基于W-SVM的民营环保企业信用风险预警模型

Financial Risk Early Warning Model of Private Environmental Companies Based on Weighted Support Vector Machine

2019年中国绿色债券发行量依旧稳居世界前列,成为民营环保企业重要融资渠道,但是2018年至今,大量环保企业信用风险事件频发为我们敲响了警钟,构建合适的民营环保企业信用风险预警机制迫在眉睫.环保产业属于新兴产业,并以国有企业为主导,民营企业的样本数据具有样本量小,维度高等特征,这导致传统的信用风险模型适用性不强.因此选用加权支持向量机模型,对不同类别样本采取不同权值,选取大量财务特征,最终构建出风险预警模型.研究发现加权支持向量机模型具有十分优秀的预警性能.环保企业本身具有资金回收周期较长并且项目前期投入较高等特点,建议加强财务管理,保障资产流动性,建立完善产业链.

In 2019,China still maintaines a steady top position in the world in terms of the volume of green bond issued,making it an important financing channel for private environmental companies.However,a large number of financial risk events involving environmental companies since 2018 ring an alarm bell for us,pointing to a pressing need to construct a financial risk early warning mechanism suitable for private environmental companies.However,the environmental industry as an emerging industry is dominated by state-owned enterprises,which means sample data of private companies is characterized by smaller size and high dimensions.Traditional credit risk models have a low adaptability in addressing these problems.In this paper,a weighted support vector machine(WSVM)is presented,and based on different weights assigned to various types of samples,a large number of financial features are selected to construct a risk early warning model.The study shows that the WSVM model has a fairly good early warning performance.Given the characteristics of environmental companies such as long payback cycle and high upfront investment,it is recommended that financial management be strengthened to ensure the liquidity of assets and to construct a sophisticated industrial chain.