CMStatistics 2016: Start Registration
View Submission - CFE
A0938
Title: Forecasting the real price of oil: Time-variation and forecast uncertainty Authors:  Christoph Funk - Justus-Liebig-University Giessen (Germany) [presenting]
Abstract: The price of oil is a driving factor of the world economy and gives a first hint on the potential economic development over time. Therefore, the purpose is to examine whether it is possible to generate an accurate forecast of the oil price and the associated uncertainty in this context. We contribute to the literature in two ways. First, the real-time out-of-sample performance of twelve individual forecasting models is investigated. This is a larger set of models than typically used in similar forecasting studies. In addition, the use of recursively constructed RMSE ratios discover potential weaknesses of the used models. Thus, several different combination approaches are tested with the result that a combination of individual models is beneficial for the forecasting performance. The second contribution is related to the forecast uncertainty of the investigated models. The forecast density provides an estimate of the probability distribution of future values and offers a complete description of the associated uncertainty. Thereby, the density is based on previous out-of-sample forecast errors and is obtained by bootstrap sampling. The model calibration will then be evaluated by checking whether the probability integral transform is uniformly distributed.