摘要
本文选取我国信用债2014-2017年全样本数据,采用时间风险模型方法实证研究了传统财务指标、公司特征以及地方环境指标对信用债券违约的影响。实证结果显示,传统财务预测模型无法很好解释我国债券违约状况。在此基础上,本文引入企业属性指标与地方经济环境指标建立了新的债券违约预测模型,并对其预测能力进行了检验和对比,发现离散风险模型拥有更好的预测能力和准确性。离散风险模型反映了违约风险随时间变化的特征,能够更好地匹配债券面板数据。伴随我国债券市场研究样本的扩大,利用离散风险模型能更为精准的对我国债券违约风险进行预测。
Based on the full sample data of China’s credit bonds from 2014 to 2017, this paper empirically studies the impact of traditional financial indicators, company characteristics, and local environmental indicators on credit bond defaults by using the hazard model, which finds that traditional financial predict models cannot explain China’s bond defaults. Non-financial indicators such as corporate nature and local economic variables will significantly affect the default risk. Based on the traditional default predict model, this paper introduces a new bond default predict model by using the indicators of corporate nature and local economic environment,and compares predictive ability and accuracy of the predict models. It finds that the new predict model produce out-of-sample forecasts that are more accurate than those of alternative models. The model reflects the time-varying characteristics of default risk, which can better match the bond panel data. With the tremendous expansion of the sample of China’s bond market, the discrete risk model can be used to predict the default risk of China’s bonds more accurately.
作者
姚红宇
施展
Yao Hongyu;Shi Zhan
出处
《投资研究》
CSSCI
北大核心
2018年第6期114-132,共19页
Review of Investment Studies
关键词
违约预测模型
债券违约
企业属性
地方经济
Default predict model
Bond default
Corporate nature
Local economic variable
