European Parliament Library

Forecasting the Nominal Brent Oil Price with VARs—One Model Fits All?, Benjamin Beckers, Samya Beidas-Strom

Abstract
We carry out an ex post assessment of popular models used to forecast oil prices and propose a host of alternative VAR models based on traditional global macroeconomic and oil market aggregates. While the exact specification of VAR models for nominal oil price prediction is still open to debate, the bias and underprediction in futures and random walk forecasts are larger across all horizons in relation to a large set of VAR specifications. The VAR forecasts generally have the smallest average forecast errors and the highest accuracy, with most specifications outperforming futures and random walk forecasts for horizons up to two years. This calls for caution in reliance on futures or the random walk for forecasting, particularly for near term predictions. Despite the overall strength of VAR models, we highlight some performance instability, with small alterations in specifications, subsamples or lag lengths providing widely different forecasts at times. Combining futures, random walk and VAR models for forecasting have merit for medium term horizons
Table Of Contents
Cover -- Contents -- Abstract -- I. Motivation -- II. Relation to the literature -- III. Baseline forecasting models -- IV. Data -- V. Forecast evaluation -- Unbiasedness -- Accuracy -- VI. Robustness -- Structural breaks -- Sample splits -- Lag length -- VII. Forecast combinations -- VIII. The 2014 oil price collapse -- IX. Conclusion -- References -- Tables -- Table 1: Overview of Forecasting Models -- Table 2: Mincer-Zarnowitz Test for Unbiasedness -- Table 3. Root Mean Square Errors (relative to the Random Walk) -- Table 4. Root Mean Square Errors for Rolling and Recursive Estimations -- Table 5: Root Mean Square Errors (relative to the Random Walk): 1997M01-2008M06 -- Table 6: Root Mean Square Errors (relative to the Random Walk): 2008M07-2014M12 -- Table 7. Root Mean Square Errors of Forecast Combinations -- Figures -- Figure 1: Prediction-Realization Diagrams for VAR and Futures Forecasts -- Figure 2: Rolling Averages of Root Mean Square Errors -- Figure 3. Ex Ante Nominal Brent Price Forecast, as of May 2014 -- Figure 4. Ex Ante Nominal Brent Price Forecasts, as of December 2014
Language
eng
Literary Form
non fiction
Physical Description
1 online resource (33 pages)
Specific Material Designation
remote
Form Of Item
online
Isbn
9781513526683

Library Locations

  • EP Library Strasbourg

    7 Place Adrien Zeller, Allée du Printemps, Strasbourg, F-67070, FR
    Borrow
  • EP Library Luxembourg

    Rue du Fort Thüngen, Luxembourg, L-1313, LU
    Borrow
  • EP Library Brussels

    60 rue Wiertz, Brussels, B-1047, BE
    Borrow