The great gasoline price conspiracy

Have you heard the latest explanation for how President Bush contrived to lower gasoline prices in time for the election?

An Econbrowser reader calling himself “Alan Greenspend”, calls our attention to a post by Russ Winter, which, in Alan’s mind, suggests how George Bush might be responsible for falling gasoline prices. I traced the source of this insight back to this story ($$$) in the September 21 Wall Street Journal:

After long shying away from oil, natural gas, metals and other raw materials, investors of all stripes — hedge funds, pension plans, endowments and individual investors — have become enamored with commodity investing. These investors, including short-term speculators, have become key in various markets, sometimes driving prices more than industrial customers who buy the materials to make things or sell services….

For evidence of these investors’ influence, consider the Goldman Sachs commodity index, one of the most popular vehicles for betting on raw materials. In July, Goldman Sachs tweaked the index’s content by cutting its exposure to gasoline. Investors tracking the index had to adjust their portfolios accordingly — which sent gasoline futures prices tumbling.

Undernews spells out what this means for those slow on the uptake:

Here we have Goldman, qua keeper of the commodities index, manipulating markets simply by adjusting index components…. It lends credence to the theory that the current well-publicized commodities decline is just a well-timed, well-orchestrated head fake to benefit the incumbents in the run up to the midterm elections – someone noted recently that Bush’s ratings vary inversely with gas prices.

And Marc Parent reminds us that

the American Treasury Secretary moved to his current job directly from being Chairman and CEO of the Goldman Sachs Group. He took a $38 million pay cut to change jobs.

As conspiracy theories go, this is really quite good. Certainly for something like the S&P500 index, which a huge number of funds simply hold according to mechanical rules, one can see a price change by as much as 3% when a company is listed or delisted, owing to a temporary flood of buys or sells. These liquidity effects can last several weeks. I doubt that mechanical indexed holding of the Sachs index operates on anything like the scale that the S&P500 is used, but I could well imagine there could be a measurable effect of something like this.

November 2006 unleaded gasoline contract. Source:


However, there are just a few holes in this theory through which some might try to pick a few nits:

  1. Between July 12 (when Winter says Goldman Sachs announced the change) and August 7, gasoline futures prices did not fall, but instead rose by 3%.
  2. When gasoline futures prices finally did begin to fall, the magnitude of the decline (25%) is far greater than could be attributed to temporary liquidity imbalances.
  3. The most important factor in the drop of gasoline prices has been the drop in the price of crude oil, in which gasoline futures would play no role.
  4. Price effects that result from a flood of sell orders on a market that cannot absorb them are by their nature inherently short-lived, ultimately reversed by the fundamentals of supply and demand. An effect operating through this mechanism in July could not conceivably determine the price of gasoline in November.

But it’s such a nice conspiracy theory, I hate to leave it twisting in the wind like this, so let me offer it some assistance. To salvage the theory, you’d have to take the view that both oil and gasoline prices have nothing to do with fundamentals, but instead are entirely determined by the market psychology of speculators. Perhaps then the Goldman Sachs move is alleged to have triggered a massive psychological response in global gasoline and oil markets (and while we’re at it, why not gold, copper and zinc as well?). I’ve earlier examined some of the problems with this view. It also has the feature that anything you observe in the world could be consistent with it. Prices go up? Must have been psychology. Prices go down? Obviously even more confirmation that psychology is responsible!

A theory that has no testable implications is to me not the most promising way to try to make sense of the world. But for some people, that lack of empirical restraint is precisely what makes such a paradigm appealing.

Technorati Tags: ,

42 thoughts on “The great gasoline price conspiracy

  1. calmo

    The latest conspiracy theory reminds me of that line about the patient who retorted that just because he was paranoid didn’t mean people weren’t following him.
    Yeah and now even more –for the entertainment value.
    Ok, somewhat like this.
    My (not so paranoid) impression is that not only is there a high correlation between w’s approval ratings and pump prices, but that w&co are aware of this and possibly other correlations (my more and growing paranoid impressions).
    Does this phenomena go back to other presidential terms? I’m not paranoid enough to check, lily liver…no, I’m too paranoid to check, Rambo.
    The worries about the Fed raising interest rates just before an election are seen (by the Rambo incumbents) to be a conspiracy against the encumbent (eg Bush I) and who are we (lily livers) to examine the falling pump prices as we approach the mid-terms, Rambo?

