Assessing the fiscal implications of banking crises

We propose a method for computing the distribution of the potential fiscal cost of a banking crisis - a key input in assessing the adequacy of a country's fiscal buffers. First, we use a cross-section of banking crises to identify the risk factors that predict the post-crisis increase in public sector debt - a measure of the overall fiscal cost of a crisis. Next, we use these risk factors to compute country-specific distributions of that cost in the event of a crisis. We find that the level and growth of credit to the private non-financial sector, foreign exchange reserves and the ratio of bank capital to assets are relevant predictors. As an illustration, we apply the method to the conditions prevailing in 2018 and find that the potential fiscal costs could be sizeable: with a probability of 95%, public debt could approach 40% of GDP on average across countries. Our illustrative estimates are probably upper bounds: while they indicate that higher bank capital can substantially reduce fiscal costs, they exclude the broader benefits of the wide-ranging reforms after the Great Financial Crisis.