The fascination with knowing the future in advance is probably as old as human culture. Only recently has forecasting developed into a widely recognized scientific field, with applications in such diverse areas as economics, finance, or weather. Thus, in particular current economic forecasting lives in a colorful land between the ivory tower of mathematical-statistical rigor and the more mundane dwellings of mass media and politics.
The 1st Vienna Workshop on Economic Forecasting was hosted on February 15-16 by the Institute for Advanced Studies Vienna (IHS) and gathered around 80 participants from many countries, covering all corners of Europe from Finland to Ireland and Spain and further abroad. 40 oral research presentations in three parallel tracks, eight poster presentations, and four invited plenary talks gave insights into many aspects of forecasting. International interest in the workshop was strong, in academia as well as in policy institutions, and only every second submission could be accepted for presentation within the constraints of a two-day event.
Complementing the presentations, the second day of the workshop featured a panel chaired by Manfred Deistler (Vienna University of Technology), where representatives of Austrian forecasting institutions (IHS, OeNB and Wifo) reflected upon their forecasting experiences against the background of the insights gained at the workshop. Whilst the participants of the panel agreed that only short-term forecasts may be informative, there was also consensus that the provision of longer-term forecasts is useful as these provide plausible and consistent scenarios that may assist policy makers, thus setting a secure frame for the inevitably elusive future.
The workshop commenced with Gianni Amisano’s (Federal Reserve Board, Washington DC) invited talk on using Bayesian techniques to forecast with high-dimensional time-series models, an approach that relies upon computer-intensive technology and that experiences widespread and increasing usage. The second plenary talk by Lucrezia Reichlin (London Business School) provided a tour on the current state of the art in inflation forecasting, touching upon tools such as unobserved components, structural vector error-correction models, factor models, and again Bayes estimation techniques. The audience learned that the so-called Phillips curve, linking inflation and monetary policy with the real economy, is still alive and well and constitutes a useful tool for modelling and forecasting.
The third invited talk by David Turner (OECD, Paris) concentrated on the visualization of forecast uncertainty by so-called fan charts that permit emphasizing the increased risk of an economic downturn in volatile episodes. Tommaso Proietti (Universitá Tor Vergata, Rome) opened the second day with the fourth and last plenary talk. He discussed shrinkage estimation of autocovariance functions of multivariate stationary processes—important elements of many prediction models. The contributed presentations, too numerous to be reported in detail, covered various topics of current concern in economic forecasting, such as nowcasting, i.e. the timely processing of new information to get an efficient assessment of the current situation, and mixed-frequency models that are capable of blending fast information from financial markets with slow data from national accounts.
The workshop was a sound success, and the IHS will continue to provide space for discussion on the multifaceted field of economic forecasting.
"What I truly appreciated about the 1st Economic Forecasting Workshop was the amazing blend between academia, public service and the finance world. Academic excellence was refreshingly coupled with practitioner-focus. The line-up of central bankers from around the world was truly impressive. The wonderful framework program and the medium-size workshop granted invaluable opportunities to connect with like-minded colleagues from around the world to reflect on the current state-of-the-art of forecasting and generate cutting-edge ideas how to develop forecasting techniques that apply to multi-faceted scenarios. The strong IHS network support will hopefully lasts beyond the workshop and many interesting connections, mind-opening events and interesting idea exchange are likely to follow." (Julia M. Puaschunder)