区间均值条件波动率模型

Sectional Mean Conditional Volatility Model

本文讨论了有、无冲击(Innovations)对资产波动率的不对称性影响,它是现有研究关于正、负冲击不对称性影响的扩展。本文以GARCH模型为基础,假设条件均值为一个区间,建立了基于区间均值的GARCH模型及其非对称形式,并讨论了该类模型的估计方法。在中国债券市场的实证分析中发现,本文提出的模型能够更好地捕捉市场上波动特征;以已实现波动率为基准,区间均值GARCH模型取得了比GARCH模型更好的预测效果。

This paper discusses the asymmetric influence of without-news and with-news innovations on asset volatility, which extends present asymmetry, effect between positive and negative innovations. Supposing that the conditional mean in conditional volatility model is in a section, this paper proposes a sectional mean GARCH model and its asymmetric form based on the general GARCH model. The empirical results on Chinese Treasury bond market show that the sectional mean GARCH model can capture the volatility dynamics better and, with the benchmark or realized volatility （RV）, the model proposed in this paper is more capable of forecasting than GARCH model.