Research On Correlation Risk And Forecast Of Asian Stock Market Portfolio

As integration continues to deepen,individual and institutional investors often choose different types of stocks across countries and regions for portfolio investment to achieve the goal of diversifying risks and achieving higher returns.However,while increasing the diversity of investors' stock market choices,it has also increased the damage caused by extreme events to financial markets and the world economy.How to choose the optimal model to achieve accurate measurement and prediction of portfolio risk in the stock market has become the current focus.Therefore,this paper combines the extreme value theory to study the risk accurate measurement and prediction of Asian stock market portfolio through different risk measurement models.This paper selects five representative stock market indices in Asia as the research object,and analyzes the measurement and prediction of relevant risks.Firstly,according to the analysis results of the correlation features of each component sequence,the EGARCH-POT model method is used to fit their edge distribution.Then,the Copula model with five different structures is used to describe the dependent structure of the five stock markets to measure the investment risk level of the portfolio;Finally,the dynamic VaR is calculated according to the model,and the rigorous return test method is used to compare the predictive ability of different models.The empirical results show that: Firstly,in the relevant risk measurement,the best-performing R-vine model structure shows that the KS11 index has strong correlation with N225 and HIS indexes;while SSE has weaker correlation effects on TWII,N225 and KS11.It shows that China has taken relevant measures to avoid the impact of events such as the Southeast Asian storm on the overall economic situation of the mainland.Third,judging the predictive ability of portfolio return rate,POT-EGARCH-R-vine-Copula on this model is the best for combined risk measurement,both precision and prediction,and can be used for more accurate measurement of stocks.The risk of market portfolios provides a technical reference.On the basis of the previous article,this paper combines Copula theory with extreme value theory,and combines the five models that are currently used the most to compare the horizontal and vertical risks of risk prediction,and compare the existing literature to obtain more accurate stock market measurement methods.It can provide a powerful technical reference for risk supervision and management of the stock market.Secondly,this paper uses the rolling window width return test method,which is superior to the commonly used non-rolling method,which provides more powerful support for the model comparison results and once again proves the more accurate risk prediction.