Selecting a model for forecasting
Chiu and al investigate the significance level of selecting variables for a forecasting model facing a non-stationary environment.
The results provide support for model selection at looser than conventional settings, albeit with many additional features explaining the forecast performance, with the caveat that retaining irrelevant variables that are subject to location shifts can worsen forecast performance.
Selecting a model for forecasting
Authors: Jennifer L. Castle, Jurgen A. Doornik and David F. Hendry
From: University of Oxford
Detecting forecast breakdowns
We develop a procedure which aims at capturing the policy cost of missing a break. We use data-based rules to find the test size that optimally trades off the costs associated with false positives with those that can result from a break going undetected for too long.
A New Approach for Detecting Shifts in Forecast Accuracy
Authors: Ching-Wai (Jeremy) Chiu, Simon Hayes, George Kapetanios, Konstantinos Theodoridis
From: Bank of England, Kings College London, Cardiff University
Judgement based forecasts
While statistical models typically deliver better forecast accuracy than economic models, communicating forecasts to decision makers often requires a narrative.[…] forecasts are more appealing to decision makers if they are underpinned by an economic rationale, which creates a tension if there is a rade-off between the theoretical and empirical coherence of forecasting models. This tension is magnified in small samples, as the potential gains in forecast accuracy through parameters bias come at the cost of reduced interpretability. Moreover, forecasts are an ingredient in counterfactual policy scenarios for which unbiased parameters are essential, for example when determining the growth payoffs from public investment.
Overfitting in Judgment-based Economic Forecasts: The Case of IMF Growth Projections
Author: Klaus-Peter Hellwig
From: IMF
