摘要
信贷评级是信贷风险管理的前提,目前我国国有商业银行都采用这一方式管理信贷风险。本文在对国有商业银行当前信用评级方法存在问题和国内外相关研究成果进行分析的基础上,提出了构建国有商业银行内部信用评级模型,提高信贷风险管理水平的建议。作者利用贷款历史数据,通过因子分析和聚类分析等方法构建内部信用评级模型:通过因子分析方法构建的模型使评级指标体系更加科学、合理,避免了反映风险信息的冗余与遗漏;聚类分析使评级模型直接与违约概率挂钩,度量风险的准确性进一步提高。论文最后对模型进行了实证分析,使其有效性得到了检验。
Credit risk management is preconditioned on credit rating,which is generally used by our state-owned commercial banks to manage credit risks at present.After an analysis is given of the problems that exist in current credit rating of state-owned commercial banks along with relevant research results achieved at home and abroad,an internal credit rating model is proposed,designed to improve credit risk management for state-owned commercial banks.The model is built on historical data on loans through factor analysis and clustering method.Factor analysis makes the rating indicator system more scientific and reasonable,preventing risk information from being inflated and omitted;whereas clustering method links rating model directly with default ratio,raising the accuracy of risk measurement.Finally,the model is test-ed empirically to verify its effectiveness.
出处
《金融论坛》
CSSCI
2004年第4期16-21,共6页
Finance Forum
