The problem of credit rationing/borrowing constraint has recently received considerable attention. Individuals who are denied any credit by a financial institution, or who find it difficult to borrow against future incomes, are said to be credit rationed or borrowing constrained in the credit markets. This dissertation tries to identify the circumstances under which individuals may be rationed (or not), and analyses the actions undertaken to overcome future constraints.

Chapter 2 analyses the problem of credit rationing as it arises in equilibrium, when borrowers differ with respect to their demands for loans. It is shown that if the principal can costlessly observe the agent’s type, then (i) the agents who meet the collateral requirements are not rationed in the sense of Stiglitz-Weiss (1981), (ii) the agents who do not meet the collateral requirements are rationed in the sense of Jaffee-Russell (1976). We further show that if the principal cannot distinguish between different agents, then the previous rationing results still hold in the second best contract which is pooling : agents of different types pick the same contract.

Chapter 3 analyses the problem of credit rationing as it emerges in a dynamic setting, when a renegotiation of the original contract may be undertaken. It is conjectured that (i) the principal uses the information revealed about an agent’s type at the time of first repayment, to design future contracts, (ii) the agents who show consistently honest behavior are never rationed, (iii) the agents who showed dishonest behavior impose a negative externality on the agents who were honest; they are rationed in later periods.

Finally, in chapter 3, we analyse the role of an exogenously imposed borrowing constraint prompting the individuals to change their life-cycle decisions. This chapter provides an explicit link between human and non-human wealth by making income endogenous through investment in human capital. The chapter also discusses the econometric aspects of the problem: the possible empirical work that can be undertaken in the future using a micro data set.