Search results

This paper studies the impact of cyclical systemic risk on future bank profitability for a large representative panel of EU banks between 2005 and 2017. Using linear local projections we show that high current levels of cyclical systemic risk predict large drops in the average bank-level return on assets (ROA) with a lead time of 3-5 years. Based on quantile local projections we further show that the negative impact of cyclical systemic risk on the left tail of the future bank-level ROA distribution is an order of magnitude larger than on the median. Given the tight link between negative profits and reductions in bank capital, our method can be used to quantify the level of "Bank capital-at-risk" for a given banking system, akin to the concept of "Growth-at-risk". We illustrate how the method can inform the calibration of countercyclical macroprudential policy instruments.

This paper proposes a semi-structural approach to identifying excessive household credit developments. Using an overlapping generations model, a normative trend level for the real household credit stock is derived that depends on four fundamental economic factors: real potential GDP, the equilibrium real interest rate, the population share of the middle-aged cohort, and institutional quality. Semi-structural household credit gaps are obtained as deviations of the real household credit stock from this fundamental trend level. Estimates of these credit gaps for 12 EU countries over the past 35 years yield long credit cycles that last between 15 and 25 years with amplitudes of around 20%. The early warning properties for financial crises are superior compared to credit gaps that are obtained from purely statistical filters. The proposed semistructural household credit gaps could therefore provide useful information for the formulation of countercyclical macroprudential policy, especially because they allow for economic interpretation of observed credit developments.

This paper proposes a framework for deriving early-warning models with optimal out-of-sample forecasting properties and applies it to predicting distress in European banks. The main contributions of the paper are threefold. First, the paper introduces a conceptual framework to guide the process of building early-warning models, which highlights and structures the numerous complex choices that the modeler needs to make. Second, the paper proposes a flexible modeling solution to the conceptual framework that supports model selection in real-time. Specifically, our proposed solution is to combine the loss function approach to evaluate early-warning models with regularized logistic regression and cross-validation to find a model specification with optimal real-time out-of-sample forecasting properties. Third, the paper illustrates how the modeling framework can be used in analysis supporting both microand macro-prudential policy by applying it to a large dataset of EU banks and showing some examples of early-warning model visualizations.

This paper presents a new database for financial crises in European countries, which serves as an important step towards establishing a common ground for macroprudential oversight and policymaking in the EU. The database focuses on providing precise chronological definitions of crisis periods to support the calibration of models in macroprudential analysis. An important contribution of this work is the identification of financial crises by combining a quantitative approach based on a financial stress index with expert judgement from national and European authorities. Key innovations of this database are (i) the inclusion of qualitative information about events and policy responses, (ii) the introduction of a broad set of non-exclusive categories to classify events, and (iii) a distinction between event and post-event adjustment periods. The paper explains the two-step approach for identifying crises and other key choices in the construction of the dataset. Moreover, stylised facts about the systemic crises in the dataset are presented together with estimations of output losses and fiscal costs associated with these crises. A preliminary assessment of the performance of standard early warning indicators based on the new crises dataset confirms findings in the literature that multivariate models can improve compared to univariate signalling models.

This paper presents a new database for financial crises in European countries, which serves as an important step towards establishing a common ground for macroprudential oversight and policymaking in the EU. The database focuses on providing precise chronological definitions of crisis periods to support the calibration of models in macroprudential analysis. An important contribution of this work is the identification of financial crises by combining a quantitative approach based on a financial stress index with expert judgement from national and European authorities. Key innovations of this database are (i) the inclusion of qualitative information about events and policy responses, (ii) the introduction of a broad set of non-exclusive categories to classify events, and (iii) a distinction between event and post-event adjustment periods. The paper explains the two-step approach for identifying crises and other key choices in the construction of the dataset. Moreover, stylised facts about the systemic crises in the dataset are presented together with estimations of output losses and fiscal costs associated with these crises. A preliminary assessment of the performance of standard early warning indicators based on the new crises dataset confirms findings in the literature that multivariate models can improve compared to univariate signalling models.