在线咨询
中国工业与应用数学学会会刊
主管：中华人民共和国教育部
主办：西安交通大学
ISSN 1005-3085  CN 61-1269/O1
工程数学学报

工程数学学报 ›› 2024, Vol. 41 ›› Issue (1): 67-87.doi: 10.3969/j.issn.1005-3085.2024.01.005

• • 上一篇    下一篇

基于特征选择和违约鉴别的我国上市公司债券违约预警模型研究

白钰铭,  姜昱汐   

  1. 大连交通大学经济管理学院，大连  116028
  • 收稿日期:2023-06-18 接受日期:2023-08-29 出版日期:2024-02-15 发布日期:2024-04-15
  • 通讯作者: 姜昱汐 E-mail: yuxi_jiang_dl@126.com
  • 基金资助:
    国家自然科学基金 (71731003)；辽宁省社会科学规划基金 (L18DTJ001).

Research on Bond Default Early Warning Model of Listed Companies in China Based on Feature Selection and Default Identification

BAI Yuming,  JIANG Yuxi   

  1. School of Economics and Management, Dalian Jiaotong University, Dalian 116028
  • Received:2023-06-18 Accepted:2023-08-29 Online:2024-02-15 Published:2024-04-15
  • Contact: Y. Jiang. E-mail: yuxi_jiang_dl@126.com
  • Supported by:
    The National Natural Science Foundation of China (71731003); the Social Science Planning Fund of Liaoning  Province (L18DTJ001).

PDF (PC)

166

摘要：

为了更好地挖掘债券违约的关键指标，确定有效的违约预警时间窗，建立对不同违约状态预测精度高且精简实用的债券违约预警模型，采用 SMOTE 方法对非均衡样本进行处理，并基于 XGBoost 方法，根据违约预警精度高和指标体系规模小反推违约预警模型，并确定最优预警时间窗。研究结果表明，当违约预警时间窗为 $t-1$ 期时预测效果较好，即用提前一年的指标数据能更好地预测债券是否违约；采用 XGBoost 的嵌入式特征选择方法建立违约预警模型，将模型训练与指标体系降维同时完成，计算更简便。通过与其他 7 个常用违约预测方法的计算结果对比，该模型违约预测精度高、降维效果好、计算速度快、可解释性强。

关键词: 债券违约预警, 预测精度, 特征选择, 非均衡样本, XGBoost, SMOTE

Abstract:

In order to select the indicators of bond default, to determine an effective default warning time window, and to establish a bond default warning model which is practically and has high prediction accuracy in different default states, SMOTE and XGBoost are applied to process imbalanced samples and determine the default warning model and the optimal warning time window according to the default warning accuracy and the number of indicators in the indicator system respectively. The results show that the prediction effect is better when the default warning time window is $t-1$, which means that using indicator data one year in advance can better predict whether bonds will default. Using the embedding feature selection of XGBoost algorithm to establish the default early warning model can complete the model training and indicator system dimension reduction simultaneously, which makes the compute easier. Comparing with the other 7 common default prediction methods, the proposed model has higher default prediction accuracy, more effective dimension reduction, less computing time, and stronger interpretability.

Key words: bond default warning, predictive accuracy, feature selection, imbalanced sample, XGBoost, SMOTE

中图分类号: