Bank Crisis Management Policies and the New Instability

Abstract

This research focuses on the determinants of banking crises. Despite all efforts to create systems of leading indicators, banking crises still occur, leading to negative consequences not only in the financial but also in the real sector of the economy. Our objective is to figure out how the characteristics of past banking crises, in particular the amount of liquidity support provided by the government, are related to the probability of subsequent banking crises. To accomplish this task, we use a classification tree methodology which we apply to a sample of 56 countries over the period 2000–2021. In addition to the characteristics of past crisis episodes, we add to the list of potential predictors of banking crises indicators related to the depth, structure and efficiency of the banking sector; institutional and macroeconomic indicators; and the index of macroprudential policy. To test the robustness of our results, we perform variable selection through Bayesian model averaging. Our results suggest that liquidity support during the past banking crisis is important for future crises. We demonstrate that excessive government support aimed at stabilizing the system during a crisis may affect future banking sector stability mainly by increasing the level of moral hazard in the financial system.

Keywords

• classification tree
• banking crisis
• bayesian model averaging
• moral hazard