Research Of Portfolio Risk Based On Copula-GARCH-MCMC Model

The research on portfolios covers correlation studies and risk measurement studies,of which the former is particularly important for the latter.However,the correlation between assets is complex and changeable.Therefore,to examine the correlation between assets,we must not only measure the degree of correlation between the return rate of assets,but also to consider the relevant structure.In view of the significant investment value of Hong Kong stocks in 2018,the article launched a portfolio risk analysis of the industry index of the Hong Kong stock market,using the Hang Seng Financial Index,the Hang Seng Property Index and the Hang Seng Industrial and Commercial Index to build a portfolio.Copula-GARCH model was used to measure the correlation structure between assets,and MCMC method was used to simulate the portfolio ratio.Then to build a Copula-GARCH-MCMC model by combining two methods mentioned.The results show that: the volatility of the returns of Hong Kong stocks in the industry stock index has cluster characteristics.The form of Copula function is flexible,and it can have a strong ability to characterize the relationship of assets with different related structural characteristics.The t-Copula function can better describe the correlation between stock indexes in the Hong Kong stock market.It implies that the relevant structure between industries was symmetrical.The trailing tails at both ends of the relevant structure indicate that extreme events are prone to occur.And it is more scientific to use the MCMC method to consider the investment ratio calculated by the winning performance of the historical yield.From the model's effectiveness,the Markowitz's approach may underestimate the risk of a portfolio,and the combination of the Copula function and the MCMC approach is more realistic,has a better reference value.