Update on U.S. gasoline prices

Two weeks ago, I commented on the tendency of U.S. retail gasoline prices to follow the price of Brent crude oil, anticipating on the basis of the price of Brent, then at $91.50, that we might expect to see average U.S. retail gasoline prices, then at $3.47, to fall an additional 35 cents/gallon. The gasoline price has since come down about 11 cents. But with Brent now surging back up near $100, this is about all we can expect.

Here’s an update of the Brent price and predicted gasoline price. With gasoline down and Brent up, the two are back to their long-run relation.

Black: average U.S. price of regular gasoline, all formulations, in dollars per gallon, weekly Jan 10, 2000 to Jul 2, 2012 (data source: EIA).
Blue: 0.84 plus 0.025 times price of Brent, in dollars per barrel, weekly Jan 7, 2000 to Jun 12, 2012 (data source: EIA) with Jun 29 value estimated from Oil-Price.net.

Ironman at Political Calculations has been kind enough to let us reproduce his tool for calculating the expected U.S. average retail gasoline price based on its long-run cointegrating relationship with the price of Brent crude oil. Just input the current (or your hypothesized) value for Brent in the box below, and click “calculate.” If you need a hint, I also include a self-updating graphic that displays the current price of Brent, courtesy of Oil-Price.net

Crude Oil Price Data
Input Data Values
Price per Barrel of Brent Crude Oil
Implied average U.S. gasoline price
Calculated Results Values
Average U.S. Price per Gallon
Code created with the assistance of Political Calculations

And here is also a self-updating plot of GasBuddy's estimate of the average U.S. retail price, though if you click its "show crude oil price" option, it uses WTI, which is less useful than Brent for predicting the price of gasoline.

New Jersey Historical Gas Price Charts Provided by GasBuddy.com


9 thoughts on “Update on U.S. gasoline prices

  1. AS

    For those of us who are using Excel to emulate your model, can you share a comparison of the regression statistics using time series software compared to the statistics shown on Excel? As I recall the regression statistics are not accurate due to the autocorrelation of the residuals? I think you introduced the Newey-West adjustment in a prior post. Thanks. Who needs video games with these fun models.

  2. Jeffrey J. Brown

    Average Brent price for first half of 2011 was about $113. Link to annual Brent prices:
    Following are the recent annual high Brent prices preceding year over year annual price declines, showing the progression in higher annual highs and higher annual lows.
    It seems likely that we we will show an year over year decline in 2012, to an average price that is, in all likelihood, well above the previous year over year decline of $62 that we saw in 20009.
    1997 to 1998: $19/$13
    2000 to 2001: $29/$24
    2008 to 2009: $97/$62
    2011 to 2012: $111/?
    The average rate of increase for new annual highs was about 13%/year (three data points), and the average rate of increase for new annual lows was about 16%/year (only two data points), suggesting that the average 2012 price might be somewhere in the vicinity of about $100.

  3. spencer

    Over the past week the CRB: raw material index has rebounded nicely, supporting your conclusion that crude oil may also be bottoming.

  4. westslope

    If I was betting on the outcome of the US presidential election, I would be watching the price of gasoline very closely. High oil prices giveth and high oil prices taketh away.

    Jeffrey: Boone Pickens is calling for $115/bbl oil by year end. I like your forecast of $100/bbl average price better. Brent Crude prices will soften once the Statoil strike is over. Pickens is also forecasting that the president will release oil from the SPR. Unlikely in my opinion.

  5. JDH

    AS: These are simply OLS coefficients and standard errors, same as in Excel. One can do more, but I was in a hurry and not particularly interested in standard errors.

  6. AS

    Dear Professor Hamilton,
    Sorry for the mundane question. I was curious after reading a bit on cointegrated models. Guess I need to buy a new Eviews program.

  7. Jeffrey J. Brown

    Regarding speculators, here is something that has always puzzled me about the “Blame the speculators” theory.
    Let’s stipulate, for the sake of argument, that the rascally speculators drove the annual average price of Brent crude from $55 in 2005 to $111 in 2011 (with a decline in 2009).
    Preceding this doubling in annual crude oil prices, there was another doubling, from $25 in 2002 to $55 in 2005.
    In response to the 2002 to 2005 doubling, the EIA shows that global crude oil production (Crude + Condensate) increased rapidly, from 67 mbpd in 2002 to 74 mbpd in 2005. At this rate of increase, we would have been at about 90 mbpd in 2011.
    However, the EIA shows that post-2005 annual global crude oil production has not exceeded 74 mbpd for six straight years, ranging between 72 and 74 mbpd.
    Here’s my question: Even if the rascally speculators drove the price of oil up from 2005 to 2011, why did we see no global crude oil supply response?

