Long-Term Financial Contracts can Mitigate the Adverse Selection Problem in Project Financing

Abstract

This paper shows that long-term contracts can be used in competitive financial markets to separate entrepreneurs of different abilities. In equilibrium poor entrepreneurs are financed with a sequence of standard-debt contracts. Good entrepreneurs are financed with a modified contract, in which the terms of the second part of the contract are contingent upon whether default is observed at the first date. Sorting is achieved through the contingent term in the contract. In equilibrium good entrepreneurs will typically pay high interest rates to start with, followed by relatively lower rates later if they are successful. The solution in the paper is contrasted with the use of collateral. Incomplete information plays an important role in explaining the existence and form of enduring relationships. When agents have private information and cannot precommit to future actions and announcements there may be a role for multiperiod contracts (see Townsend 1982 and Hart and Holmstrom 1986). Such contracts tie future contractual terms to current behaviour. Moreover, these contracts can be self-enforcing, simply relying upon the self interest of agents to see them through. This paper shows that long-term relationships may help overcome adverse selection problems. The basic idea is very simple. Entrepreneurs have different probabilities of completing successive projects successfully. Observation of the returns cannot reveal the type of project. However, repeated observation of returns does give information about the type of project, by the law of large numbers. Even if there are only two observations at successive dates it may be possible to separate types by making the terms of the contract at the second date conditional upon whether default is observed at the first. We consider an entrepreneur who is risk neutral and who has a project at each of two successive dates. The projects may require outside finance from banks. Risk neutral banks offer two-period credit contracts to the entrepreneur. We assume that there are two types of entrepreneur, one with a high probability of completing projects successfully and a second with a low probability. The probability of success is private information to the entrepreneur, it is not known by the bank. The problem of generalising the analysis to more entrepreneur types is discussed in the conclusion. In Section 2 of the paper it is shown that in this setting an adverse selection problem may arise. We consider a sequence of standard-debt contracts which specify a fixed payment at each date. The important point to note here is that with