摘要
用当前机器学习领域最为前沿的LightGBM算法研究大股东进行股权质押的影响因素。研究发现,大股东以往累计质押比例是影响大股东股权质押的最重要的因素。除此之外,大股东股权质押更多受到融资约束和资金需求的影响,具体表现为资产规模小的中小企业、增长速度快的成长性企业的大股东更有可能发生股权质押。与国有企业相比,资产规模对民营企业大股东股权质押的影响更大。加强对大股东股权质押的强制性披露,增强对中小企业及民营企业的金融服务,促进民营企业及中小企业自身良好发展,有助于缓解大股东股权质押问题、稳定资本市场。
This paper uses the most advanced LightGBM algorithm in the current machine learning field to study what makes the largest shareholder be involved in the equity pledge.The study found that the cumulative pledge ratio of the largest shareholder is the most important factor affecting the equity pledge.In addition,the equity pledge of the largest shareholder is more affected by financing constraints and the capital needs.Compared with state-owned enterprises,the scale of assets has a greater impact on the equity pledge of private enterprises.Strengthen the mandatory disclosure of the equity pledge of major shareholders,enhance the financial services for small and medium-sized enterprises and private enterprises and promote the good development of private enterprises and small and medium-sized enterprises to stabilize the capital market.
作者
宋曼
卓建伟
SONG Man;ZHUO Jianwei(Antai College of Economics&Management,Shanghai Jiao Tong University,Shanghai 200030,China)
出处
《上海管理科学》
2020年第3期35-39,共5页
Shanghai Management Science
基金
国家自然科学基金(71673186)。
关键词
股权质押
影响因素
机器学习模型
equity pledge
influencing factors
machine learning model