首页|公司财务特征与企业ESG评分的内在关联——基于随机森林方法的识别预测探究

公司财务特征与企业ESG评分的内在关联——基于随机森林方法的识别预测探究

扫码查看
ESG已成为衡量企业价值的重要理念,相对于企业自身披露的信息,外部机构的ESG评分是经过信息筛选形成关键事项议题并专业分析后形成的,旨在帮助投资者做出更明智的决策.本文实证分析了企业财务特征与机构ESG评分的内在关联,研究发现:我国上市公司财务指标特别是异常指标能在较大程度上预测ESG评分,即当前企业ESG评判仍以财务信息为主导,但预测能力逐年减弱,即ESG评分正趋于独立.在环境维度和社会维度下,财务特征对评分的预测能力显著低于综合得分,企业履行环境保护、产品责任等更多的是出于遵守法律法规的强制性规定和维护必要的声誉考虑.本文结论为完善中国特色ESG评分机制,促进其与财务信息的功能协同提供了借鉴.
The Intrinsic Association between Corporate Financial Characteristics and Corporate ESG Scores:An Exploration of Identification Prediction Based on Random Forest Approach
ESG has become an important concept for evaluating corporate value.Compared with the information disclosed by enterprises themselves,ESG scores of external organizations are formed after information screening to form key issues and professional analysis,aiming to make more informed de-cisions for investors.This paper empirically analyzes the intrinsic correlation between corporate financial characteristics and institutional ESG scores,and finds that:financial indicators of listed companies in China,especially abnormal indicators,can predict ESG scores to a greater extent,i.e.,the current corporate ESG judgment is still dominated by financial information,but the predictive ability has been weakening year by year,i.e.,the ESG scores are tending to be independent.Under the environmental and social dimensions,the predictive ability of financial characteristics on the score is significantly lower than the composite score,and the fulfillment of environmental protection and product responsibility by enterprises is more for complying with the mandatory provisions of laws and regulations and maintaining the necessary reputational considerations.The conclusions of this paper provide a reference for improving the ESG scoring mechanism with Chinese characteristics and promoting its functional synergy with financial information.

Corporate ESG ScoresFinancial CharacteristicsRandom Forest ApproachIdentify Predictions

钟玮、董淼

展开 >

中国财政科学研究院 100142

中国税务出版社有限公司 100055

企业ESG评分 财务特征 随机森林方法 识别预测

2024

会计研究
中国会计学会

会计研究

CSTPCDCSSCICHSSCD北大核心
影响因子:6.346
ISSN:1003-2886
年,卷(期):2024.(7)