文章基于2007年钢铁行业上市企业的样本数据,用主成分分析方法将筛选出的显著性指标减少为综合反映企业财务状况的5个主成分并建立信用风险识别的二元逻辑回归模型。实证结果表明,该模型中反映盈利能力和股东获得能力的资产报酬率、每股净资产及留存收益率是钢铁行业上市公司的关键性指标,利用该模型可以对上市公司一年后的信用状况进行预测。 Based on the sample data of listed companies in the iron and steel industry in 2007, the paper uses the principal component analysis method to reduce the selected salient indicators to five principal components that comprehensively reflect the financial status of enterprises and establishes a binary logistic regression model of credit risk identification. Empirical results show that the return on assets, net assets per share and retained earnings, which reflect the profitability and shareholders’ ability to obtain, are the key indicators of listed companies in the iron and steel industry. Using this model, the creditability of listed companies Make a prediction