The Energy Information Administration (EIA) of the U.S. Energy Department (DOE) developed the Short-Term Integrated Forecasting (STIFS) model to generate short-term (up to 8 quarters), monthly forecasts of U.S. supplies, demands, imports, stocks, and prices of various forms of energy. The purpose of this report is to define the energy price model in STIFS and describe its basic properties. This report documents the August 2000 version of the energy price equations in STIFS.
This report is written for persons who want to know how short-term energy markets forecasts are produced by EIA. The report is intended as a reference document for model analysts, users, and the public.
This section discusses the methodology for forecasting the various energy prices published in the Outlook. The prices are important in their own right, because they are widely used for budget planning and other purposes by Federal and local government agencies, as well as corporate planners. These prices are also used in the projections of energy supply and demand discussed in other sections of the Model Documentation.
The price of imported crude oil (RAIMUUS) is an exogenous variable which is determined by analyzing the world petroleum demand and supply balance over forecast period. Factors that influence the price include new oil fields coming into production (as well as rates of decline in existing fields), regional economic growth rates and planned shifts in production from major producing nations. The composite refiner acquisition cost of crude oil (RACPUUS), a weighted average of imported and domestic crude oil, is assumed to equal to RAIMUUS plus the historical differences between RACPUUS and RAIMUUS over the last year (June 1999-May 2000). For this documentation the difference was $0.25 per barrel. Thus:
RACPUUS = RAIMUUS + 0.25.
The refiner price of gasoline (MGWHUUS) is estimated as a function of the refiner acquisition cost of crude oil, deseasonalized day's supply of motor gasoline (MGPSPUSA(t-1)/MGTCPUSA), a dummy variable for the period of May 1994 -June 1994, when crude prices rose rapidly, yet refiner margins fell (DUM94MJ), a dummy variable for 1995 on (D95ON), considered to be a period of gains in refinery efficiency, and monthly dummy variables.
|
[Click here for Regression Results] |
The retail price of motor gasoline, ex-tax (BLS series: MGTCAUS) is a function of the refiner price (MGWHUUS) and MGWHUUS(t-1), the consumer price index as a proxy for inflation (CICPIUS), refining cost passthroughs for oxygenated gasoline, (OXFRAC), cost passthroughs for reformulated gasoline (RFRAC) cost passthroughs for regions and periods where both oxygenated and reformulated gasolines are used (OPFRAC), and monthly dummy variables:
|
[Click here for Regression Results] |
Motor Gasoline Taxes (MGTXUUS) are then added to MGTCAUS to get the pump price (MGUCUUS). Taxes on motor gasoline consist of the Federal tax of 18.3 cents per gallon and the state tax.. The most recent estimate for the average state tax is 20.07 cents per gallon (Energy Information Administration, Petroleum Marketing Monthly, May 2000, Table EN1.) Although the average tax on motor gasoline increased during the 1970's and 1980's, it has remained fairly flat over the past 3-4 years. In the forecast this tax is also projected to remain flat, averaging about 38.4 cents per gallon. Thus:
MGUCUUS = MGTCAUS + MGTXUUS
The retail price of motor gasoline, ex-tax (EIA series: MGTC1US) is a function of the refiner price (MGWHUUS) and MGWHUUS(t-1), the consumer price index as a proxy for inflation (CICPIUS), refining cost passthroughs for oxygenated gasoline, (OXFRAC), cost passthroughs for reformulated gasoline (RFRAC) cost passthroughs for regions and periods where both oxygenated and reformulated gasolines are used (OPFRAC), and monthly dummy variables:
|
Motor gasoline taxes (MGTXUUS) are then added to MGTC1US to get the pump price (MGEIRUS). Thus:
MGEIRUS = MGEI1US + MGTXUUS
The retail price of motor gasoline, ex-tax (EIA series: MGTC2US) is a function of the refiner price (MGWHUUS) and MGWHUUS(t-1), the consumer price index as a proxy for inflation (CICPIUS), refining cost passthroughs for oxygenated gasoline, (OXFRAC), cost passthroughs for reformulated gasoline (RFRAC) cost passthroughs for regions and periods where both oxygenated and reformulated gasolines are used (OPFRAC), and monthly dummy variables:
|
[Click here for Regression Results] |
The retail price of diesel fuel oil (DSTCUUS) is a function of the refiner acquisition price for crude oil (RACPUUS), days' supply of distillate fuel (DFPSPUS(t-1)/DFTCPUS, the consumer price index as a proxy for inflation (CICPIUS) , a dummy variable for the extreme weather in January 1990 (D9001), and monthly dummy variables.
|
[Click here for Regression Results] |
Diesel Motor Fuel Taxes (DSTXUUS) are then added to DSTCUUS to get the pump price (DSRTUUS). Taxes on motor gasoline consist of the Federal tax of 24.3 cents per gallon and the state tax.. The most recent estimate for the average state tax is 19.64 cents per gallon (Energy Information Administration, Petroleum Marketing Monthly, May 1998, Table EN1.) Although the average tax on diesel fuel increased during the 1970's and 1980's, it has remained fairly flat over the past 3-4 years. In the forecast this tax is also projected to remain flat, averaging about 44.5 cents per gallon. Thus:
DSRTUUS = DSTCUUS + DSTXUUS
The refiner price of distillate fuel oil (D2WHUUS), is a function of the refiner acquisition price for crude oil (RACPUUS), demand (squared) for distillate fuel (DFTCPUSA(**2)), a one-month- lag of demand (squared) for distillate fuel ((DFTCPUSA(**2))(t-1)), deviation from normal (3-year average) stock levels (DFPSPUSA(t-1)- (DFPSPUSA(t-13) + DFPSPUSA (t-25) + DFPSPUSA(t-37)), and monthly dummy variables.
