Product market competition and firms’ narrative disclosures: evidence from risk factor disclosures

Abstract

This study examines how product market competition affects firms’ narrative disclosures of Item 1A Risk Factors in 10-K filings. We find that firms in more concentrated industries tend to disclose a greater quantity of narrative risk information. Besides, such firms provide risk disclosures more similar to those of their competitors, hence reducing the quality of the disclosure. We also document similar findings for idiosyncratic risk disclosure, which is inherently more firm-specific. The results imply that firms in more concentrated industries avoid divulging risk information in their narrative disclosures by disclosing more similar information rather than by reducing the amount of risk disclosure.

Keywords:

• product market competition
• industry concentration
• proprietary cost
• risk factor disclosure
• textual analysis

JEL classification:

• D43
• M40
• M41

Acknowledgements

We gratefully appreciate all the suggestions and comments from Jeong-Bon Kim (editor) and the anonymous reviewer. We also thank Hsinchun Chen, Yu-Hung Chen, Dan Dhaliwal, Sandy Klasa, Chih-Hsien Liao, Pei-Cheng Liao, Hsin-min Lu, Stephen Lusch, Matthew Serfling, Tawei Wang, and all participants in 2012 Spring Financial Accounting Seminar in the University of Arizona and in 2013, AAA Annual Meeting for their help and comments. Ju-Chun Yen gratefully acknowledges financial support from National Science Council (now Ministry of Science and Technology) in Taiwan. The previous version of this paper is titled “Product Market Competition and Firms’ Qualitative Disclosure”.

Notes

1. Previous empirical studies have documented inconclusive results on the relation between industry concentration and quantitative disclosures. See Section 2.1 for detailed discussions.

2. Specifically, firms in more concentrated industries are closer to oligopoly and farther away from perfect competition. These firms experience higher proprietary costs due to their highly interdependent business strategies (Ali, Klasa, and Yeung, 2014). On the other hand, firms in less concentrated industries are closer to perfect competition and have lower proprietary costs since they earn nearly zero profit. We do not consider monopoly in our study since there are no existing competitors and thus no proprietary costs for a monopolistic firm.

3. For instance, Kravet and Muslu (2013) mention that the deficiency of details and specific content may not be informative: “Despite the regulatory environment, companies may easily avoid providing useful risk-related information” (5).

4. The risk factor disclosure in Item 1A also satisfies the criteria stated in Karuna (2011). Karuna (2011) indicates that a suitable proxy for disclosure in the line of studies on competition and disclosure should be forward-looking and product market-related, and verification as whether it will eventually be realized should be difficult.

5. Hoberg and Phillips (2010) and Brown, Tian, and Tucker (2013) also calculate the cross-sectional similarity of narrative disclosures but their scopes are different from ours. Hoberg and Phillips (2010) use the similarity of product descriptions to measure competition and Brown, Tian, and Tucker (2013) use the similarity of risk factor disclosures across firms to examine the spillover effect of the SEC comment letters, while our study measures the relative informativeness of narrative disclosures.

6. The last survey of industries conducted by the US Census Bureau was in 2012. However, the survey results of industry concentration in 2012 are not on the US Census website as of June, 2013. Therefore, we cut off our sample period in 2009 since we believe the actual industry status after 2010 will be closer to the survey data in 2012 than that in 2007 (if the level of industry concentration changes dramatically from 2007 to 2012).

7. We appreciate Hsin-min Lu for providing the Item 1A raw texts data-set used in his research.

8. The five risk categories are defined by Campbell et al. (2014) including financial (e.g. leverage, liquidity), tax (e.g. loss carryback, uncertain tax position), legal & regulatory (e.g. litigation, potential lawsuit), other-systematic (e.g. industry condition, market supply), and other-idiosyncratic (e.g. backlog, internal control) risks. We modify the components only within each group but do not change the categories. See Appendix 1 for the full keyword list.

9. In the classifying process, Campbell et al. (2014) first identify whether the keywords belong specifically to financial, litigation or tax risk subcategories, and then group the remaining words into other-systematic or other-idiosyncratic risk depending on whether the information is economy-wide or firm-specific (see Appendix 1 in Campbell et al. 2014). Therefore, using the idiosyncratic risk keyword list alone to capture idiosyncratic information might be incomplete since first three categories may also convey idiosyncratic information. However, excluding the first three categories should weaken our empirical results in H2-B because these categories include common information which is inherently similar within the industry.

10. We gratefully thank Sandy Klasa and Matthew Serfling for sharing the organized data-set of the US Census-based concentration ratios. The original data can be retrieved from the US Census Bureau website: http://www.census.gov/econ/concentration.html.

11. For the tests of H1-A and H1-B, we also calculated the standard errors of coefficients clustered by NAICS industries as suggested by Petersen (2009). The significance levels of coefficients under clustering remain similar.

12. LOGWDLEN includes Item 1A Risk Factors, Item 7 MD&A, Item 7A Quantitative and Qualitative Disclosures about Market Risk, and Item 3 Legal Proceedings. We thank Hsin-min Lu for providing this data.

13. Note that our four-firm concentration ratio covers public and private firms. Since detailed finance data are unavailable for private firms in Compustat, we choose only big-three firms, rather than big-four firms, within an industry year in our sample considering the possibility that the data of big private firms are missing in our sample. We admit that our definition of the leader maybe rough; however, we believe it can achieve our goal of excluding minor firms that have smaller market shares and face less oligopolistic competition and hence lower proprietary costs.