Do Firms Engage in Risk Shifting? Empirical Evidence∗

I empirically test whether rms engage in risk-shifting in a setting where corporate investment risk measures are available in SEC disclosures. Contrary to what risk-shifting theory predicts, I nd that rms reduce investment risk both when leverage increases and when they approach distress. In rm-level panel regressions I nd that rms reduce the riskiness of capital expenditures by 21.6% when leverage is high, relative to the average rm. In a second test, I use a natural experiment with exogenous shocks to leverage, and nd that rms with exogenous increases in leverage reduce risk taking. This result suggests risk reducing incentives during distress, such as borrower reputation and managerial reputation concerns, outweigh risk-shifting incentives in investment decision making. ∗I would also like to thank Todd Gormley, Edith Hotchkiss, Darren Kisgen, Nadya Malenko, Sébastien Michenaud, Je Ponti , Jon Reuter, Michael Roberts, Elena Simintzi, Phil Strahan, and Jérôme Taillard as well as seminar participants at the CEPR European Summer Symposium in Financial Markets 2013, Wharton, and the UNC Roundtable for Junior Faculty in Finance for their helpful comments and suggestions. All errors are my own. †The Wharton School, University of Pennsylvania, 3620 Locust Walk SHDH 2456, Philadelphia, PA, 19104. Email: [email protected]