Design and Renegotiation of Debt Covenants

We analyze the design and renegotiation of covenants in debt contracts as a specific example of the contractual assignment of property rights under asymmetric information. In particular, we consider a setting where managers are better informed than the lenders regarding potential transfers from debt to equity that are associated with future investments. We show that this simple adverse-selection problem leads to the allocation of greater ex-ante decision rights to the creditor (the uninformed party) – i.e., tighter covenants – than would follow under symmetric information. This corresponds well with empirical evidence indicating covenants are indeed very tight upon inception. Strict covenants in turn lead to a bias in ex-post renegotiation, with the creditor giving up these excessive rights – i.e., covenants are waived. This result contrasts with the conclusion of standard incomplete contracting models of ex-ante hold-up, where the party with the more important ex-ante investments – presumably management, rather than the creditors – is granted decision rights. ∗Gârleanu: Wharton School, University of Pennsylvania, 3620 Locust Walk, Philadelphia, PA 19104-6367, email [email protected]; Zwiebel: Graduate School of Business, Stanford University, Stanford CA 94305, email [email protected]. We are grateful to George Baker, Bernard Black, Patrick Bolton, Philip Bond, Peter DeMarzo, Itay Goldstein, Hayne Leland, Jonathan Levin, Michael Roberts, Neil Stoughton, Steve Tadelis, Ralph Winter, several anonymous referees, and seminar participants at Carnegie Mellon, Columbia University, DePaul University, Duke University, New York University, Stanford University, the University of British Columbia, UCSD, the University of Utah, USC, the American Finance Association Meetings, the NBER Conference on Organizations, the European Finance Association Meetings and the Western Finance Association Meetings, for helpful comments. In this paper we analyze the design and renegotiation of covenants in debt contracts as a particular example of the contractual assignment of property rights under asymmetric information. We start with the presumption that debtholders are less informed than an entrepreneur about potential future transfers from debt to equity, and explore the implications of this adverse selection on the design and subsequent renegotiation of debt covenants. Intuitively, we analyze the notion that debtholders may receive stronger decision rights (in the form of tighter debt covenants) than under symmetric information, in order to “protect” them from an informational asymmetry, and the implications this will in turn have for information acquisition and future covenant renegotiation. Covenants are a common feature of debt contracts, and are generally understood to protect debtholders against activities that transfer wealth from them to shareholders. Most corporate debt contracts include covenants that place restrictions on the issuing firm, and thereby effectively serve to allocate control rights, in certain states of the world. For example, a firm might not be allowed to issue new debt if net working capital is below a specified level or if an interest coverage ratio is too low. Common covenant conditions are based on firm net worth, working capital, leverage, interest coverage, and cash flow; and involve restrictions on issuing debt, paying dividends, and investing, or impose actions such as the acceleration of debt payments if the specified condition is binding.1 A striking empirical regularity, however, is less well understood: covenants are remarkably tight. Chava and Roberts (2005) documents that at the inception of a debt contract, the average covenant threshold is only about one standard deviation away from the current value of the accounting ratio in question. As a consequence, 15%-20% (depending on the type of covenant) of outstanding loans are in violation during a typical quarter, and conditional on violating a covenant, a loan is delinquent about 40% of the time. Overall, for current-ratio covenants, about 50% of firms and 42% of loans are delinquent at some point in their lives, while for the less frequently violated net-worth covenant, about 30% of the loans are delinquent at some time. Furthermore, given such tight initial covenants, loans often fall into violation very quickly. The median covenant violation occurs at the end of the first third of the life of the loan, which translates to only one year from the inception of the loan. One direct implication of this tightness is that covenants are frequently renegotiated, and when they are the conditions imposed on the firm are routinely relaxed (or waived), and virtually never strengthened.2 Our analysis yields a simple explanation based on asymmetric information See, for example, Smith and Warner (1979), Smith (1993), and Bradley and Roberts (2004). 2 In a typical covenant waiver, the debtholder allows the manager to undertake some action prohibited in an initial covenant or relaxes the covenant that is in breach, in exchange for a higher interest rate and/or new additional restrictions. Note that in principle, covenants that are not in violation could also be renegotiated to be strengthened, by prohibiting some previously allowed actions or by tightening the triggering threshold in exchange for a lower interest rate. The asymmetry we note in the text is that while the first type of renegotiation (relaxing the covenant) is common, the latter type of renegotiation (strengthening the covenant) is virtually unseen in practice. See, for