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
Trade, Misallocation, and Capital Market Integration
Laszlo Tetenyi
IWH-CompNet Discussion Papers,
No. 8,
2019
Abstract
I study how cross-country capital market integration affects the gains from trade in a model with financial frictions and heterogeneous, forward-looking firms. The model predicts that misallocation among exporters increases as trade barriers fall, even as misallocation decreases in the aggregate. The reason is that financially constrained productive exporters increase their production only marginally, while unproductive exporters survive for longer and increase their size. Allowing capital inflows magnifies misallocation, because unproductive firms expand even more, leading to a decline in aggregate productivity. Nevertheless, under integrated capital markets, access to cheaper capital dominates the adverse effect on productivity, leading to higher output, consumption and welfare than under closed capital markets. Applied to the period of European integration between 1992 and 2008, I find that underdeveloped sectors experiencing higher export exposure had more misallocation of capital and a higher share of unproductive firms, thus the data is consistent with the model’s predictions. A key implication of the model is that TFP is a poor proxy for consumption growth after trade liberalisation.
Read article
Should Forecasters Use Real‐time Data to Evaluate Leading Indicator Models for GDP Prediction? German Evidence
Katja Heinisch, Rolf Scheufele
German Economic Review,
No. 4,
2019
Abstract
In this paper, we investigate whether differences exist among forecasts using real‐time or latest‐available data to predict gross domestic product (GDP). We employ mixed‐frequency models and real‐time data to reassess the role of surveys and financial data relative to industrial production and orders in Germany. Although we find evidence that forecast characteristics based on real‐time and final data releases differ, we also observe minimal impacts on the relative forecasting performance of indicator models. However, when obtaining the optimal combination of soft and hard data, the use of final release data may understate the role of survey information.
Read article
02.10.2019 • 20/2019
Joint Economic Forecast Autumn 2019: Economy Cools Further – Industry in Recession
Berlin, October 2, 2019 – Germany’s leading economics research institutes have revised their economic forecast for Germany significantly downward. Whereas in the spring they still expected gross domestic product (GDP) to grow by 0.8% in 2019, they now expect GDP growth to be only 0.5%. Reasons for the poor performance are the falling worldwide demand for capital goods – in the exporting of which the Germany economy is specialised – as well as political uncertainty and structural changes in the automotive industry. By contrast, monetary policy is shoring up macroeconomic expansion. For the coming year, the economic researchers have also reduced their forecast of GDP growth to 1.1%, having predicted 1.8% in the spring.
Oliver Holtemöller
Read
Pricing Sin Stocks: Ethical Preference vs. Risk Aversion
Stefano Colonnello, Giuliano Curatola, Alessandro Gioffré
European Economic Review,
2019
Abstract
We develop an ethical preference-based model that reproduces the average return and volatility spread between sin and non-sin stocks. Our investors do not necessarily boycott sin companies. Rather, they are open to invest in any company while trading off dividends against ethicalness. When dividends and ethicalness are complementary goods and investors are sufficiently risk averse, the model predicts that the dividend share of sin companies exhibits a positive relation with the future return and volatility spreads. An empirical analysis supports the model’s predictions. Taken together, our results point to the importance of ethical preferences for investors’ portfolio choices and asset prices.
Read article
Predicting Free-riding in a Public Goods Game – Analysis of Content and Dynamic Facial Expressions in Face-to-Face Communication
Dmitri Bershadskyy, Ehsan Othman, Frerk Saxen
IWH Discussion Papers,
No. 9,
2019
Abstract
This paper illustrates how audio-visual data from pre-play face-to-face communication can be used to identify groups which contain free-riders in a public goods experiment. It focuses on two channels over which face-to-face communication influences contributions to a public good. Firstly, the contents of the face-to-face communication are investigated by categorising specific strategic information and using simple meta-data. Secondly, a machine-learning approach to analyse facial expressions of the subjects during their communications is implemented. These approaches constitute the first of their kind, analysing content and facial expressions in face-to-face communication aiming to predict the behaviour of the subjects in a public goods game. The analysis shows that verbally mentioning to fully contribute to the public good until the very end and communicating through facial clues reduce the commonly observed end-game behaviour. The length of the face-to-face communication quantified in number of words is further a good measure to predict cooperation behaviour towards the end of the game. The obtained findings provide first insights how a priori available information can be utilised to predict free-riding behaviour in public goods games.
Read article
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
Accounting Quality in Banking: The Role of Regulatory Interventions
Manthos D. Delis, Iftekhar Hasan, Maria Iosifidi, Lingxiang Li
Journal of Banking and Finance,
2018
Abstract
Using the full sample of U.S. banks and hand-collected data on enforcement actions over 2000–2014, we analyze the role of these interventions in promoting several aspects of accounting quality. We find that enforcement actions issued for both risk-related and accounting-related reasons lead to significant improvements in accounting quality. This improvement is consistently found for earnings smoothing, big-bath accounting, timely recognition of future loan losses, the association of loan loss provisions with future loan charge offs, loss avoidance, and cash flow predictability and earnings persistence. Most of the effects are somewhat more potent in the crisis period and survive in several sensitivity tests. Our findings highlight the imperative role of regulatory interventions in promoting bank accounting quality.
Read article
Central Bank Transparency and the Volatility of Exchange Rates
Stefan Eichler, Helge Littke
Journal of International Money and Finance,
2018
Abstract
We analyze the effect of monetary policy transparency on bilateral exchange rate volatility. We test the theoretical predictions of a stylized model using panel data for 62 currencies from 1998 to 2010. We find strong evidence that an increase in the availability of information about monetary policy objectives decreases exchange rate volatility. Using interaction models, we find that this effect is more pronounced for countries with a lower flexibility of goods prices, a lower level of central bank conservatism, and a higher interest rate sensitivity of money demand.
Read article
Connecting to Power: Political Connections, Innovation, and Firm Dynamics
Ufuk Akcigit, Salomé Baslandze, Francesca Lotti
NBER Working Paper,
No. 25136,
2018
Abstract
How do political connections affect firm dynamics, innovation, and creative destruction? To answer this question, we build a firm dynamics model, where we allow firms to invest in innovation and/or political connection to advance their productivity and to overcome certain market frictions. Our model generates a number of theoretical testable predictions and highlights a new interaction between static gains and dynamic losses from rent-seeking in aggregate productivity. We test the predictions of our model using a brand-new dataset on Italian firms and their workers, spanning the period from 1993 to 2014, where we merge: (i) firm-level balance sheet data; (ii) social security data on the universe of workers; (iii) patent data from the European Patent Office; (iv) the national registry of local politicians; and (v) detailed data on local elections in Italy. We find that firm-level political connections are widespread, especially among large firms, and that industries with a larger share of politically connected firms feature worse firm dynamics. We identify a leadership paradox: when compared to their competitors, market leaders are much more likely to be politically connected, but much less likely to innovate. In addition, political connections relate to a higher rate of survival, as well as growth in employment and revenue, but not in productivity – a result that we also confirm using a regression discontinuity design.
Read article
