidly,and its economic scale has been increasing.The Chinese capital market has gradually become a necessary channel to support the financing of the real economy.However,that the stock market has greater volatility as well as the synchronization between traders is stronger makes stock price collapse more frequently compared with foreign developed capital markets.This influences macro-economy and the allocation of investors’ resources to a great extent.Therefore,the research on stock price collapses becomes indispensable.Current research often use negative yield skewness coefficient,fluctuating ratio coefficient and the binary variable of whether stock prices collapse to measure the stock price collapse risk.The risk of stock price collapse is often regarded as the systemic risk to explain its mechanism,and a few of literatures use other innovative mathematical models to describe stock price collapsing events.This paper studies the probability of stock price collapse from the perspective of individual stocks and discovers the relationship between the probability and the expected cross-sectional return rate of stocks.This paper is based on 3243 listed companies,324,605 observed value on the Chinese A-share market’s companies from 2000 to 2020.Based on the window periods of 3 months,6 months and 12 months,this paper predicts the new indexes of stock price collapse probability by using multinomial logit model.Then through using the ROC/AUC model and event study method to inspect the probability index on the accuracy of predicting the collapsing event in share prices.Through portfolio analysis way and FM regression to analysis the relationship between the probability of stock price collapse and the stock return in the future.The conclusions are as follows :(1)The characteristic indexes of market level and company level can better predict the probability of stock price collapse.(2)The event study method found that the cumulative abnormal return rate of stocks increased significantly in a period of time before the event of stock collapse;After the collapse event day,the cumulative abnormal return on stocks declined steadily.(3)It is found that the higher the predicted probability of stock price collapse,the lower the cross-sectional return rate of stock in the next period.After adding common pricing factors as control variables,the results still show a significant negative relationship.Similarly,the probability of stock price collapse predicted under different window period variables has reached the above consistent conclusion,which is not affected by small market value samples and market sentiment.This paper provides a new perspective on the assessment of stock price collapse risk,and innovatively introduces the predicted probability of stock price collapse under different window periods,proving that the index based on the probability of collapse events at the level of individual shares can depict the prediction relationship between stock future returns.The conclusions of this paper provide important advice to investors and listed companies on investment management,which fosters health and stableness of China capital market development.