Evaluating Concentration and Distribution Measures of IMF Financial Soundness Indicators

Abstract

This paper is a first approach to assessing the analytical usefulness of Concentration and Distribution Measures (CDMs) of IMF Financial Soundness Indicators (FSIs) in financial surveillance, using an experimental data collection of the IMF from 36 countries for up to 8 years (2007-2014). Besides illustrating the use made of CDMs in recent policy and academic work, we show econometrically annually over 2008-14 that a range of these CDMs can help to predict system wide vulnerabilities, with appropriate control variables to reduce omitted variable bias. Overall, the exercise lends support to the IMF’s intention to collect CDM data on a regular basis, and supports the argument made in IMF (2013) that CDMs would “allow policy makers and Fund staff to better identify potential build-up of systemic risks, thus providing additional inputs for macro-financial managemen