Skip to toolbar

Nanyang Business School Forum on Risk Management and Insurance

Market Risk Disclosure and Crash Risk: Evidence from Textual Analysis

by | Dec 24, 2018 | Asset Pricing, Disclosure, Market Risk | 0 comments

Tags: Disclosure, Market Risk, Crash Risk, Latent Dirichlet Allocation, More from: Shingo Goto, Xin Luo, Zhao Wang

Editor’s Note: Presented at the 2018 China International Risk Forum. Posted by Zhao Wang, University of Rhode Island. Shingo Goto, Associate Professor of Finance, University of Rhode Island; Xin Luo, Assistant Professor of Accounting, Marquette University.

Investors, financial analysts, regulators, and other market participants generally agree on the necessity of improving the quality of disclosures that firms make to the public about their exposures to market risk, including interest rate risk, foreign currency exchange rate risk, commodity price risk, equity price risk, and so on. Enhancing the quality of market risk disclosures should help investors improve the process of security valuation and analysis (CFA Institute 2016) and reduce investors’ panics and sensitive trading behaviors in response to unfavorable changes in market conditions (e.g., Rajgopal 1999; Linsmeier et al. 2002; Thornton and Welker 2004).

With the objective of enhancing the quality of corporate financial disclosures, the US Securities Exchange Commission (SEC) issued Financial Reporting Release (FRR) No. 48 in 1997, which mandates firm managers to narratively disclose quantitative and qualitative information about the risk of loss arising from market prices and rates for debt, equity, currency, commodity, and other traded instruments as well as their derivatives in Item 7A of Form 10-K filings. Managers should also describe how they manage or hedge the risk exposures associated with market fluctuations.

While more than 20 years have passed since the SEC mandated narrative market risk disclosures in Item 7A, people have different views on the usefulness of Item 7A. For example, investors and financial analysts surveyed by CFA Institute (2016) generally express a low level of satisfaction with current market risk disclosures, owing to a large amount of boilerplate information in narrative market risk disclosures. There are widely-held perceptions that, while firms’ market risk disclosures may have the appearance of valid disclosures, in actuality many of them only provide routine, redundant, non-specific, and boilerplate information (e.g., Abraham and Shrieves 2014; CFA Institute 2016; SEC 2016). In a recent concept release, SEC (2016) calls for further discussions on the usefulness of current market risk disclosures and on strategies to improve their informativeness.

This study provides a new angle to the regulatory debate on the informativeness of narrative market risk disclosures in Item 7A of Form 10-K filings. To shed light on the ongoing discussion, we employ a language processing algorithm based on LDA (Blei et al. 2003; Bao and Datta 2014), to cluster the textual information in Item 7A into eight latent topics. We then examine how the disclosure of each topic (topic assignment or topic proportion) is related to the firm’s future stock price crash risk.

The majority of the latent topics extracted from Item 7A via LDA do not provide significant information about future crash risk. However, this study still uncovers two latent topics that are negatively significantly correlated with future stock price crash risk. These two latent topics are associated with exposures of the firms’ cash flows to risks in their input costs or/and output prices. The first topic is about the “commodity price risk and derivatives” and the second topic is about the “risks in product prices and materials costs.” In particular, the former topic is correlated with future crash risk even when we control for other predictors of crash risk. This suggests that the disclosure of “commodity price risk and derivatives,” conveys significant and incremental information about the firms’ future crash risk.

A further analysis suggests that the negative correlation between the disclosure (topic assignment) of the “commodity price risk and derivatives” and future crash risk is concentrated among low-accrual firms than among high-accrual firms. We also find that, among high-accrual firms, firms that spend a large proportion of Item 7A in disclosing “risks in product prices and materials costs” have a significantly lower crash risk. By conditioning on accruals, we can enhance the information content of the disclosures of “commodity price risk and derivatives” and “risks in product prices and materials costs” substantially.

Despite a few limitations and shortcomings, evidence in the paper provides a strong case against the commonly-held perceptions that Item 7A provides only boilerplate information. In particular, this study shows that some firms’ narrative market risk disclosures, especially those related to the “commodity price risk and derivatives” and “risks in product prices and materials costs,” convey relevant information that is negatively and significantly correlated with future stock price crash risk. Clearly, they convey useful, non-redundant, and non-boilerplate information to the public, which should help investors and analysists improve the process of security analysis and valuation, stock selection, and portfolio risk management.

The complete paper is available at:


Submit a Comment

Your email address will not be published. Required fields are marked *