摘要
应对气候变化是推动经济高质量发展和生态文明建设的重要抓手,企业的碳减排信用风险关系到投资者、银行和政府等利益相关者的资金保障。因此,有效预警中国制造型企业碳减排信用风险成为当前的研究热点。首先基于熵权TOPSIS与耦合协调度模型验证制造型企业碳减排信用风险预警的必要性;其次,通过与多模型对比,证明MLP深度学习模型在企业碳减排信用风险预警中的优越性;最后,通过压力测试方法确定利益相关者的重点防控企业和行业类型。结果表明:第一,多数制造型企业处于中度、高度协调度耦合阶段,即企业“碳排放水平”因素会显著影响自身的碳减排信用风险;第二,制造型企业碳减排信用风险对“预期碳税力度τ”具有敏感性,而碳税对“石油加工、炼焦及核燃料加工业”的冲击最为明显;第三,本文提出的MLP深度学习模型预测效果最佳,其预测准确率达到98.60%,相较于其他模型平均提升了15.57%,充分表明了该模型的可行性和实用价值。
Addressing climate change is a crucial lever for promoting high-quality economic development and the construction of an ecological civilization.The carbon emission credit risk of enterprises is directly related to the financial security of stakeholders such as investors,banks,and governments.Therefore,effectively warning about the carbon emission credit risk of manufacturing enterprises in China has become a current research hotspot.Firstly,this study validates the necessity of carbon emission credit risk warning for manufacturing enterprises based on the entropy weight TOPSIS and coupling coordination degree model.Secondly,through a comparison with multiple models,it demonstrates the superiority of the MLP deep learning model in carbon emission credit risk warning for enterprises.Lastly,the stress testing method is employed to identify the key control enterprises and industry types for stakeholders.The results indicate the following:Firstly,most manufacturing enterprises are in a stage of moderate to high coupling coordination,meaning that the“carbon emission level”significantly affects their carbon emission credit risk.Secondly,the carbon emission credit risk of manufacturing enterprises is sensitive to the“expected carbon tax intensity(τ),”with the petroleum refining,coking,and nuclear fuel processing industry being the most affected by carbon taxes.Finally,the MLP deep learning model proposed in this study achieves the best predictive performance,with an accuracy rate of 98.60%,which is an average increase of 15.57%compared to other models.This demonstrates the feasibility and practical value of the proposed model.
作者
龙志
陈湘州
滕熙玉
LONG Zhi;CHEN Xiang-zhou;TENG Xi-yu(Hunan University of Science and Technology,Xiangtan,Hunan 411201;Hunan Strategic Emerging Industries Research Base,Xiangtan,Hunan 411201)
出处
《怀化学院学报》
2024年第2期92-105,共14页
Journal of Huaihua University
基金
国家社会科学基金一般项目“持续调控背景下房地产市场利益分配协调机制及政策研究”(13BJY057)
国家社会科学基金一般项目“生态产品价值实现超级基金的制度设计与运行机制研究”(20BGL201)
湖南省社会科学基金一般项目“双碳目标下的我国企业碳资产价值转化机制研究”(22YBA141)。
关键词
“双碳”目标
企业碳减排信用风险
熵权TOPSIS
深度学习
压力测试
“double carbon”target
corporate carbon emission reduction credit risk
entropy weight TOPSIS
deep learning
stress test
