Does Machine Learning Help us Predict Banking Crises?
Johannes Beutel, Sophia List, Gregor von Schweinitz
Journal of Financial Stability,
December
2019
Abstract
This paper compares the out-of-sample predictive performance of different early warning models for systemic banking crises using a sample of advanced economies covering the past 45 years. We compare a benchmark logit approach to several machine learning approaches recently proposed in the literature. We find that while machine learning methods often attain a very high in-sample fit, they are outperformed by the logit approach in recursive out-of-sample evaluations. This result is robust to the choice of performance metric, crisis definition, preference parameter, and sample length, as well as to using different sets of variables and data transformations. Thus, our paper suggests that further enhancements to machine learning early warning models are needed before they are able to offer a substantial value-added for predicting systemic banking crises. Conventional logit models appear to use the available information already fairly efficiently, and would for instance have been able to predict the 2007/2008 financial crisis out-of-sample for many countries. In line with economic intuition, these models identify credit expansions, asset price booms and external imbalances as key predictors of systemic banking crises.
Read article
19.09.2019 • 19/2019
Long-term effects of privatisation in eastern Germany: award-winning US economist begins large-scale research project at the IWH
It is one of the most prestigious awards in the German scientific community: the Max Planck-Humboldt Research Award 2019 endowed with €1.5 million goes to Ufuk Akcigit, Professor of Economics at the University of Chicago. At the Halle Institute for Economic Research (IWH), Akcigit aims to use innovative methods to investigate why the economy in eastern Germany is still lagging behind that in western Germany – and what role the privatisation process 30 years ago played in this.
Reint E. Gropp
Read
An Evaluation of Early Warning Models for Systemic Banking Crises: Does Machine Learning Improve Predictions?
Johannes Beutel, Sophia List, Gregor von Schweinitz
Abstract
This paper compares the out-of-sample predictive performance of different early warning models for systemic banking crises using a sample of advanced economies covering the past 45 years. We compare a benchmark logit approach to several machine learning approaches recently proposed in the literature. We find that while machine learning methods often attain a very high in-sample fit, they are outperformed by the logit approach in recursive out-of-sample evaluations. This result is robust to the choice of performance measure, crisis definition, preference parameter, and sample length, as well as to using different sets of variables and data transformations. Thus, our paper suggests that further enhancements to machine learning early warning models are needed before they are able to offer a substantial value-added for predicting systemic banking crises. Conventional logit models appear to use the available information already fairly effciently, and would for instance have been able to predict the 2007/2008 financial crisis out-of-sample for many countries. In line with economic intuition, these models identify credit expansions, asset price booms and external imbalances as key predictors of systemic banking crises.
Read article
The Privatisation Activities of the Treuhandanstalt and the Transformation of the East German Corporate Landscape: A New Dataset for First Explorations
Alexander Giebler, Michael Wyrwich
IWH Technical Reports,
No. 1,
2018
Abstract
Even nearly 30 years after the fall of the Berlin Wall, the privatisation and transformation of East Germany's business landscape is controversially discussed in the media and politics. The privatisation process led to enormous structural changes, which were associated with massive job losses. In particular, the stagnating regional development of East Germany is often blamed on the “long shadow” of the privatisation activities of the Treuhandanstalt (THA). From a scientific perspective, however, there are hardly any contributions dealing with the effects of privatisation activities. The IWH-Treuhand Privatisation Micro Database introduced in this technical report is novel as such that it provides comprehensive information on employment and turnover figures for formerly state-owned enterprises for the early 1990s.
Read article
Within Gain, Structural Pain: Capital Account Liberalization and Economic Growth
Xiang Li, Dan Su
New Structural Economics Working Paper No. E2018010,
2018
Abstract
This paper is the first to study the effects of capital account liberalization on structural transformation and compare the contribution of within term and structural term to economic growth. We use a 10-sector-level productivity dataset to decomposes the effects of opening capital account on within-sector productivity growth and cross-sector structural transformation. We find that opening capital account is associated with labor productivity and employment share increment in sectors with higher human capital intensity and external financial dependence, as well as non-tradable sectors. But it results in a growth-reducing structural transformation by directing labor into sectors with lower productivity. Moreover, in the ten years after capital account liberalization, the contribution share of structural transformation decreases while that of within productivity growth increases. We conclude that the relationship between capital account liberalization and economic growth is within gain and structural pain.
Read article
15.02.2018 • 1/2018
Presseeinladung: „Von der Transformation zur Europäischen Integration: Optimieren durch Evaluieren – Wirtschaftsförderung im Qualitätscheck“
Unter dem Titel „Von der Transformation zur Europäischen Integration: Optimieren durch Evaluieren – Wirtschafts-förderung im Qualitätscheck“ präsentiert das Leibniz-Institut für Wirtschaftsforschung Halle (IWH) am Mittwoch, dem 21. Februar 2018 gemeinsam mit Wissenschaftlerinnen und Wissenschaftlern anderer Forschungsinstitute sowie Universitäten Forschungsergebnisse zu verschiedenen Aspekten der Evaluation von Wirtschaftsfördermaßnahmen.
Read
Von der Transformation zur europäischen Integration:
Wachstumsfaktor Bildung besser nutzen – ein Tagungsbericht
Gerhard Heimpold
Wirtschaft im Wandel,
No. 2,
2017
Abstract
Unter dem Titel „Von der Transformation zur europäischen Integration: Wachstumsfaktor Bildung besser nutzen“ hat das Leibniz-Institut für Wirtschaftsforschung Halle (IWH) gemeinsam mit Partnern aus Forschungseinrichtungen und Universitäten in Deutschland am 22. Februar 2017 Forschungsergebnisse zur besseren Nutzung von Bildung als Wachstumsfaktor vorgestellt und diskutiert. Der Präsident des IWH, Professor Reint E. Gropp, Ph.D., unterstrich, dass es Investitionen in Humankapital seien, die langfristig das Wirtschaftswachstum treiben. Andere Länder investierten deutlich mehr in Humankapital als Deutschland. Dies sollte zu denken geben.
Read article
15.02.2017 • 11/2017
Presseeinladung: „Von der Transformation zur Europäischen Integration: Wachstumsfaktor Bildung besser nutzen“
Unter dem Titel „Von der Transformation zur Europäischen Integration: Wachstumsfaktor Bildung besser nutzen“ wird das Leibniz-Institut für Wirtschaftsforschung Halle (IWH) am Mittwoch, dem 22. Februar 2017 gemeinsam mit Wissenschaftlerinnen und Wissenschaftlern aus mit dem IWH vernetzten Instituten und Universitäten Forschungsergebnisse zu verschiedenen Aspekten des Wachstumsfaktors Bildung präsentieren.
Read
Von der Transformation zur europäischen Integration: Auf dem Weg zu mehr Wachstumsdynamik – ein Tagungsbericht
Gerhard Heimpold
Wirtschaft im Wandel,
No. 2,
2016
Abstract
Unter dem Titel „Von der Transformation zur europäischen Integration: Auf dem Weg zu mehr Wachstumsdynamik“ hat das Leibniz-Institut für Wirtschaftsforschung Halle (IWH) gemeinsam mit Partnern aus Universitäten in Mitteldeutschland am 22. Februar 2016 Forschungsergebnisse zu den Folgen des Strukturwandels, zur Erzielung von mehr Wachstumsdynamik und den wirtschaftspolitischen Rahmenbedingungen hierfür präsentiert.
Read article