  8. Jeffrey J. Brown

    Here is my explanation for a doubling in global crude oil prices from 2005 to 2011:
    Following is a rough summary of my updated net exports paper, in three parts, focusing on the Export Capacity Index (ECI), which is the ratio of total petroleum liquids production to liquids consumption in oil exporting countries.
    Projections are almost always, to some degree, wrong, but I don’t think the general direction of the GNE and ANE curves are wrong, and I think that declining GNE & ANE are the key factors affecting the global economy, although most of the Cornucopian analysts don’t seem to have yet discovered “Net Export Math,” or they are carefully trying to avoid it.
    I think that this is the key question: Will improved technology allow oil importing and exporting countries to sufficiently increase production to offset (and exceed) generally increasing consumption in the exporting and developing countries?
    So far, the answer has been no, which contributed to the doubling in annual Brent prices from $55 in 2005 to $111 in 2011, which in turn contributed to the current economic slowdown.
    Part One: IUKE + VAM Case Histories
    I have generally used three members of AFPEC–the Association of Former Petroleum Exporting Countries–as models for net export declines. They are the IUKE countries–Indonesia, UK and Egypt.
    We had three new members of AFPEC last year–Vietnam, Argentina and Malaysia (VAM).
    I just had the production, consumption and net export data (BP) summed for all six countries, from 1986 to 2011, inclusive. I also calculated their combined ECI (Export Capacity Index, or the ratio of production of total petroleum liquids production to liquids consumption).
    Their final combined production peak was in 1998 at 7.0 mpbd, with consumption of 4.3 mbpd, and net exports of 2.7 mbpd, with an ECI of 1.64. In 2008, their production was 4.9 mbpd, consumption was 5.0 mbpd, and net imports were 0.1 mbpd, with an ECI of 0.98 (net importer status).
    A 29% decline in production, plus rising consumption, caused them to collectively go from net exports of 2.7 mbpd in 1998, or a billion barrels per year, to net importer status 10 years later, in 2008.
    Actual post-1998 combined Cumulative Net Exports (CNE) were about 4.4 Gb.
    The initial three year rate of decline in the ECI ratio was from 1.64 in 1998 to 1.45 in 2001, a decline rate of 4.1%/year. Extrapolated, this would put them at an ECI of 1.0 (zero net oil exports) in about 12 years, around 2010. They actually hit zero net exports in 2008.
    Estimated post-1998 CNE, using the initial three year rate of decline in the ECI ratio, were 5.0 Gb. Actual post-1998 CNE were 4.4 Gb.
    In other words, the actual net export decline was faster than what the initial three year projection predicted. This is a little more clear on the following graph which shows the 1986 to 2011 Export Capacity Index (ECI), which is the combined ratio of total petroleum liquids production to liquids consumption for the six countries:
    Note that the 1988 North Sea Piper Alpha accident contributed a significant decline in UK production, which caused the overall ECI ratio to temporarily decline.
    I think that these six diverse and geographically diversified former oil exporting countries give us a pretty decent model for Global Net Exports of oil (GNE).
    Part Two: Global Net Exports of Oil (GNE)
    Our data base calculates the combined net exports (in terms of total petroleum liquids) from the top 33 net oil exporters in 2005. We primarily used the BP data base, plus minor input from the EIA. Net exports are defined as total petroleum liquids production less liquids consumption in oil exporting countries.
    The following graph shows the combined ECI ratio for the top 33 net oil exporting countries in 2005:
    The 2008 to 2011 decline in the ECI ratio was faster than what the 2005 to 2008 rate of decline (2.2%/year) predicted we would see. Note that production from the top 33 net exporters in 2005 has been on an “Undulating Plateau” since 2005; however, three of the top 33 net oil exporters in 2005 (countries with net exports of 100,000 bpd or more) were Vietnam, Argentina and Malaysia, which were net importers in 2011.
    Based on the 2005 to 2011 rate of decline in the Top 33 ECI ratio, estimated post-2005 CNE (Cumulative Net Exports) for the (2005) Top 33 net oil exporters are about 445 Gb. Cumulative top 33 net exports from 2006 to 2011 inclusive were about 96 Gb, putting estimated post-2005 Top 33 CNE about 22% depleted.
    Part Three: Available Net Exports of oil (ANE)
    ANE are defined as GNE less Chindia’s combined net oil imports. The following graph shows the ratio of GNE to Chindia’s Net Imports (CNI):
    The 2008 to 2011 decline in the GNE/CNI ratio was faster than what the 2005 to 2008 rate of decline (7.7%/year) predicted we would see.
    Based on the 2005 to 2011 rate of decline in the GNE/CNI ratio, estimated post-2005 Available CNE (Cumulative Net Exports) are about 168 Gb. Cumulative ANE for 2006 to 2011 inclusive were about 81 Gb, putting estimated post-2005 Available CNE about 48% depleted.
    The 2005 to 2011 rate of decline in the GNE/CNI ratio, if extrapolated, suggests that China and India would be consuming 100% of Global Net Exports of oil in the year 2030, which is 18 years from now. I don’t think that will actually happen, but that is the trend line, and the rate of decline in the GNE/CNI ratio accelerated from 2008 to 2011, versus 2005 to 2008.

Comments are closed.