Multivariate Copula-GARCH Model and Its Applications in Financial Risk Analysis
Zhang Shi-ying
Abstract
Zhang Shi-ying
Abstract
Combined copula techniques with GARCH model,multivariate Copula-GARCH model is provided to avoid defects of classical risk analysis models.Not only is non-linear dependence between financial markets able to be caught,but also flexible multivariate distribution which can be use to analyze portfolio Value-at-Risk is able to receive from this model.Monte Carlo techniques of portfolio based on copula techniques are fully discussed in this paper.The empirical results getting from Shanghai stock markets indicate that Copula-GARCH model with different marginal distributions and Monte Carlo techniques given in this paper is feasible and effective.
OpenAlex reports 7 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Combined copula techniques with GARCH model,multivariate Copula-GARCH model is provided to avoid defects of classical risk analysis models.Not only is non-linear dependence between financial markets able to be caught,but also flexible multivariate distribution which can be use to analyze portfolio Value-at-Risk is able to receive from this model.Monte Carlo techniques of portfolio based on copula techniques are fully discussed in this paper.The empirical results getting from Shanghai stock markets indicate that Copula-GARCH model with different marginal distributions and Monte Carlo techniques given in this paper is feasible and effective.
Key concepts: Copula (linguistics), Econometrics, Autoregressive conditional heteroskedasticity, Value at risk, Multivariate statistics, Monte Carlo method, Multivariate t-distribution, Marginal distribution