Research On Stock Market Risk Measurement Based On Extreme Value Theory

After decades of development,our country stock market has been expanding in size,but at the same time,the risks it faces are becoming more and more diverse and complex.How to measure risk scientifically and effectively is the focus of many investors,and it is also the core issue faced by financial regulators in building a risk management system.Therefore,it is of great practical significance to accurately measure the risk of the stock market to maintain the interests of investors and promote the stability of the stock market.At present,the VaR method,as a mainstream risk measurement method,is valued by the financial community.Compared with traditional risk management methods,it can quantify risks more scientifically and accurately,and has higher use value.Therefore,it is of great theoretical significance to use the VaR method to study the risk status of the stock market and to realize effective risk management.The article reviews the current research status of risk measurement both domestically and internationally,and finds that in the face of the volatile stock market,accurately measuring stock market risk has become a key issue in financial risk management.Therefore,the paper selects the daily closing price of the Hang Seng Index for nearly 14 years from November 10,2008 to November 8,2022 as the basic data,and R software was used for data processing to obtain a logarithmic rate of return with good statistical characteristics sequence.First of all,in order to make up for the shortcomings caused by the neglect of extreme events in the traditional risk measurement model,the extreme value theory is introduced,and the POT model is constructed to measure the risk of the stock market.It does not need to set the distribution of the model in advance,so there is no risk of model setting,and it can effectively describe the tail distribution of the sequence.Secondly,considering the common volatility aggregation of stock market return series,the EGARCH model is introduced to eliminate heteroscedasticity.Then,in order to further improve the accuracy of the risk measurement model,the thick-tailed characteristics of the data are combined with the volatility aggregation,and the EGARCH-POT model is constructed to estimate the VaR value.Finally,based on the POT model,EGARCH model,and EGARCH-POT model in the extreme value theory,the VaR values at the 95% and 99% confidence levels are estimated to measure the stock market risk.In order to compare the accuracy and effectiveness of the VaR value,the VaR value is tested by the Kupiec failure frequency test method,and the three risk measurement models are analyzed and compared according to the test results.The empirical results show that: compared with the POT model and the EGARCH model,the improved EGARCH-POT model has higher superiority and accuracy in measuring the VaR value at two confidence levels,and can better capture the stock market Risk characteristics have important reference value and certain practical application value for individual investors and financial institutions to control market risk.