摘要
本文选取了我国195家农村合作金融机构作为研究对象,基于2010年23个CAMELs指标值和监管评级数据,应用Bayes判别分析和Logistic回归分析建立了两个监管评级判定模型,并使用样本金融机构2011年的数据验证了模型的可靠性。研究结果表明,银行监管评级结果主要由23个指标中的10个决定,监管工作中应重点关注这些指标的变化;两种模型对银行监管评级的判定能力较强,有助于简化监管评级工作和缩短评级时滞,从而帮助监管部门尽早采取应对措施。
Based on 195 China’s rural cooperative banks,this paper uses 23 CAMELs indicator values and supervisory ratings of these banks in 2010 to establish two supervisory rating models based on Bayes discriminant and logistic regression methods, and then verifies the reliability of the model by the sample banks’ data in 2011. The research result shows that a bank’s supervisory rating is mainly determined by 10 of the 23 indicators which should be taken seriously in supervisory work. The two models run well for bank rating which can help regulatory authorities simplify regulatory rating work, shorten its process, and take supervisory actions as soon as possible .
出处
《金融监管研究》
2012年第8期71-85,共15页
Financial Regulation Research
