首页|"方以类聚、物以群分":分级模型与中国存款保险定价

"方以类聚、物以群分":分级模型与中国存款保险定价

扫码查看
本文基于国际主流存款保险辖区最新的分级框架,设计了一个融合基准区间、评价体系、费率体系和检验体系的分级定价方案。首先,基于DRSK模型引入一种新的银行违约概率估计方法,对我国存款保险费率基准区间进行抽样估计。其次,融合风险矩阵评分、监管评分、经营质量评分和资本质量评级,提出兼顾"风险匹配"与"质量匹配"的银行评分体系。再次,综合考虑我国银行类型差异、风险差异和保费结构差异,提出兼顾"悬崖效应"与"级距效应"且符合我国国情的分级费率体系。最后,基于回溯法和匹配法证实分级方案在上市银行样本中的可行性,使用来自S&P、NUS信用研究计划和国内评级机构的映射违约率与累积正确率证实分级方案在非上市银行样本中的可行性。相关工作为我国存款保险制度完善与银行风险防控提供决策支持。
Birds of a Feather Flock Together:The Stratified Model and China's Deposit Insurance Pricing
Based on the latest stratified model from major international deposit insurance jurisdictions,this study develops a strat-ified pricing scheme for China that integrates benchmarking ranges,evaluation systems,rating structures,and validation mecha-nisms.Firstly,a new method for estimating bank default probabilities using the DRSK model is introduced,providing a benchmark interval estimation for deposit insurance rates based on sample data.Secondly,a scoring system that balances"risk matching"and"quality matching"is proposed,incorporating risk matrix scores,regulatory ratings,operational quality assessments,and capital quality ratings.Thirdly,accounting for the differences in bank types,risks,and premium structures,a stratified rating system is de-signed to address both"cliff effects"and"gradational effects"in line with national conditions.Finally,the feasibility of the pro-posed scheme is validated for listed banks using back-testing and match-testing methods,with further confirmation for non-listed banks through mapped default rates and cumulative accuracy ratios from S&P,the NUS Credit Research Initiative,and domestic rating agencies.This work provides decision-making supports for improving China's deposit insurance system and strengthening banking risk control.

deposit insurance pricingstratified modelcliff effectsback-testingmatch-testing

苏育洲、姚帅、王擎

展开 >

澳门科技大学商学院

西南财经大学金融学院

西南财经大学中国金融研究院

存款保险定价 分级框架 悬崖效应 回溯法 匹配法

2024

保险研究
中国保险学会

保险研究

CSSCICHSSCD北大核心
影响因子:1.072
ISSN:1004-3306
年,卷(期):2024.(12)