Determinants of Exchange Market Pressures in Different Exchange Rate Regimes: Bayesian Model Averaging Evidence

The purpose of this paper is to determine the most significant leading indicators of the currency crisis in two groups of countries, countries with the floating exchange rate regimes and countries with the non-floating ones. A unique set data that covers 43 countries and their currency crises during 1999-2014 and 64 indicators are used. Moreover, estimation technique that is robust to model uncertainty, i.e. Bayesian Model Averaging is applied. Two types of the warning systems are used. Therefore, in the first system, warning variables of currency crisis are investigated while in the second ones, determinants of the exchange market pressure volatility are examined. Morever, these systems involve monitoring the evaluation of several leading indicators from the financial, political, structural, trade and other sectors. Overall, the results indicate that Oil price plays a pivotal role as a warning variable in countries with floating exchange rate regimes in both systems. Just the same, for the countries that experienced non-floating exchange rate regime prior to the crisis and during of it, changes in exchange market pressure index is a significant leading indicator of currency crises.