Anthony Garratt, Shaun P. Vahey, and Yunyi Zhang, "Real-time Forecast Combinations for the Oil Price", Journal of Applied Econometrics, Vol. 34, No. 3, 2019, pp. 456-462. The data used in the article are stored in a real-time dataset for the oil market published at https://www.niesr.ac.uk/real-time-forecastcombinations-oil-price. The file gvz-data.zip contains the variables used for the paper only. All files are CSV files in DOS format, so Unix/Linux users should use "unzip -a". There are 16 time-series variables, which we split into variables with and without revisions. The 11 revised variables are: (i) Brent (BRENT) (ii) West Texas Intermediate (WTI) and (iii) US refiners acquisition cost for crude oil imports (RAC), oil prices; (iv) world crude oil production; inventories for (v) U.S. crude oil (vi) U.S. petroleum and (vii) OECD petroleum; (viii) the ratio betwen OECD and U.S. petrolum inventories; (ix) the U.S. consumer price index (CPI) for all urban consumers, (x) the world real economic activity index (World rea index), and (xi) the 24-month cumulative growth rate of real shipping rates proposed by Hamilton (2018) (24-M shapping rate growth Hamilton(2018)). The remaining 5 variables (in Table 2), available without any revisions are: (i) NYMEX WTI Light Sweet Crude Oil Futures (WTI futures (with real-time backcast)), (ii) Intercontinental exchange (ICE) Brent crude oil futures (Brent futures (with real-time backcast)), (iii) gasoline prices, (iv) heating oil prices, and (v) an index of industrial raw materials (Industry raw prices (non-oil)). The more detailed explanations of the data set, including data resources, nowcasting and backcasting methods are summarised in a pdf file named GVZ-data-documentation.pdf (not for publication). Please address any questions to: Yunyi Zhang Dept. of Finance Warwick Business School University of Warwick CV4 7AL United Kingdom