Abstract
Has the global economic crisis resulted in countries shifting their exchange rate regimes and, if so, in what way? Focusing on the relevant period of 2008–2012, and using the IMF’s Annual Report on Exchange Arrangements and Exchange Restrictions classification of exchange rate regimes and database, we calculate exchange rate regime transition probabilities and test their statistical significance. Even though there is some evidence of state dependence, in the sense that transitions are relatively infrequent, we do find that these are significant, especially in the direction of fixity. Our testing procedure employs the Wilson (1927) statistic, which is appropriate for drawing inference based on relatively rare events. By examining all transitions in detail, we also find further evidence that countries that shift often flip back to their previous regime.
JEL Classification:
Notes
1 For detailed definitions of the IMF’s exchange rate arrangements see Habermeier et al. (Citation2009).
2 This is a residual, and, hence, less interesting group – the IMF includes all observations that could not be classified into any of the other exchange rate arrangements.
3 The binomial assumption is appropriate, as the potential outcome of each transition, say from k, is binary: there is either a shift towards a given regime, say l with probability pk,l, or towards the remaining regime with probability 1 – pk,l.
4 See Wilson (Citation1927).
5 For example, under the binomial distribution, a Wald confidence interval may assume negative values, as is the case here when we consider transitions from other managed arrangements to flexible regimes at the 1% confidence level. A disadvantage of the Wilson transition probabilities is that they may not necessarily add up to 100%.
6 This procedure is consistent with Beaver et al. (Citation2008).