| In 2015,Chinese stock market experienced abnormal fluctuations,with frequent crashes,thousands of shares fell by the daily limit,or even all stocks fell by the daily limit.The allocation of resources in the financial market almost failed,which had a huge impact on the entire economic system.At present,the development of Chinese stock market is not mature,and the regulatory system needs to be improved.Scholars usually focus on the question of extreme risk in financial markets,there is little literature analysis our country financial market crash the time-varying characteristics and inner mechanism of the risk,however,scientific measure market collapse risk and predict the collapse events not only helps to prevent the occurrence of financial crisis,at the same time help provides beneficial reference to markets such as investment decision-making practice problems.Therefore,this paper intends to analyze and study the bubble and crash risk of Chinese stock market based on the Log-Period Power Low(LPPL)model.Specifically,This paper based on the Shanghai composite index and the CSI 500 index as the research object,using the genetic algorithm instead of the Tabu search algorithm to estimate the parameters,using relative error analysis,unit root test and parameter sensitivity test analysis model fitting effect and prediction ability,and the reliability of genetic algorithm,and further more the market shares and the predicted results of small and medium-sized market value of company stock whether there is a statistically significant difference.The empirical results show that LPPL model can effectively fit the stock market bubble and predict the time point of bubble bursting(collapse).The genetic algorithm used in this paper is more reliable than the original search algorithm.The estimates of the collapse points of the Shanghai composite index and the China securities 500 index are close,so there is no significant difference in the sequence of the two crashes in Chinese stock market. |
PDF Full Text Request