Professors Lai and Chang have provided valuable extensions to our analysis of exchange market unification. The use of the “terminal curve” AA in the graphical presentation of the unification process shows clearly the inverse relationship between the reform date and the unified exchange rate. Their contribution shows that the dynamic predictions of our model are potentially richer than we originally indicated.
However, having shown that almost anything can happen—depending on the position of the terminal curve AA and the reform date T—makes a comparison of the predictions of the model with the empirical evidence on exchange market unification all the more difficult. An examination of the most probable outcomes would therefore prove useful. The first issue is to determine the likely position of the terminal curve AA. Using the results established in our paper, it can be shown that the condition under which the curve AA is steeper than the pre-reform saddlepath is given by
If θ, the share of foreign reserves in the domestic money stock, is small—a likely situation for most developing countries contemplating exchange rate unification—this condition is likely to hold for reasonably high values of α (the interest elasticity of money demand), γ, and v. Thus, in practice, the parallel market exchange rate is likely to jump upward upon announcement of the reform date, as emphasized in our paper.
The second issue that requires further investigation in the context of the unification process relates to the length of the transition period. In the context of the extended model presented in the second part of our paper, there exists a logical way of determining endogenously the optimal date of reform. Essentially, the idea is to minimize a loss function that highlights the trade-off involved in continuing the “old” regime—which entails a variety of costs, such as price distortions and diversion of export remittances to the parallel market—and implementing an immediate reform, because of the real costs that a short transition period or an “overnight” float may entail. As shown in the second part of our paper, the immediate increase in domestic wages and prices that takes place when the unified rate is expected to be more depreciated than the initial parallel exchange rate leads to an appreciation of the real exchange rate during the transition period, and a fall in output. A formal solution to this problem would further extend our understanding of exchange rate reforms.