论文部分内容阅读
基于中国股市逐笔交易数据,有效利用全样本不规则采样数据,采用二尺度已实现波动率(TSRV)方法对我国股票市场分离噪音下的日波动率进行估计,并将TSRV方法拓展至日内高频时段,结合RV模型进行了对比研究。研究结果显示逐笔交易数据相对传统超高频数据包含更多的交易信息,噪音的存在严重的干扰了波动率估计。TSRV方法在中国市场条件日间和日内都具有较好的适用性,能有效剔出微观结构噪音对波动率估计的影响,提高波动率的估计精度和稳定性。实际应用中,TSRV方法对低频采样频率的选取具有很好的鲁棒性。
Based on the transaction-by-transaction data of Chinese stock market, this paper makes an efficient use of the irregular sample data of the whole sample, estimates the daily volatility under the noise of China’s stock market using the second-order realized volatility (TSRV) method and expands the TSRV method to intraday high Frequency period, combined with the RV model were compared. The results show that the transaction by transaction data contains more transaction information than traditional UHF data, and the existence of noise seriously interferes with the volatility estimation. The TSRV method has good applicability during the daytime and the daytime of the Chinese market condition, effectively removing the influence of the microstructure noise on the volatility estimation and improving the estimation accuracy and stability of the volatility. In practical application, the TSRV method has good robustness to low-frequency sampling frequency selection.