In this paper we examine whether accounting conservatism facilitates or detracts from the efficiency of debt contracting. We consider both “unconditional” and “conditional” conservatism as discussed in the literature. In both cases, our analysis does not support the positive relationship between accounting conservatism and the efficiency of debt contracting, as suggested by Watts [2003], and as hypothesized in numerous empirical studies. In fact, we find the opposite can be true. Under very plausible conditions, we find that accounting conservatism, that affects the information content of accounting reports, actually decreases the efficiency of debt contracts.
Our notion of efficiency is derived endogenously from the joint optimality of the debt covenant and the corresponding interest rate on debt, but differs from the usual efficiency notion assumed in empirical studies. We also develop and use a statistical characterization of conservatism that is new in the literature but which, we believe, is consistent with the widely accepted definition of accounting conservatism originally proposed by Basu [1997].