Essays on overlending and capital flight.

The dissertation has two parts. The first part argues that overlending by international banks to LDC borrowers prior to 1982 should be understood as the result of profit maximizing behavior under specific circumstances. In the credit market model developed, creditors assign a positive probability to an ex post involvement by a third agent in a formally two agent contract. This results in an ex ante Pareto inefficient amount of credit extended. The deviation from the efficient amount is defined as overlending and it increases as the perceived probability of ex post intervention rises. Under certain conditions, a positive probability leads to projects being completely debt financed and the adoption of the riskiest project available. The empirical investigation identifies three types of international credit market relationships. For each of these types, variables were identified that contained information relevant to the formation of an expectation of the probability of intervention by a third party. Descriptive statistics and regressions were used to investigate the relationship between those chosen variables (as proxies for the probability of bailout) and the amount of lending undertaken. This study also discusses actions undertaken by the U.S. government and multi-lateral organizations vis-a-vis the LDC debt crisis. The policies implemented seem to have amounted to relatively small transfers of resources from (mostly) OECD taxpayers to the banks with LDC exposure and transfers of risk from these banks to the taxpayers. A number of alternative Congressional and other proposals were also analyzed; nearly all alternative proposals would involve such transfers. The study also discusses LDC government policies toward private sector debt owed to the banks and the issue of policies and reforms needed to avoid a repetition of overlending and the moral hazard behind it. The second part addresses the phenomenon of LDC residents investing abroad at the same time they are borrowing from abroad and investing in LDC projects. The conditions under which this phenomenon takes place are investigated using: (i) a variant of the st and ard intertemporal optimization model assuming full information and (ii) a principal-agent model which assumes that effort expended on projects is unobservable. Both models are used to discuss the behavior of capital flight, investment of savings in the LDC, LDC capital formation, external private borrowing and foreign loan interest rates under different shocks. Moreover, different policies aimed at reducing or eliminating capital flight are discussed.PhDFinanceEconomicsUniversity of Michiganhttp://deepblue.lib.umich.edu/bitstream/2027.42/162449/1/9013904.pd