摘要
银行股和房地产股在我国资本市场中都处于重要地位,行业间和行业内业务关联性高,波动也呈现出一定的规律性和相似性。本文以16家银行股和9家房地产股为研究对象,计算已实现波动率,并将K-means模型拓展为适用于时间序列的DTW-K-means,对波动进行聚类。结果表明,DTW聚类模型可以有效发现银行和房地产股价波动聚集特征,将有类似波动的企业聚在一起,也可区分具有独特风险特征的银行或企业。本文结论可为监管决策部门了解银行与房地产之间的关联提供数据借鉴,为分类实施监管政策提供参考。
Banks and real estate stocks are in the core position in China’s capital market. The business correlation between industries and within the industry is high, and the fluctuations also show certain regularity and similarity. This paper takes the leading 16 banks and 9 real estate stocks as the research objects, calculates the realized volatility, reflects the risk status, and innovatively expands the K-means model to DTW-K-means suitable for time series clustering. The results show that the DTW clustering model can effectively find the fluctuation clustering characteristics of bank and real estate stock prices and has typical industry clustering characteristics. At the same time, classification and clustering can explore the heterogeneity of corporate volatility, cluster companies with similar volatility, and distinguish banks or companies with unique risk characteristics.The conclusion can provide data reference for regulatory decision-making departments to understand the relationship between banks and real estate and provide a reference for implementing regulatory policies by classification.
作者
吴忠睿
吴金旺
邬华阳
WU Zhong-rui;WU Jin-wang;WU Hua-yang(School of Financial Management,Zhejiang Finance College,Hangzhou 310018;School of Statistics and Mathematics,Zhejiang Gongshang University,Hangzhou 310018;Department of Radio,Television and Information Technology,Organizing Committee of the 19th Asian Games in 2022,Hangzhou 310018)
出处
《财务与金融》
2022年第5期37-44,共8页
Accounting and Finance
基金
“浙江金融职业学院2021年度基本科研业务费青年科研一般项目”(编号:2021YB33)
“浙江省金融教育基金会课题”(课题编号:2022Y01)
“2022年度浙江省金融学会重点研究课题”。
作者简介
吴忠睿,男,浙江金融职业学院金融管理学院教师,研究方向:人工智能;吴金旺,男,浙江金融职业学院金融管理学院副院长,教授,研究方向:金融科技、风险管理;邬华阳,男,2022年第19届亚运会组委会广播电视和信息技术部运行保障处工程师,研究方向:大数据治理。
