I would like to thank Phil Berger, Dave Burgstahler, John Core, Ilia Dichev, Weili Ge, Leslie Hodder, Jack Hughes, Raffi Indjejikian, Mark Lang, Dave Larcker, Charles Lee, Roby Lehavy, Russell Lundholm, Dawn Matsumoto, Brian P. Miller, Venky Nagar, Marlene Plumlee, Grace Powell, Jonathan Rogers, Haresh Sapra, Cathy Shakespeare, Terry Shevlin, Doug Skinner, Abbie Smith, K. R. Subramanyam, Brett Trueman, Jenny Tucker, Regina Wittenberg Moerman, Sarah Zechman, the participants of the 2008 University of Michigan Traverse City Conference, the 2008 North American Summer Meeting of the Econometric Society, the 2009 FARS mid-year conference, the New York University 2009 Summer Camp, and the Deutsche Bank Asset Management seminar, and the workshop participants at Chicago, Notre Dame, University of Washington, UCLA, Indiana, Florida, Emory, and Stanford for their helpful comments and Julie Miller for her editorial help. Special thanks go to Merle Erickson (the editor) and an anonymous reviewer. I also thank Mike Chen, Qian Fang, Qian Li, Chris Liong, Doris Lu, Amy McMahon, Bharat Modi, Qimin Ni, Darin Styles, Kathy Wang, Tina Wei, Sarah Xia, Quan Yuan, and especially Rencheng Wang and Winnie Wei for their excellent research assistance and Ernst, and Young Faculty Development Fund and the Harry Jones Endowment for Earnings Quality at the University of Michigan for financial support.
The Information Content of Forward-Looking Statements in Corporate Filings—A Naïve Bayesian Machine Learning Approach
Article first published online: 16 AUG 2010
©, University of Chicago on behalf of the Accounting Research Center, 2010
Journal of Accounting Research
Volume 48, Issue 5, pages 1049–1102, December 2010
How to Cite
LI, F. (2010), The Information Content of Forward-Looking Statements in Corporate Filings—A Naïve Bayesian Machine Learning Approach. Journal of Accounting Research, 48: 1049–1102. doi: 10.1111/j.1475-679X.2010.00382.x
- Issue published online: 13 JUL 2010
- Article first published online: 16 AUG 2010
- Accepted manuscript online: 13 JUL 2010 12:00AM EST
- Received 14 January 2009; accepted 8 June 2010
This paper examines the information content of the forward-looking statements (FLS) in the Management Discussion and Analysis section (MD&A) of 10-K and 10-Q filings using a Naïve Bayesian machine learning algorithm. I find that firms with better current performance, lower accruals, smaller size, lower market-to-book ratio, less return volatility, lower MD&A Fog index, and longer history tend to have more positive FLSs. The average tone of the FLS is positively associated with future earnings even after controlling for other determinants of future performance. The results also show that, despite increased regulations aimed at strengthening MD&A disclosures, there is no systematic change in the information content of MD&As over time. In addition, the tone in MD&As seems to mitigate the mispricing of accruals. When managers “warn” about the future performance implications of accruals (i.e., the MD&A tone is positive (negative) when accruals are negative (positive)), accruals are not associated with future returns. The tone measures based on three commonly used dictionaries (Diction, General Inquirer, and the Linguistic Inquiry and Word Count) do not positively predict future performance. This result suggests that these dictionaries might not work well for analyzing corporate filings.