国债超额收益与波动率分解

发布时间：2018-07-06 14:29

本文选题：国债超额收益 + 跳跃风险　；参考：《厦门大学》2014年硕士论文

【摘要】：本文将债券的已实现波动率分解成两个部分：一部分是连续部分的波动率,也就是所谓的已实现二次幂变差,另一部分是非连续部分的波动率,也就是已实现跳跃的均值。我们发现：使用二次幂变差的技术从美国30年国债远期的高频数据中提取出已实现二次幂变差均值的滚动估计量,并将该估计量加入到Jonathan Wright and Hao Zhou (2009)的回归中,能够在跳跃因子的基础上显著提高该回归对国债超额收益率的预测效果。在样本外预测中,把二次幂变差均值包含进去将使均方根预测误差在Jonathan Wright and Hao Zhou (2009)的基础上减少12-27%。另外,我们发现跳跃部分的波动率对国债超额收益有很好的预测作用,但是对当期的利率期限结构几乎没有解释作用,而连续部分的波动率不仅对国债超额收益有很好的预测作用,同时能够驱动当期利率期限结构的变化。这个结果说明了跳跃部分的波动率是一种纯粹的不可张随机波动率,而连续部分的波动率则更大程度上表现为一种相对不纯粹的不可张随机波动率,我们称之为部分不可张随机波动率。这种性质使得连续部分的波动率可能会成为连接债券市场与固定收益衍生品市场的桥梁。
[Abstract]:In this paper, the realized volatility of bonds is decomposed into two parts: one is the volatility of the continuous part, that is, the so-called realized quadratic power variation, and the other is the volatility of the discontinuous part, that is, the mean value of the realized jump. We find that the rolling estimator of the realized mean value of the quadratic variation is extracted from the high frequency data of the United States 30-year Treasury bonds by using the technique of quadratic variation, and the estimate is added to the regression of Jonathan Wright and Hao Zhou (2009). On the basis of jump factor, it can improve the prediction effect of this regression on the excess yield of national debt. In the extrasample prediction, the error of root mean square prediction can be reduced by 12-27 points on the basis of Jonathan Wright and Hao Zhou (2009). In addition, we find that the volatility of the jump part has a good predictive effect on the excess yield of the national debt, but it has little effect on explaining the term structure of the current interest rate. The volatility of the continuous part can not only predict the excess yield of the national debt but also drive the change of the term structure of the current interest rate. The results show that the volatility of the jump part is a pure intensible random volatility, while the volatility of the continuous part is more likely to be a relatively non-pure intensible random volatility. We call it partial intensible random volatility. This makes the volatility of successive parts a bridge between the bond market and the fixed-income derivatives market.
【学位授予单位】：厦门大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F832.51;F224

【共引文献】