中证白酒指数不同波动下的在险价值——基于GARCH族模型和加权历史模拟法

The value at risk under different fluctuations of China security liquor index: based on GARCH family model and weighted historical simulation method

选取中证白酒指数1550天收盘价数据,基于T分布和广义误差分布(GED)GARCH族模型,对比发现GED分布下CGARCH-M模型能够体现收益率的非正态性、异方差性和波动率杠杆效应,对中证白酒指数收益率拟合效果更优;利用加权历史模拟法和CGARCH-M-GED模型对指数在险价值进行计算,发现低波动期和不稳定波动期加权历史模拟法易高估市场风险,在高波动期模型无效;根据预测失败率和Kupiec检验结果,非对称GARCH模型在各波动时期均能得到中证白酒指数更优的风险预测效果.

Selecting the closing price data of China Security Liquor Index for 1550 days and based on T-distribution and GARCH family models, it is found that CGARCH-Mean model under Generalized error distribution(GED) can reflect the abnormality, heteroscedasticity and volatility leverage effect of return rate. The model has a better fitting effect on the yield of China Security liquor index. By aid of the weighted historical simulation method and CGARCH-M-GED model, its value at risk is calculated. It is found that the weighted historical simulation method is easy to overestimate the market risk in low fluctuation period and unstable fluctuation periods, yet it is invalid in high fluctuation periods. According to the prediction failure rate and Kupiec test, the asymmetrical GARCH model can get better risk prediction effect of China Security Liquor Index in each fluctuation period.