Search

This paper shows that bailouts of private agents can optimally take the form of asset purchases, even if this also means paying off external asset holders, in the presence of borrowing constraints and asymmetric information on liquidity needs. The combination of these two ingredients make direct compensation through loans and/or net transfers imperfect. Thus, when more constrained agents are also more exposed to the asset, the compensation through asset purchases becomes desirable. Anticipating these purchases, private agents engage in a collective bet on the defaulting asset, leading to an equilibrium implicit guarantee, where even an intrinsically worthless asset can be traded at a positive price.

Coordination failure among owners of heterogeneous debt types increases distress costs. Covenants reduce expected distress costs by lowering the probability of liquidity shortages, increasing liquidation values, and incentivizing creditor monitoring. We predict and find that new debt contracts include more covenants when borrowers' existing debt structures are more heterogeneous. Our findings suggest that covenants are not only used to address creditor-shareholder conflicts but also to reduce the expected costs of coordination failure among creditors. Further, our results indicate a dynamic component missing from static debt structure models: Debt heterogeneity entails additional covenants (i.e., constraints) when raising future debt.

We build a dynamic model of investment and financing decisions to study the choice between bonds and bank loans in a firm's marginal financing decision and its effects on corporate investment. We show that firms with more growth options, higher bargaining power in default, operating in more competitive product markets, and facing lower credit supply are more likely to issue bonds. We also demonstrate that, by changing the cost of financing, these characteristics affect the timing of investment. We test these predictions using a sample of U.S. firms and present new evidence which is strongly supportive of our theory.