Forecast Combination and Interpretability Using Random Subspace
Boris Kozyrev
IWH Discussion Papers,
No. 21,
2024
Abstract
This paper investigates forecast aggregation via the random subspace regressions method (RSM) and explores the potential link between RSM and the Shapley value decomposition (SVD) using the US GDP growth rates. This technique combination enables handling high-dimensional data and reveals the relative importance of each individual forecast. First, it is possible to enhance forecasting performance in certain practical instances by randomly selecting smaller subsets of individual forecasts and obtaining a new set of predictions based on a regression-based weighting scheme. The optimal value of selected individual forecasts is also empirically studied. Then, a connection between RSM and SVD is proposed, enabling the examination of each individual forecast’s contribution to the final prediction, even when there is a large number of forecasts. This approach is model-agnostic (can be applied to any set of predictions) and facilitates understanding of how the aggregated prediction is obtained based on individual forecasts, which is crucial for decision-makers.
Read article
Bank Market Power, Factor Reallocation, and Aggregate Growth
R. Inklaar, Michael Koetter, Felix Noth
Journal of Financial Stability,
2015
Abstract
Using a unique firm-level sample of approximately 700,000 firm-year observations of German small and medium-sized enterprises (SMEs), this study seeks to identify the effect of bank market power on aggregate growth components. We test for a pre-crisis sample whether bank market power spurs or hinders the reallocation of resources across informationally opaque firms. Identification relies on the dependence on external finance in each industry and the regional demarcation of regional banking markets in Germany. The results show that bank markups spur aggregate SME growth, primarily through technical change and the reallocation of resources. Banks seem to need sufficient markups to generate the necessary private information to allocate financial funds efficiently.
Read article
Banking Market Competition, Opaque Firms, and the Reallocation Component of Aggregate Growth
R. Inklaar, Michael Koetter, Felix Noth
Abstract
Read article
The Quantity Theory Revisited: A New Structural Approach
Makram El-Shagi, Sebastian Giesen
Abstract
While the long run relation between money and inflation is well established, empirical evidence on the adjustment to the long run equilibrium is very heterogeneous. In this paper we show, that the development of US consumer price inflation between 1960Q1 and 2005Q4 is strongly driven by money overhang. To this end, we use a multivariate state space framework that substantially expands the traditional vector error correction approach. This approach allows us to estimate the persistent components of velocity and GDP. A sign restriction approach is subsequently used to identify the structural shocks to the signal equations of the state space model, that explain money growth, inflation and GDP growth. We also account for the possibility that measurement error exhibited by simple-sum monetary aggregates causes the consequences of monetary shocks to be improperly identified by using a Divisia monetary aggregate. Our findings suggest that when the money is measured using a reputable index number, the quantity theory holds for the United States.
Read article
