利率期限结构预测对债券组合风险管理具有重要影响。动态Ndson—Siegel类利率期限结构预测模型可能不稳定,本文对其进行一阶差分修正。实证结果表明一阶差分修正后模型对利率期限结构进行预测的误差显著减小。基于利率期限结构预测。将债券的预测收益率引入Nelson—Siegel久期向量匹配利率风险管理模型,实证结果表明在Nelson—Siegel久期向量匹配的条件下引入预测收益率匹配的约束条件后,可改善预测步长较长或目标债券期限较长情形下的利率风险对冲效果。
Interest rates term structure forecasting is crucial for bonds portfolio risk management. In this paper, dynamic Nelson - Siegel - style models are modified by modeling the differentiated series instead of original parameter series. The empirical study shows that the new models forecast interest rates term structure better significantly. Based on that, a new model of managing interest rates risk of bond portfolio with interest rates term structure forecasting information included is proposed, which also proved to be better in hedging interest rates risk of mid - term or long - term bonds.