我国房地产企业的信用风险评价研究

Research on Credit Risk Evaluation of China’s Real Estate Enterprises

房地产行业在我国国民经济中占有重要地位,但近年来房地产企业违约现象却层出不穷。通过偿债能力、盈利能力、运营能力、发展能力、宏观经济等5个方面共计23项指标,运用主成分分析法对2010年至2020年间我国121家上市房地产企业的数据进行处理,分别采用Logistic模型、随机森林、XGBoost算法,建立我国房地产企业信用风险评价模型。结果显示,这三个模型都具有较好的违约风险预测效果,准确率均在90%以上,其中Logistic模型的准确率最高,其次为XGBoost算法。

The real estate industry plays an important role in national economy in China,however,endless defaults of the real estate enterprises have emerged in recent years.This paper analyzes the data of 121 Chinese listed real estate enterprises from 2010 to 2020 by using the principal component analysis method through 23 indicators in five aspects including solvency,profitability,operation ability,development ability and macro economy.Logistic model,random forest and XGBoost algorithm are respectively adopted,and credit risk evaluation models of Chinese real estate enterprise are established.The results show that the three models all have good default risk prediction effect,and the accuracy rate is above 90%.The accuracy rate of Logistic model is the highest,followed by XGBoost algorithm.