Christiane Baumeister, Lutz Kilian, and Thomas K. Lee, "Inside the 
Crystal Ball: New Approaches to Predicting Gasoline Prices at the Pump", 
Journal of Applied Econometrics, Vol. 32, No. 2, 2017, pp. 275-295.

The data used in this article were collected from various sources.
Below we describe the sources and construction of the data used 
in our analysis. The raw data that are publicly available are in 
the files bkl_data_monthly.txt and bkl_data_quarterly.txt, which are 
ASCII files in DOS format. They are zipped in the file bkl_data.zip. 
Unix/Linux users should use "unzip -a".

The real-time database is not publicly available, so we only
include the last vintage of data (September 2014) with nowcasts
for the period Oct 1973 to Sept 2014. The monthly data are contained 
in the txt file in the same order as they are listed below.


DESCRIPTION OF DATA SOURCES AND CONSTRUCTION

1. GASOLINE

U.S. retail price of gasoline:
The monthly nominal price in $/gallon for motor gasoline, all grades
is reported in Table 9.4 of the Monthly Energy Review. This series 
starts in Jan 1978 and has been backcast to Oct 1973 based on the 
growth rate of the price for leaded regular gasoline.
http://www.eia.gov/totalenergy/data/monthly/#prices

U.S. gasoline consumption:
Monthly data on U.S. gasoline consumption are constructed as the sum of 
commercial, industrial and transportation sales based on data in Tables 
3.7a, b, c in the Monthly Energy Review.
http://www.eia.gov/totalenergy/data/monthly/#petroleum

Gasoline spot price:
The spot price for delivery of regular gasoline in New York Harbor is 
obtained from the U.S. Energy Information Administration (EIA). This 
series only starts in Jan 1986.
https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=eer_epmru_pf4_y35ny_dpg&f=m

Gasoline futures prices:
These data are from Bloomberg and are proprietary. They are available 
only to subscribers.


2. OIL

Refiner's acquistion cost of crude oil imports:
The monthly nominal refiner's acquistion cost for crude oil imports (RAC)
in $/barrel is reported in Table 9.1 of the Monthly Energy Review.
http://www.eia.gov/totalenergy/data/monthly/#prices
RAC_WTI indicates that WTI has been used for nowcasting.
RAC_Brent indicates that Brent has been used for nowcasting. 

Spot price for West Texas Intermediate (WTI) crude oil:
The monthly nominal WTI spot price is obtained from the FRED database of
the St. Louis Fed.
https://research.stlouisfed.org/fred2/search?st=WTI

Spot price for Brent crude oil:
The monthly nominal Brent spot price is obtained from the U.S. Energy
Information Administration. This series starts in June 1987. We backcast
the Brent series with the growth rate of the RAC (see above).
https://www.eia.gov/dnav/pet/hist/LeafHandler.ashx?n=pet&s=rbrte&f=m

Global crude oil production:
Monthly data for global crude oil production are obtained from the
Monthly Energy Review.
http://www.eia.gov/totalenergy/data/monthly/#international

Proxy for global crude oil inventories:
The monthly inventory data are constructed by multiplying U.S. crude oil 
inventories by the ratio of OECD petroleum inventories to U.S. petroleum 
inventories obtained from the Monthly Energy Review.
http://www.eia.gov/totalenergy/data/monthly/#petroleum
http://www.eia.gov/totalenergy/data/monthly/#international


3. ECONOMIC ACTIVITY

Global real economic activity measure (REA):
The global real activity index is constructed from data on global dry
cargo ocean shipping freight rates as described in Kilian (2009).
http://www-personal.umich.edu/~lkilian/reaupdate.txt

Chicago Fed National Activity Index (CFNAI):
The CFNAI is provided by the Chicago Fed.
https://www.chicagofed.org/research/data/cfnai/historical-data


4. INFLATION

U.S. consumer price index (US CPI):
Real-time data for the monthly seasonally adjusted U.S. consumer
price index for all urban consumers are obtained from Economic Indicators 
published by the Council of Economic Advisers. These data are made 
available by the FRASER data base of the St. Louis Fed. Additional 
real-time CPI data were obtained from the macroeconomic real-time data
base of the Federal Reserve Bank of Philadelphia.
https://fraser.stlouisfed.org/title/1
https://www.philadelphiafed.org/research-and-data/real-time-center 

Blue Chip: 
This is not a publicly available dataset.


5. OTHER FORECASTS

Monthly forecasts from Michigan Survey of Consumers:
The average expected change in the nominal price of gasoline in the 
Michigan Survey of Consumers was kindly provided by Soren Anderson
and is not reported here.

Quarterly forecasts from EIA:
EIA forecasts are from the Short-Term Energy Outlook (STEO) published
by the U.S. Energy Information Administration.
https://www.eia.gov/forecasts/steo/outlook.cfm
https://www.eia.gov/forecasts/steo/report/prices.cfm
The forecasts for retail gasoline prices in $/gallon for the period 
1991Q4 to 2014Q3 are contained in bkl_data_quarterly.txt in following 
format:

column 1: release date of the STEO
column 2: corresponding quarter
column 3: quarter of the forecast
column 4: last observation at Q(-1)
column 5: nowcast at Q0
columns 6-9: forcasts for 1-4 quarters ahead (Q1,Q2,Q3,Q4)


Please address any questions to:

Christiane Baumeister
Department of Economics
University of Notre Dame
722 Flanner Hall
Notre Dame, IN 46556
cjsbaumeister [AT] gmail.com