Risk Exposures from Risk Disclosures: What They Said and How They Said It

We extract information from 10-K risk disclosures: topic models measure what is discussed, and context models measure how it is discussed. We find that both contain significant predictive information about future aggregate risk exposures, even controlling for (structured) firm characteristics. For market exposure, only management's choice of context is useful; for other factors, tradable and nontradable, both context and topic information is useful. Further, we show that topics and contexts are statistically distinct. We analyze their relationship and shed light on those methods while economically interpreting our results.

Keywords: Text Analysis, Asset Pricing, Factor Model, Conditional Betas, Word embedding

JEL Classification: C23, C53, G11, G12, G17

Suggested Citation: Suggested Citation

Mazumder, Rahul and Pruitt, Seth and Ross, Landon, Risk Exposures from Risk Disclosures: What They Said and How They Said It (August 2023). Available at SSRN: https://ssrn.com/abstract=4637697 or http://dx.doi.org/10.2139/ssrn.4637697