论文部分内容阅读
本文基于最近发展起来的非参数高频数据波动估计和跳跃识别方法,将波动中的连续成分和跳跃成分分离开来,在月度频率上进行风险收益权衡和波动非对称性检验。文章得出以下几点结论:首先,中国股市的跳跃存在明显的聚类特征(特别是2008年左右),已实现方差所代表的市场整体风险对收益率并没有明显的解释效力;其次,跳跃成分对收益率有稳健的预测作用,跳跃波动与收益率负相关;最后,跳跃特别是负向跳跃更为准确地反映波动的非对称性,并可以提高对波动的预测效果。
Based on the recently developed nonparametric high frequency data fluctuation estimation and jump identification method, this paper separates the continuous components and the jumping components in volatility, and conducts risk return balance and wave asymmetry test on the monthly frequency. The article draws the following conclusions: First, the jump in China’s stock market has obvious clustering characteristics (especially around 2008), and the overall market risk represented by the variance has no obvious explanatory power on the rate of return. Second, Compositions have a robust predictive effect on the yield. The jump volatility is negatively correlated with the yield. Finally, the jump, especially the negative jump, reflects the asymmetry of volatility more accurately and improves the forecasting effect on volatility.