关键词: SVAR模型, 独立成分分析, 因果结构, GARCH模型, 波动率

Abstract: 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 statistical independent 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.

Key words: SVAR model, independent component analysis (ICA), causal structure, GARCH model, volatility