Abstract:
We propose a model of exchange-rate regime choice which accounts for the existence of a continuous range of regimes, the need for real exchange-rate adjustment in response to shocks, the existence of capital account shocks and of balance-sheet effects, the sensitivity of prices to the nominal exchange rate, and the need for a commitment to make any given regime sustainable. Non-ordered Logit estimations on a cross-section sample of 126 emerging and developed countries before and after 1997-1998 currency crises broadly support our approach.. Specifically, we find that there is still a case for intermediate regimes in countries where the interest rate channel is weaker and which do not depend too much on commodities. The empirical model correctly predicts up to 83% of observed exchange rate regimes, and the recent "hollowing out" of intermediate regimes. It also provides a benchmark to assess the recent changes in individual exchange rate regimes.