基于NS族模型的中国公司债信用利差预测

本文以中国公司债为研究对象, 基于NS族模型研究了信用利差的预测问题。通过对不同期限、不同信用评级公司债信用利差的样本内外预测效果进行实证比较, 得到主要结论如下:(1)模型对中长期公司债信用利差的预测误差低于短期公司债。(2)不同信用评级公司债信用利差的预测效果受剩余到期期限的影响:1年期的AAA级公司债的预测误差低于AA+和AA级公司债; 5年期的AA+级公司债的预测误差低于AAA和AA级公司债; 10年期的AA级公司债的预测误差低于AAA和AA+级公司债。成果为各经济主体预测信用利差提供了具体思路和方法, 有利于做出合理的金融决策。

Forecasting Credit Spread of Chinese Corporate Bonds Based on NS-group Models

Taking Chinese corporate bonds as the research objects,based on the NS-group models,this paper studies the prediction of credit spread.By empirically comparing the prediction effects of in sample and out of sample on the credit spread of corporate bonds with different maturities and credit ratings,the main conclusions are as follows:The first is that the error of models in predicting the credit spread of medium-term and long-term corporate bonds is smaller than that of short-term corporate bonds.The second is that the prediction effects on the credit spread of corporate bonds with different credit ratings vary with maturities.The results provide concrete ideas and methods to predict credit spread for different economic entities,which are conducive to making reasonable financial decisions.