IMF Programs, Shadow Banks and Financial Crisis: Empirical Evidence

This chapter presents in detail our empirical analysis employed to test Hypothesis 1 which posits that IMF programs are likely to be positively associated with the “twin” outbreak of currency and sudden reversal crisis—that is, financial crisis—in Fund-assisted developing states in which the market concentration of domestic private shadow banks is high. We begin by discussing the time-series cross-sectional (TSCS) sample of 72 developing countries from 1985 to 2015 used to test Hypothesis 1. We discuss the operationalization of the binary dependent variable in this hypothesis—the joint outbreak of sudden foreign portfolio investment reversals (“sudden reversal crisis”) and “currency crisis.” Put together, the concurrent outbreak of these two types of financial crises constitutes the binary financial crisis dependent variable employed to test our first hypothesis. We then describe the operationalization of the independent variable: the market concentration of domestic private shadow banks. We then present the results from estimation of various probit and bivariate probit models that we use to evaluate Hypothesis 1. Overall, these models account for a variety of econometric challenges that emerge when testing the first hypothesis. The statistical results in these models strongly support Hypothesis 1.