主管:中国科学院
主办:中国优选法统筹法与经济数学研究会
   中国科学院科技战略咨询研究院

中国管理科学 ›› 2025, Vol. 33 ›› Issue (2): 1-15.doi: 10.16381/j.cnki.issn1003-207x.2023.0794cstr: 32146.14.j.cnki.issn1003-207x.2023.0794

• •    下一篇

企业净资产收益率操纵度量模型与识别方法

孙庆文1,2(), 靳玉1,2, 郑立梅3, 孙子涵4   

  1. 1.河北经贸大学工商管理学院,河北 石家庄 050061
    2.河北经贸大学公司治理与企业成长研究中心,河北 石家庄 050061
    3.河北建筑设计研究院有限责任公司,河北 石家庄 050011
    4.西澳大学商学院,澳大利亚 珀斯市 WA600 9
  • 收稿日期:2023-05-11 修回日期:2023-11-04 出版日期:2025-02-25 发布日期:2025-03-06
  • 通讯作者: 孙庆文 E-mail:13832101028@126.com
  • 基金资助:
    河北省社会科学发展研究课题(20230602002)

Measurement Model and Identification Method of Return on Equity Manipulation: Empirical Research Based on the Data of Chinese A-share Listed Companies

Qingwen Sun1,2(), Yu Jin1,2, Limei Zheng3, Zihan Sun4   

  1. 1.School of Business Administration,Hebei University of Economics and Business,Shijiazhuang 050061,China
    2.Research Center for Corporate Governance and Enterprise Growth,Hebei University of Economics and Business,Shijiazhuang 050061,China
    3.Finance Department,Hebei Institute of Architectural Design & Research,Shijiazhuang 050011,China
    4.Business School,The University of Western Australia,Perth WA600 9,Australia
  • Received:2023-05-11 Revised:2023-11-04 Online:2025-02-25 Published:2025-03-06
  • Contact: Qingwen Sun E-mail:13832101028@126.com

摘要:

净资产收益率是反映企业盈利能力的重要财务指标。上市公司出于提高估值水平、摆脱外部融资约束、达到行权激励条件等目的,往往会使用多种手段对净资产收益率进行操纵。因此,准确度量和识别净资产收益率操纵的程度和手段,对于投资者正确评估公司价值、政府部门有效监管违规行为具有重要意义。本文从影响净资产收益率的多视角出发,采用文献研究、统计分析和实证研究等方法,构建了净资产收益率操纵度量模型。在度量出上市公司净资产收益率操纵程度的基础上,构造了嫌疑样本备选集,并采用临界值动态分位法识别和确定有显著操纵行为的嫌疑样本。实证研究结果表明,本文所构造的度量模型以及临界值动态分位法可有效地识别出嫌疑样本,且识别精度较高。

关键词: 净资产收益率操纵, 盈余管理, 现金股利管理, 度量模型, 识别方法

Abstract:

Return on equity is an important financial indicator that reflects the profitability of an enterprise. In order to improve the firm value, get rid of external financing constraints, and meet the salary incentive conditions, listed companies often use a variety of means to manipulate the ROE. Since the performance results achieved by ROE manipulation are unsustainable and are likely to erode future operating results and then lead to a decline in firm value, if not effectively identified, it will disrupt the normal order of the capital market and infringe on the interests of investors and creditors. Therefore, the precise measurement of the extent of ROE manipulation and accurate identification of suspected samples are of great significance for investors to correctly assess the value of the company and for government departments to effectively supervise violations.After combing through the previous studies, it is not difficult to find that there is a lack of systematic research focusing on ROE manipulation. Further, there is no quantitative research on the degree of ROE manipulation. Besides, in the research on suspect sample identification, no identification method has been found that can effectively identify suspect samples and their manipulation methods. However, enterprises will consider the cost-effectiveness of each manipulation method and choose different combinations of methods for ROE manipulation. The economic consequences will vary greatly depending on the combinations. Given this, it will focus on the ROE indicator, construct an ROE measurement model, and a suspect sample identification method to provide new methods and clues for ROE manipulation behavior identification in this paper.Through literature research and mathematical analysis, an ROE manipulation measurement model is constructed starting by analyzing multiple manipulation methods affecting ROE. Firstly, the accrual earnings management, real earnings management, and cash dividend management measurement models are constructed respectively. Secondly, according to the ROE calculation formula stipulated by the CSRC, the impact of the three manipulation methods on the ROE is calculated in total, and the measurement model of the degree of ROE manipulation is constructed. Based on the measurement of the extent of ROE manipulation, the suspect samples and manipulation methods are identified. Firstly, the samples with a degree of manipulation of ROE greater than 0 are preliminarily defined as suspect samples. Secondly, the extent of manipulation of each method of the suspect samples is standardized to make samples in different scales and industries comparable. Thirdly, the quantile threshold adjustment method is used to determine the final suspect samples.The data of China's A-share listed companies from 2014 to 2019 as the research samples are selected to test the validity of the measurement model and identification method in the following four steps. Firstly, the identification ability test of each measurement model; Secondly, taking the real samples of earnings management disclosed by the CSRC as the standard to test the identification errors; Thirdly, a comparative analysis of identification error of measurement models and identification method; Fourthly, from the perspective of the economic consequences of ROE manipulation, the effectiveness of constructed models and identification method is further tested. The empirical research results show that the measurement models and the identification method constructed in this paper can effectively identify suspect samples with high recognition accuracy.The research in this paper is of great significance for maintaining the order of the capital market. The construction of the ROE manipulation measurement models and identification method is helpful for investors and creditors to better understand and use financial information, then restrain the ROE manipulation behavior of listed company managers driven by self-interest, and protect the interests of investors.

Key words: return on equity manipulation, earnings management, cash dividend management, measurement model, identification method

中图分类号: