| The purpose of this dissertation is to explore Argentina's experience with foreign exchange controls between 1985 and 1989, and in particular, the behavior of black market spot and futures exchange rates.;In this respect, two separate issues are addressed. The second chapter deals with short run dynamics and volatility clustering. In contrast to the experience in developed economies, it is shown that volatility clustering is the result of a process which has a more precise or known form: the official exchange rate policy.;The Autoregressive Conditional Heteroskedasticity Model (ARCH), as developed by Engel (1982), and extended by Bollerslev (1986) is used to show how changes in the conditional mean and variance of the official exchange rate policy affected the depreciation rate in the spot black market. It is shown that in general, changes in that process introduced at the same time as financial markets were closed, tended to reduce the mean return and volatility in the spot black market rate.;The second issue explored is the market efficiency or unbiasedness hypothesis. This theory states that in an efficient market, the futures exchange rate is an unbiased predictor of the future spot rate.;This hypothesis has been rejected in the literature. One interpretation for the existence of excess returns states that this is enough evidence of market inefficiency, in the sense that agents are not exploiting all the available information (Bilson 1981, Mankiw and Summers (1984)). An alternative one, (see Hansen and Hodrich (1983) and Domowitz and Hakkio (1985)) is that risk averse agents, when facing uncertainty, require a premium to hold risky currency forward, thus driving a wedge between futures and expected spot exchange rates.;The Argentine evidence shows that the futures exchange rate has predicted accurately the direction in which the spot exchange rate has changed. However, the futures rate has been a biased predictor in accordance to the evidence in developed economies. The source of bias is shown to depend, among other things, on the specific exchange rate regime which prevailed during this period.;So, in this respect, an optimizing asset pricing model is presented and different representations for the risk premium are derived. The Generalized Method of Moments, as developed by Hansen (1982), is used to recover those parameters which characterize preferences. Evidence of risk aversion is found when a more disaggregated measure of monthly consumption is used. Finally, the model is numerically solved with the Parameterized Expectations algorithm, as developed by Marcet (1989), with the purpose of verifying how well a nonlinear rational expectations model replicates some of the properties observed in Argentine macroeconomic data. |
