Modelling Official And Parallel Exchange Rates In Colombia Under Alternative Regimes: A Non-Linear Approach

In this paper we look at the long-run relationship between the parallel and the official exchange rate in Colombia over two regimes; a crawling peg period and a more flexible crawling band one. The existence of cointegration between the parallel and official exchange rates in Colombia is consistent with previous findings for other developing economies, and offers support for the view that the parallel market for foreign exchange is not informationally efficient, because past values of the two rates (and of the disequilibrium error) could be used for forecasting the parallel exchange rate. The fact that the parallel rate cointegrates with the official one also implies that the latter has a role to play in the evolution of the former. This should be kept in mind when the monetary authorities affect with their decisions the behaviour of the official exchange rate. Further, we look at the short-run adjustment process of the parallel rate both in a linear and a non-linear context. According to the empirical results, there is strong evidence in favour of non-linear adjustment over the crawling peg but not over the crawling band period. This should not come as a surprise. The first period has witnessed the operation of strict foreign controls that have caused distortions in the transition back to equilibrium, once disequilibrium has occurred. The non-linear adjustment reported in the paper, provides an empirical evidence of the complicated structure under which the exchange rate market operated. With the abolition of exchange rate controls and the introduction of more flexibility in the exchange rate market over the second period, these distortions have gradually been eliminated. As a result, the transition back to equilibrium does not longer seem to exhibit any complicated non-linear structure. Thus, the modelling exercise has showed that the change from the crawling peg exchange rate regime to a crawling band one did not affect the long-run equilibrium relationship between the official and parallel exchange rates in Colombia, but changed radically the short-run dynamics.