摘要
本文首先运用多元统计技术对企业财务指标进行分析,选择其中的7个财务指标作为建立判别函数的计量参数,运用多元判别技术建立了4水平的线性判别函数。通过对培训样本的交叉回代和检测样本的检验,该模型具有较强的预测能力。说明所建立的信用风险评估模型对国有商业银行的贷款评价有较强的实用价值,可进一步增强国有商业银行信用评估的科学性,减少银行信用风险暴露,降低增量不良贷款,提高银行信贷资产质量,推动国有商业银行的进一步发展。
This paper analysis the financial ratios of companies with multi-statistics, Selects seven finance ratios to measure the company's financial performance. With which to construct four-level discrimination functions to measure credit risk of enterprises. According to being tested with review and cross-validation, These models are applied valuable, So these models can be applied to analyzing credit risk of enterprises for bank or other financial institutions. Which is a powerful exploration to credit risk quantitative evaluation of our country.
出处
《预测》
CSSCI
2004年第4期65-68,共4页
Forecasting
关键词
信用风险评估
判别分析
预测
credit risk evaluation
discrimination analysis
prediction