Multivariate volatility estimation of SVAR-GARCH model.

*(Chinese. English summary)*Zbl 1449.62275Summary: We considered the multivariate volatility of SVAR-GARCH model, and proposed a new method for estimating volatility. Firstly, the causal structure and error item of the statistical independence were solved by independent component analysis (ICA) method, and the relationship between the conditional covariance matrix of the residual term and the conditional covariance matrix of the error term was established. Then, the impulse response of the conditional volatility of multivariable GARCH model was estimated by using the estimation results of univariate GARCH model and the causal structure of recognition, and the estimation of multivariate volatility was realized. This method could effectively reduce the estimated parameters. The experimental results show that the volatility estimated by the new method is consistent with the law of energy futures market.