Summary:
Information manipulation by listed companies is not rare in the stock market. The signals about such manipulation are usually transmitted to investors through various forms, including quantitative performance and qualitative tone. As accounting becomes more stringently reviewed, the violation cost of traditional financial performance fraud has conspicuously increased, which makes disclosure tone a popular object for companies to manipulate. Further, the deviation between performance and tone becomes a disclosure strategy. The core issue we hope to discuss is how the disclosure strategy of a company's tone deviating from its performance (abbreviated as tone deviation) affects investors' judgment of a company's true earnings and then their investment decisions, eventually resulting in short-term stock price fluctuations. The literature often treats performance and tone as two isolated parts of information transmission, arguing that irrational investors are easily misled into adopting irrational investment strategies by an abnormally positive tone. However, this can hardly explain the decline in stock price usually observed when a company's tone is extremely inconsistent with its performance. In this regard, different from the assumption of irrationality, we believe that investors are rational enough to adopt a cross-validation strategy in the face of multi-information. Based on a quick cross-validation of the company's performance and tone, an investor will consider the authenticity of the information disclosed to be higher if the performance and tone are relatively consistent; however, if the tone, the performance, or both deviate from the reality, the investor will reduce their evaluation of the stock and become more inclined to sell. We construct a novel linear regression model to test the views above. The A-share listed companies in mainland China on the Shanghai and Shenzhen stock exchanges from 2007 to 2020 are selected as our research sample. Using an event study framework, we characterize the investors' decision-making by the cumulative abnormal return (CAR) within a period of time after the issuance of the annual report. In terms of the core explanatory variable, we first analyze the text data of MD&A in the annual report to obtain a standardized index of net tone. On this basis, the net tone and performance indices are separately divided into 10 grades at the year and industry level, and a difference index can thus be obtained to examine the impact of the tone deviation strategy. Through the benchmark model and a series of robustness tests, we reach three main conclusions. First, the benchmark results based on the full sample show that tone deviation leads to a significant short-term decline in stock returns, and the difference between a company's tone and performance level has a nonlinear relationship with CAR. Second, the grouped test results by the dominant investor type show that retail investors who are at an information disadvantage rely more on cross-validation to assist themselves in making decisions. Third, with the deterioration of the quality of disclosure and the whole market information environment, the credibility of performance indicators will apparently decline, which impels institutional investors to adopt the same cross-verification strategy, resulting in the phenomenon of institutional retail. Based on the conclusions above, we put forward two policy recommendations. First, listed companies should pay attention to the rationality of investors and try to ensure that the tone of any information that may be sent to investors is consistent and fair, to reduce the possibility of misleading investors. Second, regulatory authorities should issue guidelines to regulate the tone of information disclosed by listed companies as soon as possible to strengthen the supervision of excessive tone deviation. The contributions of this paper consist of three aspects. First, this paper provides a new theoretical basis for market effectiveness in a multi-information environment. It indicates that stock prices can still reflect the real state of operation of listed companies to a considerable extent given the existence of cross-validation, which means that an efficient market is still established. Second, this paper proposes a new perspective to understand investors' rationality. By cross-verifying two types of information, investors can make more reasonable decisions and judgments at a lower cost of information cognition. Third, this paper improves the quantitative method of measuring tone deviation. Different from traditional methods that use the residuals of tone regressed on performance, we synthesize tone and performance into a quantitative index by grading and differencing. This method intuitively reflects the subjective information gap perceived by investors when they receive diverse information, and eliminates the common ambiguity found during tone analysis, providing a methodological contribution to follow-up research in this field.
王勇, 窦斌, 宋培睿, 何昕晟. 管理层语调偏离会影响投资者决策吗?——基于我国上市公司文本与财务数据的经验研究[J]. 金融研究, 2023, 513(3): 169-187.
WANG Yong, DOU Bin, SONG Peirui, HE Xinsheng. Does Tone Deviation Influence Investors' Decisions? Empirical Research on Text and Financial Data of Chinese Listed Companies. Journal of Financial Research, 2023, 513(3): 169-187.
Baginski, S., E. Demers, C. Wang, and J. Yu. 2016, “Contemporaneous Verification of Language: Evidence from Management Earnings Forecasts” ,Review of Accounting Studies, 21(1):165～197.
[16]
Barber, B. M., and T. Odean. 2008, “All That Glitters: The Effect of Attention and News on the Buying Behavior of Individual and Institutional Investors” ,The Review of Financial Studies, 21(2):785～818.
[17]
Bikhchandani, S., and S. Sharma. 2000, “Herd Behavior in Financial Markets”, IMF Economic Review, 47(3):279～310.
[18]
Boehmer, E., C. M. Jones, J. Wu, and X. Zhang. 2020, “What Do Short Sellers Know?”, Review of Finance, 24(6):1203～1235.
[19]
Davis, A. K., J. M. Piger, and L. M. Sedor. 2012, “Beyond the Numbers: Measuring the Information Content of Earnings Press Release Language” ,Contemporary Accounting Research, 29(3):845～868.
[20]
Eagly, A. H., and S. Chaiken. 1993. The Psychology of Attitudes. Published by Harcourt Brace Jovanovich College Publishers.
[21]
Huang, X., S. H. Teoh, and Y. Zhang. 2014, “Tone Management”, The Accounting Review, 89(3):1083～1113.
[22]
Larcker, D. F., and A. A. Zakolyukina. 2012, “Detecting Deceptive Discussions in Conference Calls”, Journal of Accounting Research, 50(2):495～540.
[23]
Li, F. 2010, “The Information Content of Forward-Looking Statements in Corporate Filings—a Naïve Bayesian Machine Learning Approach” ,Journal of Accounting Research, 48(5):1049～1102.
[24]
Loughran, T. I. M., and B. McDonald. 2011, “When Is a Liability Not a Liability? Textual Analysis, Dictionaries, and 10-Ks”, The Journal of Finance, 66(1):35～65.
[25]
Miyazaki, A. D., D. Grewal, and R. C. Goodstein. 2005, “The Effect of Multiple Extrinsic Cues on Quality Perceptions: A Matter of Consistency”, Journal of Consumer Research, 32(1):146～153.
[26]
Osgood, C. E., and P. H. Tannenbaum. 1955, “The Principle of Congruity in the Prediction of Attitude Change”, Psychological Review, 62(1):42～55.
[27]
Price, S. M., J. S. Doran, D. R. Peterson, and B. A. Bliss. 2012, “Earnings Conference Calls and Stock Returns: The Incremental Informativeness of Textual Tone”, Journal of Banking & Finance, 36(4):992～1011.
[28]
Ryan, L. V., and A. K. Buchholtz. 2001, “Trust, Risk, and Shareholder Decision Making: An Investor Perspective on Corporate Governance”, Business Ethics Quarterly, 11(1):177～193.
[29]
Tetlock, P. C., M. Saar-Tsechansky, and S. Macskassy. 2008, “More Than Words: Quantifying Language to Measure Firms' Fundamentals”, The Journal of Finance, 63(3):1437～1467.
[30]
Weiner, B. 1985, “An Attributional Theory of Achievement Motivation and Emotion”, Psychological Review, 92(4):548～573.
