论文部分内容阅读
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。
The traditional stochastic volatility model (SV model) only discloses the randomness of potential volatility from a macro-level perspective. Based on the modified mixed hypothesis model (MMDH model), this paper extends the single-factor SV model to a two-factor stochastic volatility model and gives a new economic explanation for each of the volatility factors. Through the calibration analysis of the high frequency data and daily data in China’s stock market, the results of the calibration are consistent with the theoretical assumptions and found that the stock price volatility has a positive correlation with the information arrival process and the liquidity risk. Finally, this paper compares the two-factor SV model and the traditional SV model by using the effective moment estimation method (EMM). The model fitting statistics show that the former is superior to the latter. The score t ratio indicates the persistence of macro-factor control fluctuations, The liquidity factor of the market microstructure mainly determines the thick tail of fluctuation.