“Implicit Government Guarantees in the U.S. Life Insurance Sector”

To what extent do investors view life insurers as “too big to fail?” I first examine market reactions to a U.S. Treasury announcement that raised expectations about government backstops for the industry. I find that a subset of large life insurers benefited from significant protection against bankruptcy, with implied probabilities of a government rescue ranging from 21% to 37% at the one-year horizon. Next, I structurally estimate a model of a life insurance sector with emergency bailouts. The estimates imply that for the 2001-2020 period the investor expectation of the probability of a bailout is 9.1% for large insurers. Indicative of prominent time-variation, the standard deviation of these bailout probabilities is 10.3%. A counterfactual analysis reveals modest average levels of moral hazard in risk-taking practices. Structural estimates for small life insurers imply more limited support. Overall, the results are consistent with the presence of a significant and time-varying “too big to fail” subsidy for large life insurers.

“Environmental Externalities of Debt Financing: Evidence from the Dust Bowl” (with Virginia Traweek)

It is well understood that the incentives associated with convex levered payoffs can be a source of agency frictions between borrowers and lenders. Less explored is how these incentives interact with the environmental externalities that borrowers might impose on less immediate stakeholders. Using Agricultural Census data, we show that counties with more levered farms subsequently experienced more severe soil erosion during the Dust Bowl. We explore evidence that this is driven in part by a “gambling for resurrection” motive among farmers struck by falling crop prices. The results highlight an important channel through which debt policy can shape environmental outcomes.