|
[Click here for Regression Results] |
The residential heating price (D2RCUUS) is estimated as a function of the dependent variable lagged one month, the refiner price of distillate fuel oil (D2WHUUS), the producer price for non energy and food products as a proxy for inflation (WPIINUS), and monthly dummy variables:
|
[Click here for Regression Results] |
Retail residual fuel oil prices are estimated as a function of the dependent variable lagged one month, the refiner acquisition cost of crude oil (RACPUUS) and monthly dummy variables.
|
[Click here for Regression Results] |
RFEUDUS is the price of residual fuel oil to electric utilities, expressed in dollars per million Btu. The conversion factor for residual fuel oil is 6.287 million Btu per barrel. In addition there is a $0.05 per million Btu difference between the electric utility price and the retail price. Thus:
RFEUDUS = (RFTCUUS / 6.287) + 0.05
The price of kerosene jet fuel (JKTCUUS) is a function of the dependent variable lagged one month, the refiner acquisition cost of crude oil (RACPUUS), days' supply of jet fuel ((JFPSPUS(t-1)/ JFTCPUS)), the producer price for non energy and food products as a proxy for inflation (WPIINUS), a dummy variable for extremely cold weather during December 1989 and January 1990 (DUMCOLD), and monthly dummy variables.
|
[Click here for Regression Results] |
The price of consumer grade propane (PRTCUUS) is estimated as a function of the dependent variable lagged one month, the price of heating oil (D2RCUUS) which captures the weather effects, and monthly dummy variables.
|
[Click here for Regression Results] |
The producer price index for petroleum products (W57IUSA) is estimated in deseasonalized form, as a function of the refiner prices of motor gasoline (MGWHUUSA), distillate fuel oil (D2WHUUSA), and kerosene jet fuel (JKTCUUSA).
|
[Click here for Regression Results] |
The price of coal to electric utilities (CLEUDUS) is estimated as a function of the dependent variable lagged one month, the price of diesel fuel oil (DSTCUUS) which captures the transportation costs, mining productivity (CLMRMUS), and monthly dummy variables.
|
[Click here for Regression Results] |
The spot wellhead price for natural gas (NGSPUUS) is estimated as a function of deviations from normal of storage {(GASVAR) where GASVAR= NGUSPUS(t-1) -NGUSPNM(t-1)}; the dependent variable lagged one month; dummy variables for anomalous circumstances: Hurricane Andrew-which shut down some production in September and October 1992 (D_HAND), extremely hot weather in the Southwest in August and September 1993 (D_9308N9), very cold weather and ice storms which caused some transmission failures in February 1994 (D_294), and a dummy variable for a spot price panic resulting from low inventories and expectations of cold weather in February 1996 (DUM9602), and monthly dummy variables.
|
[Click here for Regression Results] |
The composite wellhead price for natural gas (NGWPUUS) is estimated as a function of the dependent variable lagged one month; the spot price of natural gas (NGSPUUS) and a dummy variable for a spot price panic resulting from low inventories and expectations of cold weather in February 1996 (DUM9602), and monthly dummy variables.
|
[Click here for Regression Results] |
The price of natural gas to electric utilities (NGEUDUS) is estimated as a function of the wellhead price (NGWPUUS), the price of residual fuel oil to electric utilities (RFEUDUS), and monthly dummy variables.
|
The price of natural gas to residential users (NGRCUUS) is estimated as a function of the dependent variable lagged one month; the wellhead price lagged on month (NGWPUUS(t-1)), the producer price for non energy and food products as a proxy for inflation (WPIINUS), and monthly dummy variables.
|
[Click here for Regression Results] |
The price of natural gas to electric utilities (NGCCUUS) is estimated as a function of the dependent variable lagged one month, the wellhead price (NGWPUUS), the producer price for non-energy and food products as a proxy for inflation (WPIINUS) and monthly dummy variables.
|
[Click here for Regression Results] |
The price of natural gas to electric utilities (NGICUUS) is estimated as a function of the dependent variable lagged one month, the wellhead price (NGWPUUS), the price of residual fuel oil to electric utilities (RFEUDUS), and monthly dummy variables.
|
[Click here for Regression Results] |
The price of electricity to residential users (ESRCUUS) is estimated as a function of the dependent variable lagged one month; the volume weighted cost of fossil fuel prices to electric utilities (AFUEUUS) lagged 2 months, the producer price for non energy and food products as a proxy for inflation (WPIINUS), and monthly dummy variables.
|
[Click here for Regression Results] |
File last modified: July 21, 2000
This page's URL: http://www.eia.doe/emeu/steo/pub/document/textpr.html