(Crossposted from Hyndsight)

Today at the International Symposium on Forecasting, I announced the awards for the best paper published in the International Journal of Forecasting in the period 2012-2013.

We make an award every two years to the best paper(s) published in the journal. There is always about 18 months delay after the publication period to allow time for reflection, citations, etc. The selected papers are selected by vote of the editorial board. The best paper wins an engraved bronze plaque and US$1000. Any other awards are in the form of certificates.

For 2012-2013, 11 papers nominated of which five were short-listed for the award. The five short-listed papers were:

This year, we have made two awards. The best paper award goes to Francis Diebold and Kamil Yilmaz for their 2012 paper on estimating spillovers between markets. The nomination of the paper included the following citation.

This is a methodological paper developing ways to estimate spillovers from one market to others. Diebold and Yilmaz use a generalized vector autoregressive framework in which forecast error variance decompositions are invariant to the variable ordering. Even though they used the method to look at volatility spillovers internationally in the time domain, the procedure is usable more generally including with cross-sectional data having spatial interconnections. This is an important contribution with significant repercussions across financial markets.

An “outstanding paper award” goes to Véronique Genre, Geoff Kenny, Aidan Meyler and Allan Timmermann for their 2013 paper on combining expert forecasts. The following citation explains why the paper was nominated.

This is a topic that has been of fundamental interest to forecasters in all fields: can one beat the simple average when combining forecasts. The authors do an excellent job exploring that issue in a data set that had not previously been fully exploited.

The authors explore an extensive set of methods to show that, on aggregating forecasts, a simple average is a benchmark that is very difficult to beat by more sophisticated aggregation schemes. Although the finding per~se is not new (we have numerous studies examining the “forecast combination puzzle”), the rigorous approach to comparison of methods makes this manuscript very relevant.

Share Post!

Subscribe to Blog