nep-for New Economics Papers
on Forecasting
Issue of 2019‒04‒29
five papers chosen by
Rob J Hyndman
Monash University

  1. Going with your gut: the (in)accuracy of forecast revisions in a football score prediction game By Carl Singleton; J. James Reade; Alasdair Brown
  2. Assessing Macro-Forecaster Herding: Modelling versus Testing By Michael P. Clements
  3. IW Financial Expert Survey: Second Quarter 2019 By Demary, Markus
  4. TF-MIDAS: a new mixed-frequency model to forecast macroeconomic variables By Bonino-Gayoso, Nicolás; García-Hiernaux, Alfredo
  5. Forecasting in Big Data Environments: an Adaptable and Automated Shrinkage Estimation of Neural Networks (AAShNet) By Ali Habibnia; Esfandiar Maasoumi

  1. By: Carl Singleton (Department of Economics, University of Reading); J. James Reade (Department of Economics, University of Reading); Alasdair Brown (School of Economics, University of East Anglia)
    Abstract: We study individuals who each chose to predict the outcome of fixed events in an online competition, namely all football matches during the 2017/18 season of the English Premier League. We ask whether any forecast revisions the individuals chose to make (or not), before the matches began, improved their likelihood of predicting correct scorelines and results. Against what theory might expect, we show how revisions tended towards significantly worse forecasting performance, suggesting that individuals should have stuck with their initial judgements, or their 'gut instincts'. This result is robust to both differences in the average forecasting ability of individuals and the predictability of matches. We find evidence that this is because revisions to the forecast number of goals scored in football matches are generally excessive, especially when these forecasts were increased rather than decreased.
    Keywords: Judgement revision, Prediction making, Forecasting behaviour, Expectations
    JEL: C53 C23 D84
    Date: 2019–04–09
  2. By: Michael P. Clements (Henley Business School, University of Reading)
    Abstract: We draw on fixed-event and fixed-horizon survey expectations to better understand macro-economic forecasting behaviour. Fixed-event forecasts facilitate testing for herding behaviour,while fixed-horizon forecasts lend themselves to modelling the effects of consensus forecasts on individual forecasts. By pursuing these two approaches simultaneously for each individual forecaster, we can determine when the significance of the consensus forecasts in explaining an individuals forecasts is consistent with enhancing forecast accuracy, and when it reects strategic behaviour in response to other motives.
    Keywords: macro-forecasting, (anti-)herding, ?xed-event forecasts, ?xed-horizon forecasts.
    JEL: E37
    Date: 2018–07
  3. By: Demary, Markus
    Abstract: Pessimism determines the experts' predictions for the second and third quarter of 2019 which can be inferred from the downward revisions of the experts' forecasts. All in all, more downward revisions than upward revisions can be found in the forecasts indicating that the experts have interpreted the incoming information between end of December 2018 and end of March 2019 as bad news. Part or the forecast revision for the interest rates is due to their subdued inflation and growth outlook. All experts have revised their growth outlooks for Germany and the Euro Area downwards. So were Inflation forecasts for Germany revised downwards by 12 experts and inflation outlooks for the Euro Area were revised downwards by 8 experts. The other part of the interest rate forecast revisions were due to revisions about the future part of monetary policy interest rates, which also reflect a subdued inflation and growth outlook. While no experts expect the ECB to change its monetary policy, 12 experts have revised their forecasts for the federal funds rate downwards.
    JEL: G12 G17
    Date: 2019
  4. By: Bonino-Gayoso, Nicolás; García-Hiernaux, Alfredo
    Abstract: This paper tackles the mixed-frequency modeling problem from a new perspective. Instead of drawing upon the common distributed lag polynomial model, we use a transfer function representation to develop a new type of models, named TF-MIDAS. We derive the theoretical TF-MIDAS implied by the high-frequency VARMA family models and as a function of the aggregation scheme (flow and stock). This exact correspondence leads to potential gains in terms of nowcasting and forecasting performance against the current alternatives. A Monte Carlo simulation exercise confirms that TF-MIDAS beats UMIDAS models in terms of out-of-sample nowcasting performance for several data generating high-frequency processes.
    Keywords: Mixed-Frequency models, TF-MIDAS, U-MIDAS, Nowcasting, Forecasting
    JEL: C18 C51 C53
    Date: 2019–03–30
  5. By: Ali Habibnia (Virginia Tech); Esfandiar Maasoumi (Emory University)
    Abstract: This paper considers improved forecasting in possibly nonlinear dynamic settings, with high-dimension predictors ("big data" environments). To overcome the curse of dimensionality and manage data and model complexity, we examine shrinkage estimation of a back-propagation algorithm of a deep neural net with skip-layer connections. We expressly include both linear and nonlinear components. This is a high-dimensional learning approach including both sparsity L1 and smoothness L2 penalties, allowing high-dimensionality and nonlinearity to be accommodated in one step. This approach selects significant predictors as well as the topology of the neural network. We estimate optimal values of shrinkage hyperparameters by incorporating a gradient-based optimization technique resulting in robust predictions with improved reproducibility. The latter has been an issue in some approaches. This is statistically interpretable and unravels some network structure, commonly left to a black box. An additional advantage is that the nonlinear part tends to get pruned if the underlying process is linear. In an application to forecasting equity returns, the proposed approach captures nonlinear dynamics between equities to enhance forecast performance. It offers an appreciable improvement over current univariate and multivariate models by RMSE and actual portfolio performance.
    Date: 2019–04

This nep-for issue is ©2019 by Rob J Hyndman. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at For comments please write to the director of NEP, Marco Novarese at <>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.