结构突变下沪深300指数的波动率预测与评价

考察波动的结构突变性对模型估计和预测能力的影响，并通过SPA检验评估几种GARCH模型的预测能力优劣。研究发现，我国股市收益率的波动在样本期内发生了4次结构突变，波动存在着伪持续现象，且这种突变影响了模型的预测能力；SPA 检验表明，短期预测上，经结构突变修正的 GARCH 模型具有较高的预测能力，验证了结构突变对波动预测的重要性；长期预测上，基于滚动时间窗口下的 GARCH 模型具有较好的预测能力，通过调整时间窗口的方法来消除结构突变的影响有助于提升模型预测能力。

Forecasting the volatility of CSI 300 Index with Structural Breaks and GARCH models

The author takes account into the influence of the structural breaks into the GARCH model and the forecasting of the volatility, uses the SPA test to evaluate the ability of prediction of several GARCH models. The research shows the evidence of structural breaks in the unconditional variance of the stock returns series over the period and the high levels of persistence in the parameter estimates of the GARCH（1,1） model across the sub-samples. The impact of structural breaks on the accuracy of volatility forecasts has largely been ignored in previous research. The SPA test shows that, in the short-run volatility forecasting, the GARCH model with structural breaks performs best, whereas the GARCH model with rolling windows performs well in the long term prediction. By adjusting the estimation window, it can eliminate the influence of structural breaks and help improve prediction ability.