通过深度学习测量名字的浓度

IF 9.8 1区经济学 Q1 BUSINESS, FINANCE

International Review of Financial Analysis Pub Date : 2025-09-13 DOI:10.1016/j.irfa.2025.104598

Eva Lütkebohmert , Julian Sester

{"title":"通过深度学习测量名字的浓度","authors":"Eva Lütkebohmert , Julian Sester","doi":"10.1016/j.irfa.2025.104598","DOIUrl":null,"url":null,"abstract":"<div><div>We suggest a new deep learning approach for the quantification of name concentration risk in loan portfolios. Our approach is tailored for small portfolios and allows for both an actuarial as well as a mark-to-market definition of loss. The training of our neural network relies on Monte Carlo simulations with importance sampling which we explicitly formulate for the CreditRisk + and the ratings-based CreditMetrics model. Numerical results based on simulated as well as real data demonstrate the accuracy of our new approach and its superior performance compared to existing analytical methods for assessing name concentration risk in small and concentrated portfolios.</div></div>","PeriodicalId":48226,"journal":{"name":"International Review of Financial Analysis","volume":"107 ","pages":"Article 104598"},"PeriodicalIF":9.8000,"publicationDate":"2025-09-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Measuring name concentrations through deep learning\",\"authors\":\"Eva Lütkebohmert , Julian Sester\",\"doi\":\"10.1016/j.irfa.2025.104598\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>We suggest a new deep learning approach for the quantification of name concentration risk in loan portfolios. Our approach is tailored for small portfolios and allows for both an actuarial as well as a mark-to-market definition of loss. The training of our neural network relies on Monte Carlo simulations with importance sampling which we explicitly formulate for the CreditRisk + and the ratings-based CreditMetrics model. Numerical results based on simulated as well as real data demonstrate the accuracy of our new approach and its superior performance compared to existing analytical methods for assessing name concentration risk in small and concentrated portfolios.</div></div>\",\"PeriodicalId\":48226,\"journal\":{\"name\":\"International Review of Financial Analysis\",\"volume\":\"107 \",\"pages\":\"Article 104598\"},\"PeriodicalIF\":9.8000,\"publicationDate\":\"2025-09-13\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"International Review of Financial Analysis\",\"FirstCategoryId\":\"96\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S1057521925006854\",\"RegionNum\":1,\"RegionCategory\":\"经济学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"BUSINESS, FINANCE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"International Review of Financial Analysis","FirstCategoryId":"96","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S1057521925006854","RegionNum":1,"RegionCategory":"经济学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"BUSINESS, FINANCE","Score":null,"Total":0}

引用次数: 0

摘要

我们提出了一种新的深度学习方法来量化贷款组合中的名称集中风险。我们的方法是为小型投资组合量身定制的，并允许精算和按市值计价的损失定义。我们的神经网络的训练依赖于蒙特卡罗模拟的重要性采样，我们明确地为CreditRisk +和基于评级的CreditMetrics模型制定了重要采样。基于模拟数据和实际数据的数值结果表明，该方法在评估小型和集中投资组合中的名称集中风险方面具有较好的准确性和较好的性能。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Measuring name concentrations through deep learning

We suggest a new deep learning approach for the quantification of name concentration risk in loan portfolios. Our approach is tailored for small portfolios and allows for both an actuarial as well as a mark-to-market definition of loss. The training of our neural network relies on Monte Carlo simulations with importance sampling which we explicitly formulate for the CreditRisk + and the ratings-based CreditMetrics model. Numerical results based on simulated as well as real data demonstrate the accuracy of our new approach and its superior performance compared to existing analytical methods for assessing name concentration risk in small and concentrated portfolios.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

International Review of Financial Analysis BUSINESS, FINANCE-

CiteScore

10.30

自引率

9.80%

发文量

366

期刊介绍： The International Review of Financial Analysis (IRFA) is an impartial refereed journal designed to serve as a platform for high-quality financial research. It welcomes a diverse range of financial research topics and maintains an unbiased selection process. While not limited to U.S.-centric subjects, IRFA, as its title suggests, is open to valuable research contributions from around the world.