随机波动模型的扩展:理论与中国股市的实证

An Extension to Stochastic Volatility Model:the Theory and Empirical Analysis of Chinese Stock Market

传统随机波动模型(SV模型)仅从宏观基本面角度揭示了潜在波动的随机性。本文基于修正混合分布假设模型(即MMDH模型),将单因素SV模型拓展为两因素随机波动模型,并赋予每个波动因素新的经济意义解释。通过对中国股市高频数据和日数据进行了校准分析,所得校准结果与理论假设保持一致,并发现股价波动与信息到达过程和流动性风险均成正相关。最后,本文使用有效矩估计方法(EMM)比较了两因素SV模型和传统SV模型,其模型拟合统计量显示前者绝对优于后者;其得分t比率表明宏观因素控制波动的持久性,而市场微结构的流动性因素主要决定波动的厚尾性。

In the setting of the modified mixture distribution assumption model, we de- velop a general theoretical model for the new SV model with two factors, which contains mi- crostructure economic explanation. Choosing high frequency data of actively trading stocks in China for calibration, we find that the results are consistent with the theoretical assumptions, and stochastic volatility is positively correlated with new information arrival process and the liquidity. We then use the efficient method of moments to identify the adequacy of the models and find that the performance of the new SV model is better, the scores of t ratio of which in- dicates that the macro fundamental factor controls the persistency of the volatility, while the market microstructure factor leads to the fat tail.