Analysis And Application Of Risk Measurement Based On GARCH Models

Among the many methods for measuring the risk of securities investment,the Va R method is currently the mainstream method in the relevant research fields.Due to the characteristics of peaks and thick tails in financial time series,traditional methods such as variance-covariance can no longer accurately estimate Va R values.Fitting and estimating Va R values based on GARCH models has become a more reasonable and effective risk measurement method.At the same time,when measuring the risk of an asset portfolio,there is a time-varying correlation between different assets.Therefore,the multivariate GARCH model is used to describe the dynamic characteristics of the volatility of portfolio returns,which can provide more effective risk information.This article uses the Va R value estimated by the multivariate DCC-GARCH model to measure the risk of the investment portfolio,and establishes a portfolio selection model based on the minimization of the Va R value.This paper conducts an empirical analysis based on five stocks in the liquor sector.The normal distribution,t distribution,and GED distribution were selected to establish the GARCH(p,q)model,EGARCH(p,q)model,and TGARCH(p,q)model respectively,and the corresponding Va R value was estimated.At the same time,the Kupiec failure rate method was used to test The accuracy of Va R value estimation.When constructing the investment portfolio,this article combines the DCC-GARCH model and the Va R method to dynamically study the risk of the investment portfolio,and calculates the optimal portfolio through optimization algorithms.Through comparison,it is found that under the same expected rate of return and a certain confidence level,the DCC-GARCH-Va R model is used to calculate the expected rate of return of the portfolio,and the result is greater than the expected rate of return of the portfolio with the weights calculated by the mean-Va R model,and There is a higher Sharpe ratio.Therefore,it can be considered that the results obtained by the DCC-GARCH-Va R model are more helpful for investors to make new decisions and improve the rate of return of their investment portfolios.Considering that the estimation error generated by the parameter estimation will lead to the change of the entire portfolio income and increase unnecessary losses,this paper applies the mean square error MSE method to conduct a sensitivity analysis of the estimation error,and it is found that the mean square error depends on the DCC-GARCH model The ratio of the smallest eigenvalue to the largest eigenvalue of the estimated covariance matrix.The greater the difference between the maximum value and the minimum value of the eigenvalue of the covariance matrix,the greater the mean square error value,which indicates that the accuracy of the portfolio weight estimation is lower.