随机久期免疫:一种非参数方法

Stochastic Duration-matching Immunization:A Nonparametric Method

参数利率期限结构模型的多样性和计算的复杂性增加了随机久期免疫方法的使用成本,影响了随机久期免疫方法的实用性、准确性和稳定性。在单因子HJM模型的假设下提出了引入非参数方法的随机久期进行利率风险的免疫,显著降低了随机久期的模型设定误差问题对免疫结果的影响,从而改善了随机久期方法的免疫效果。通过实证结果的检验,该方法比参数方法估计的随机久期具有更好的免疫效果。

Parametric interest rate term structure model has many different types, and the calculations of these models are usually very complex. These problems increased the costs of using the stochastic duration-matching immunization strategies and reduced the practicability, and stability of the immunization performance. Under a Heath-Jarrow-Morton framework, we use nonparametric method to estimate the stochastic duration in order to reduce the effects of model specification error. This nonparametric estimation improves the immunization efficiency of the stochastic duration-matching method. The empirical results show that the method proposed in this paper has more accuracy result in the interest rate risk immunization.