Optimal debt contracts and moral hazard along the business cycle

Abstract

We analyze the Pareto optimal contracts between lenders and borrowers in a model with asymmetric information. The model generalizes the Rothschild-Stiglitz pure adverse selection problem by including moral hazard. Entrepreneurs with unequal “abilities” borrow to finance alternative investment projects which differ in degree of risk and productivity. We determine the endogenous distribution of projects as functions of the amount of loanable funds, when lenders have no information about borrowers’ ability and technological choices. Then, we embed these results in a dynamic competitive economy and show that the average quality of the selected projects in equilibrium may be high in recessions and low in booms. This phenomenon may generate (a) multiple steady states, (b) a smaller impact of exogenous shocks on output relative to the full information case, (c) endogenous fluctuations

Endogenous Credit Cycles

“…In terms of informational asymmetry, we will follow [10] and [31] by assuming that borrowers' types cannot be observed either directly or through realized project returns. Hence, all agents other than the owner of the project can only verify whether the latter was successful or not.…”

“…Convexity may then restored by the introduction of random contracts, which are defined as a probability ω ∈ [0, 1] together with a pair of pooling and separating contracts. To be more precise, we follow [31] and introduce random contracts as follows:…”

“…First of all, investment is allocated inefficiently across technologies. Moreover, as (31) illustrates, an increase in wages will affect not just the total amount of investment but also its allocation between sectors, since β will also change as credit expands. When an increase in collateralizable wealth cannot be fully absorbed by G − type entrepreneurs without violating incentive compatibility, the interest rate must fall and B − type investment must expand in order to restore market clearing.…”

“…In this sense, we believe that credit cycles in our framework arise through a natural mechanism which does not rest on the fact that net worth is inconsequential for determining equilibrium levels of investment (as in [6] or [31]) or on assuming that it behaves in a counter-cyclical manner (as in [39]). In our model, booms are investment-driven and are made possible because good projects cross-subsidize their less productive counterparts, a feature which also seems consistent with evidence regarding bank lending standards along the business cycle (see [4]).…”

“…Besides the aforementioned considerations regarding net worth, it is also the case that our mechanism does not rely on particular relative price changes 3 or on assumptions regarding the mix of adverse selection and moral hazard. There are many similarities between our model and that of Reichlin and Siconolfi [31], for example: the latter, however, relies on a particular mix between these two types of asymmetric information, which ultimately makes it difficult to identify the underlying assumptions that generate different effects. Moreover, the source of our reversion mechanism, regime switches between pooling and separating contracts in a context of multi-dimensional screening, is novel and interesting in its own right.…”