Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Research has suggested that a rapid pace of nonfinancial borrowing reliably precedes financial crises, placing the pace of debt growth at the center of frameworks for the deployment of macroprudential policies. I reconsider the role of asset-prices and current account deficits as leading indicators of financial crises. Run-ups in equity and house prices and a widening of the current account deficit have substantially larger (and more statistically-significant) effects than debt growth on the probability of a financial crisis in standard crisis-prediction models. The analysis highlights the value of graphs of predicted crisis probabilities in an assessment of predictors. Nonetheless, the economic and statistical significance of the results is much stronger when the set of crises around the world in the late 2000s are included in the analysis, highlighting challenges in empirical analyses of (infrequent) financial crises.