【摘要】 以二次幂变差的测量为理论基础,研究了上证综指已实现波动率中的跳跃行为。将已实现波动率分解为连续样本路径方差和离散跳跃方差,研究了跳跃方差序列的统计特征,并且应用HAR-RV-C J模型对上证综指的已实现波动率进行预测。研究结果发现,几乎所有日、周和月已实现波动率的可预测性都来自连续样本路径方差,表明二次变差中的连续样本路径成分是中国股市已实现波动率预报的决定因素。更多还原

【Abstract】 Building on theoretical results for bi-power variation measures,the jump behavior of realized volatility for Shanghai synthesized index is examined.Realized volatility is divided into two parts: the continuous sample path variation and the discontinuous jump variation.The statistical feature of jump variation is studied.The realized volatility of shanghai synthesized index is predicted by using a HAR-RV-CJ volatility forecasting model.We find that almost all of the predictability in daily,weekly,and monthly realized volatility comes from continuous sample path variation,which indicates that non-jump component of quadratic variation is the most important determinant for forecasting realized volatility in Chinese stock markets.更多还原