Forecasting Economic Activity Using a Neural Network in Uncertain Times: Monte Carlo Evidence and Application to the
German GDP
Oliver Holtemöller, Boris Kozyrev
IWH Discussion Papers,
No. 6,
2024
Abstract
In this study, we analyzed the forecasting and nowcasting performance of a generalized regression neural network (GRNN). We provide evidence from Monte Carlo simulations for the relative forecast performance of GRNN depending on the data-generating process. We show that GRNN outperforms an autoregressive benchmark model in many practically relevant cases. Then, we applied GRNN to forecast quarterly German GDP growth by extending univariate GRNN to multivariate and mixed-frequency settings. We could distinguish between “normal” times and situations where the time-series behavior is very different from “normal” times such as during the COVID-19 recession and recovery. GRNN was superior in terms of root mean forecast errors compared to an autoregressive model and to more sophisticated approaches such as dynamic factor models if applied appropriately.
Read article
Supranational Rules, National Discretion: Increasing versus Inflating Regulatory Bank Capital?
Reint E. Gropp, Thomas Mosk, Steven Ongena, Ines Simac, Carlo Wix
Journal of Financial and Quantitative Analysis,
No. 2,
2024
Abstract
We study how banks use “regulatory adjustments” to inflate their regulatory capital ratios and whether this depends on forbearance on the part of national authorities. Using the 2011 EBA capital exercise as a quasi-natural experiment, we find that banks substantially inflated their levels of regulatory capital via a reduction in regulatory adjustments — without a commensurate increase in book equity and without a reduction in bank risk. We document substantial heterogeneity in regulatory capital inflation across countries, suggesting that national authorities forbear their domestic banks to meet supranational requirements, with a focus on short-term economic considerations.
Read article
People
People Doctoral Students PhD Representatives Alumni Supervisors Lecturers Coordinators Doctoral Students Afroza Alam (Supervisor: Reint Gropp ) Julian Andres Diaz Acosta…
See page
People
People Job Market Candidates Doctoral Students PhD Representatives Alumni Supervisors Lecturers Coordinators Job Market Candidates Tommaso Bighelli Job market paper: "The…
See page
Supranational Rules, National Discretion: Increasing versus Inflating Regulatory Bank Capital?
Reint E. Gropp, Thomas Mosk, Steven Ongena, Ines Simac, Carlo Wix
Abstract
We study how higher capital requirements introduced at the supranational and implemented at the national level affect the regulatory capital of banks across countries. Using the 2011 EBA capital exercise as a quasi-natural experiment, we find that affected banks inflate their levels of regulatory capital without a commensurate increase in their book equity and without a reduction in bank risk. This observed regulatory capital inflation is more pronounced in countries where credit supply is expected to tighten. Our results suggest that national authorities forbear their domestic banks to meet supranational requirements, with a focus on short-term economic considerations.
Read article
HIP, RIP, and the Robustness of Empirical Earnings Processes
Florian Hoffmann
Quantitative Economics,
No. 3,
2019
Abstract
The dispersion of individual returns to experience, often referred to as heterogeneity of income profiles (HIP), is a key parameter in empirical human capital models, in studies of life‐cycle income inequality, and in heterogeneous agent models of life‐cycle labor market dynamics. It is commonly estimated from age variation in the covariance structure of earnings. In this study, I show that this approach is invalid and tends to deliver estimates of HIP that are biased upward. The reason is that any age variation in covariance structures can be rationalized by age‐dependent heteroscedasticity in the distribution of earnings shocks. Once one models such age effects flexibly the remaining identifying variation for HIP is the shape of the tails of lag profiles. Credible estimation of HIP thus imposes strong demands on the data since one requires many earnings observations per individual and a low rate of sample attrition. To investigate empirically whether the bias in estimates of HIP from omitting age effects is quantitatively important, I thus rely on administrative data from Germany on quarterly earnings that follow workers from labor market entry until 27 years into their career. To strengthen external validity, I focus my analysis on an education group that displays a covariance structure with qualitatively similar properties like its North American counterpart. I find that a HIP model with age effects in transitory, persistent and permanent shocks fits the covariance structure almost perfectly and delivers small and insignificant estimates for the HIP component. In sharp contrast, once I estimate a standard HIP model without age‐effects the estimated slope heterogeneity increases by a factor of thirteen and becomes highly significant, with a dramatic deterioration of model fit. I reach the same conclusions from estimating the two models on a different covariance structure and from conducting a Monte Carlo analysis, suggesting that my quantitative results are not an artifact of one particular sample.
Read article
Banks Response to Higher Capital Requirements: Evidence from a Quasi-natural Experiment
Reint E. Gropp, Thomas Mosk, Steven Ongena, Carlo Wix
Review of Financial Studies,
No. 1,
2019
Abstract
We study the impact of higher capital requirements on banks’ balance sheets and their transmission to the real economy. The 2011 EBA capital exercise is an almost ideal quasi-natural experiment to identify this impact with a difference-in-differences matching estimator. We find that treated banks increase their capital ratios by reducing their risk-weighted assets, not by raising their levels of equity, consistent with debt overhang. Banks reduce lending to corporate and retail customers, resulting in lower asset, investment, and sales growth for firms obtaining a larger share of their bank credit from the treated banks.
Read article
Bank Response to Higher Capital Requirements: Evidence from a Quasi-natural Experiment
Reint E. Gropp, Thomas Mosk, Steven Ongena, Carlo Wix
Abstract
We study the impact of higher capital requirements on banks’ balance sheets and its transmission to the real economy. The 2011 EBA capital exercise provides an almost ideal quasi-natural experiment, which allows us to identify the effect of higher capital requirements using a difference-in-differences matching estimator. We find that treated banks increase their capital ratios not by raising their levels of equity, but by reducing their credit supply. We also show that this reduction in credit supply results in lower firm-, investment-, and sales growth for firms which obtain a larger share of their bank credit from the treated banks.
Read article
Impulse Response Analysis in a Misspecified DSGE Model: A Comparison of Full and Limited Information Techniques
Sebastian Giesen, Rolf Scheufele
Applied Economics Letters,
No. 3,
2016
Abstract
In this article, we examine the effect of estimation biases – introduced by model misspecification – on the impulse responses analysis for dynamic stochastic general equilibrium (DSGE) models. Thereby, we use full and limited information estimators to estimate a misspecified DSGE model and calculate impulse response functions (IRFs) based on the estimated structural parameters. It turns out that IRFs based on full information techniques can be unreliable under misspecification.
Read article
Effects of Incorrect Specification on the Finite Sample Properties of Full and Limited Information Estimators in DSGE Models
Sebastian Giesen, Rolf Scheufele
Journal of Macroeconomics,
June
2016
Abstract
In this paper we analyze the small sample properties of full information and limited information estimators in a potentially misspecified DSGE model. Therefore, we conduct a simulation study based on a standard New Keynesian model including price and wage rigidities. We then study the effects of omitted variable problems on the structural parameter estimates of the model. We find that FIML performs superior when the model is correctly specified. In cases where some of the model characteristics are omitted, the performance of FIML is highly unreliable, whereas GMM estimates remain approximately unbiased and significance tests are mostly reliable.
Read article
