Capital versus Performance Covenants in Debt Contracts

In the paper, Capital versus Performance Covenants in Debt Contracts, which was recently made publicly available on SSRN, we propose a simple classification of financial covenants into two distinct groups: performance covenants and capital covenants. Performance covenants rely on measures of profitability and efficiency whereas capital covenants rely on information about sources and uses of capital, i.e., balance sheet information. We argue that capital covenants align incentives between borrowers and lenders by limiting the amount of debt in the borrower’s capital structure. In contrast, performance covenants act as tripwires that transfer control to lenders when performance deteriorates and hence incentive conflicts between shareholders and lenders become more acute.

We document that performance covenants and capital covenants are negatively correlated and used by borrowers in different situations. More importantly, performance covenants are used in conjunction with negative covenants that place direct restrictions on certain managerial actions, whereas capital covenants are not, in line with capital covenants’ incentive alignment effects. In addition, performance covenants are significant predictors of contract renegotiations, consistent with performance covenants acting as tripwires, whereas capital covenants are not.

We also examine how the contractibility of accounting information affects the choice of covenant mix. We find that performance covenants are preferred to capital covenants when accounting information is a good summary measure of credit risk. This result is consistent with our hypothesis that capital covenants are a more robust mechanism to address underlying agency problems when accounting information gives a poor description of credit risk, and suggests that the contractibility of accounting information has an important effect on the design of covenant packages. This finding suggests in turn that contractibility of accounting information has an important effect on credit market access. In line with this conjecture, we observe that the use of tripwire covenants is positively correlated with the level of debt.

Finally, it needs to be pointed out that covenants and other contract features are chosen simultaneously. Therefore, our results do not allow causal inferences to be drawn regarding the effects of firm and contract characteristics on the choice of covenants; the effects should be interpreted as associations. Identification of the causal effects (which requires exogenous instruments that are not readily available) is outside the scope of this paper.

The full paper is available for download here.