Discrete-Time Survival Models with Neural Networks for Age–Period–Cohort Analysis of Credit Risk

IF 2 Q2 BUSINESS, FINANCE Risks Pub Date : 2024-02-03 DOI:10.3390/risks12020031
Hao Wang, Anthony Bellotti, Rong Qu, Ruibin Bai
{"title":"Discrete-Time Survival Models with Neural Networks for Age–Period–Cohort Analysis of Credit Risk","authors":"Hao Wang, Anthony Bellotti, Rong Qu, Ruibin Bai","doi":"10.3390/risks12020031","DOIUrl":null,"url":null,"abstract":"Survival models have become popular for credit risk estimation. Most current credit risk survival models use an underlying linear model. This is beneficial in terms of interpretability but is restrictive for real-life applications since it cannot discover hidden nonlinearities and interactions within the data. This study uses discrete-time survival models with embedded neural networks as estimators of time to default. This provides flexibility to express nonlinearities and interactions between variables and hence allows for models with better overall model fit. Additionally, the neural networks are used to estimate age–period–cohort (APC) models so that default risk can be decomposed into time components for loan age (maturity), origination (vintage), and environment (e.g., economic, operational, and social effects). These can be built as general models or as local APC models for specific customer segments. The local APC models reveal special conditions for different customer groups. The corresponding APC identification problem is solved by a combination of regularization and fitting the decomposed environment time risk component to macroeconomic data since the environmental risk is expected to have a strong relationship with macroeconomic conditions. Our approach is shown to be effective when tested on a large publicly available US mortgage dataset. This novel framework can be adapted by practitioners in the financial industry to improve modeling, estimation, and assessment of credit risk.","PeriodicalId":21282,"journal":{"name":"Risks","volume":"3 1","pages":""},"PeriodicalIF":2.0000,"publicationDate":"2024-02-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Risks","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.3390/risks12020031","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"BUSINESS, FINANCE","Score":null,"Total":0}
引用次数: 0

Abstract

Survival models have become popular for credit risk estimation. Most current credit risk survival models use an underlying linear model. This is beneficial in terms of interpretability but is restrictive for real-life applications since it cannot discover hidden nonlinearities and interactions within the data. This study uses discrete-time survival models with embedded neural networks as estimators of time to default. This provides flexibility to express nonlinearities and interactions between variables and hence allows for models with better overall model fit. Additionally, the neural networks are used to estimate age–period–cohort (APC) models so that default risk can be decomposed into time components for loan age (maturity), origination (vintage), and environment (e.g., economic, operational, and social effects). These can be built as general models or as local APC models for specific customer segments. The local APC models reveal special conditions for different customer groups. The corresponding APC identification problem is solved by a combination of regularization and fitting the decomposed environment time risk component to macroeconomic data since the environmental risk is expected to have a strong relationship with macroeconomic conditions. Our approach is shown to be effective when tested on a large publicly available US mortgage dataset. This novel framework can be adapted by practitioners in the financial industry to improve modeling, estimation, and assessment of credit risk.
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
利用神经网络的离散时间生存模型进行信贷风险的年龄-周期-队列分析
生存模型已成为信用风险评估的常用方法。目前大多数信贷风险生存模型都使用基本的线性模型。这在可解释性方面是有好处的,但在实际应用中却受到限制,因为它无法发现数据中隐藏的非线性和交互作用。本研究使用内嵌神经网络的离散时间生存模型作为违约时间的估计器。这为表达变量间的非线性和交互作用提供了灵活性,从而使模型的整体拟合度更高。此外,神经网络还可用于估计年龄-周期-队列(APC)模型,从而将违约风险分解为贷款年龄(到期日)、发放时间(年份)和环境(如经济、运营和社会影响)的时间组成部分。这些模型既可以作为一般模型,也可以作为针对特定客户群的局部 APC 模型。局部 APC 模型揭示了不同客户群的特殊情况。由于环境风险预计与宏观经济条件有密切关系,因此相应的 APC 识别问题是通过正则化和将分解的环境时间风险成分与宏观经济数据拟合相结合来解决的。我们的方法在一个大型公开的美国抵押贷款数据集上进行了测试,结果表明是有效的。金融业的从业人员可以利用这一新颖的框架来改进信用风险的建模、估算和评估。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
求助全文
约1分钟内获得全文 去求助
来源期刊
Risks
Risks Economics, Econometrics and Finance-Economics, Econometrics and Finance (miscellaneous)
CiteScore
3.80
自引率
22.70%
发文量
205
审稿时长
11 weeks
期刊最新文献
Funding Illiquidity Implied by S&P 500 Derivatives Dynamics of Foreign Exchange Futures Trading Volumes in Thailand Automated Machine Learning and Asset Pricing What Drives Banks to Provide Green Loans? Corporate Governance and Ownership Structure Perspectives of Vietnamese Listed Banks Trends and Risks in Mergers and Acquisitions: A Review
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1