Side-by-side Chinese-English


刘逸爽1 陈艺云1

(1.华南理工大学经济与贸易学院, 广东广州 510006)

【摘要】选取2011~2015年139家ST的上市公司作为研究样本, 以财务困境公司和正常公司年报的文本内容为基础, 经过中文分词处理, 利用国内外常用的情感词典来衡量文本内容所传递的管理层语调, 然后与传统财务比率变量相结合, 采用Logistic回归、决策树和支持向量机三种方法来构建信用风险预警模型, 对语调变量加入前后模型的预测能力进行实证检验, 结果表明, 文本内容传递的管理层语调确实提高了信用风险预警模型的效力, 描述性文容提供了定量财务数据所不能反映的增量信息。因此, 为了防范信用和债务危机, 有必要构建嵌入管理层语调文本分析的信用风险预警系统和风险评估模型。

【关键词】 管理层语调;信用风险预警;文本分析;情感分析;机器学习;


【基金资助】 国家社会科学基金一般项目 (15BJY149) ; 广东省自然科学基金博士启动项目 (2015A030310156) ; 广东省软科学研究计划项目 (2016A070705013) ; 广东省哲学社会科学“十二五”规划学科共建项目 (GD14XYJ05) ;

Tone at the top and credit risk warning for listed companies: textual analysis of company annual reports

LIU Yishuang1 CHEN Yiyun1

(1.School of Economics and Commerce, South China University of Technology, Guangzhou, China 510006)

【Abstract】Based on a sample of 139 ST listed companies from 2011 to 2015, this study applied Chinese word segmentation to the textual content of annual reports issued by companies in financial distress and those who are not. A sentiment dictionary commonly used in China and abroad was used to measure the management’s tone at the top, conveyed by textual content. Traditional financial ratio variables were also incorporated to construct a credit risk early warning model using logistic regression, decision trees, and a support vector machine. Empirical tests were performed on the predictive ability of the model before and after the inclusion of tone variables. The results indicate that the management’s tone at the top, conveyed by textual content, actually improves the effectiveness of the credit risk early warning model, and the descriptive text provides additional information not otherwise reflected by quantitative financial data. Therefore, to prevent credit and debt crises, it is necessary to construct a credit risk early warning system and a risk assessment model that analyzes the management’s tone at the top.

【Keywords】 tone at the top; credit risk warning; textual analysis; emotion analysis; machine learning;


【Funds】 General project of The National Social Science Fund of China (15BJY149); Doctoral Research Project of the Natural Science Foundation of Guangdong Province (2015A030310156); Soft Science Research Project of Guangdong Province (2016A070705013); Project of the Twelfth Five-Year Plan of Joint-Construction of Philosophy and Social Sciences in Guangdong Province (GD14XYJ05);

Download this article

    [1] Standard & Poor’s, 2003. Ratings and Ratios: Corporate Ratings Criteria, Standard & Poor’s Report.

    [2] Lu H., Tsai F., Chen H., Hung M. Li S., 2012. Credit Rating Change Modeling Using News and Financial Ratios. ACM Transactions on Management Information System, Vol. 3, No. 3: 1–30.

    [3] Lu Y. C., Shen C. H., and Wei Y. C., 2013. Revisiting Early Warning Signals of Corporate Credit Default Using Linguistic Analysis. Pacific-Basin Finance Journal, Vol. 24: 1–21.

    [4] Bian, H., Wan, C., Liu, D. et al. Computer Science (计算机科学), (11) (2013a).

    [5] Bian, H., Wan, C. & Wan, J. Journal of Jiangxi University of Finance and Economics (江西财经大学学报), (3) (2013b).

    [6] Song, B., Zhu, J. & Li, X. Journal of Central University of Finance and Economics (中央财经大学学报), (6) (2015).

    [7] Tennyson B. M., Ingram R. W., and Dugan M. T., 1990. Assessing the Information Content of Narrative Disclosures in Explaining Bankruptcy. Journal of Business Finance & Accounting, 17 (3): 391–410.

    [8] Bloomfield R. J., 2002. The “Incomplete Revelation Hypothesis” and Financial Reporting. Accounting Horizons, Vol. 16, No. 3: 233–243.

    [9] Cecchini M., Aytug H., Koehler G. J., and Pathak P., 2010. Making Words Work: Using Financial Text as a Predictor of Financial Events. Decision Support Systems. Vol. 50, No. 7: 164–175.

    [10] Lin, L. & Xie, D. Journal of Finance and Economics (财经研究), (7) (2016).

    [11] Xie, D. & Lin, L. Accounting Research (会计研究), (2) (2015).

    [12] Yang, Y., Zhou, Y. & Zhou, Z. Big Data Research (大数据), (1) (2017).

    [13] Loughran T., and Mcdonald B., 2016. Textual Analysis in Accounting and Finance: A Survey. Journal of Accounting Research. Vol. 54, No. 4: 1187–1230.

    [14] Loughran T., Mcdonald B., 2011. When is a Liability Not a Liability? Textual Analysis, Dictionaries, and 10-Ks. Journal of Finance, Vol. 66, No. 1: 35–65.

    [15] Henry E., 2008. Are Investors Influenced by how Earnings Press Releases are Written? Journal of Business Communication, Vol. 45, No. 4: 363–407.

    [16] Shi, X., Ren, R. & Xiao, Y. Chinese Journal of Management Science (中国管理科学), (4) (2006).

This Article


CN: 44-1696/F

Vol 33, No. 04, Pages 46-54

July 2018


Article Outline


  • 1 Introduction
  • 2 Empirical research design
  • 3 Samples and data
  • 4 Empirical results and analyses
  • 5 Conclusions and implications
  • References