  2. PaulS

    Calmo, we keep getting stuff like this graph cited on The Oil Drum the other day, that leads to impressions like yours, and reinforces a general impression of godly powers of the President. But all I see in the graph is that something is steadily going down while something else is steadily going up. The graphs don’t track each other in detail because there are no details, just some noise, and not very much at that. So what? There are lots of things that have gone fairly steadily up or down over the last five years.
    JDH, is there any legitimate statistical test that would assign any significance to a graph like this? If so, when and why would anyone consider such a test to be of any genuine use?

  3. JDH

    Paul, you’re absolutely correct that a correlation or regression coefficient relating two trending series like these is quite meaningless. However, one can fix most of the problems by including lags of both the left-hand and right-hand variables in the regression, or by looking at the correlation between the growth rates rather than the levels of the series. There’s been enough casual yakking about this correlation, somebody has probably attempted this properly, but I don’t remember seeing it.

  4. wcw

    The main problem with the Goldman-did-it argument that I see is that, pace Paulson, Goldman is the only investment bank on the street that is culturally Democratic. Rubin was the rule, not the exception. I can see Goldman gaming their index to make a couple bucks, but politics was almost certainly not a consideration.
    However, your (3) is slightly wrong. Oil at $62 is down $15 from its peak, while crack spreads are down over $15. The more important of these factors is the latter, followed closely by the former. The “most” important factors, of course, are those which influenced the price of crude and the margins earned by refiners.

  5. HZ

    Does the spot price drive the future price? Or the other way around? Is oil&gas supply/demand too inelastic that psychology plays an unusually big role in amplifying price movements?

  6. JohnnyQ

    Oil and gas prices are done going down, now time to go back up. The fact that they have not gone down anymore in the last 10 days is VERY telling. 90 barrels by Christmas.

  7. sharkbait

    Its hard to see how Goldman did it. But that doesnt mean that the administration cannot influence oil prices. They have friends in Saudi Arabia — our biggest ally. One phone call to the prince is probably sufficient to lower price of oil – temporarily. The Saudis want the republicans in power. Why not help out a friend in need. If it goes back up after the election … you’ll know.

  8. Aaron Krowne

    Looking at the data from past seasons, the end of the summer demand plateu for gas tends to be on the order of 10% in terms of retail price. We had 17% this time. So the question is not whether the GSCI reweighting was responsible for all of this drop, but how much it accounted for. It cannot possibly be zero, by virtue of the existence of GSG alone.

    The Amaranth pop, hurricane fizzle-out, and anti-effect of the MTBE phase-out all likely contributed to the price drop. That doesn’t seem to leave too much left. I’d say 2-5% at max.

    As for the timing issue for the “conspiracy theory”, I don’t think the announced time of the change is necessarily the best indicator; probably the substantial price effects wouldn’t occur until the August-September contract roll began. This would coincide better with the observed price drop… almost perfectly.

    What I want to know is what the market capitalization of GSCI-trackers is; this would allow one to get some estimates for hard numbers on how much selling of gas futures had to take place. Anyone know how to get this data? It seems like it’d be tricky, because there are bound to be many informal trackers, not just GSG (there’s a Chicago Merc contract, though).

    By the way, does anyone know what the excuse was for re-weighting down by more than 70% one component of a supposedly “market-weighted” index? Did the market suddenly change by that much?

  9. Aaron Krowne

    By the way, the WSJ story is explicitly wrong, the index was not changed in July. It was changed for September. If there is a conspiracy, it wouldn’t be surprising if the WSJ was part of the disinformation campaign, which has been successful in casting confusion upon the analysis in this blog post =)

  10. esb

    GS saw enough people leaning the wrong way and decided to knock out their teeth. From time to time this happens and it is just a part of the game.
    The reason so many were leaning that way (long and unhedged) had more to do with a huge bet on the July-September weather in the Gulf than anything else.
    Now regarding politics and our good friends the Saudis, just remember that you don’t refer to someone as “Bandar Bush” without a reason, esp. if your name is Bush 41. If anyone made the calls it would have been 41. I would just love to listen to some of those tapes.
    And as for a solid conspiratorial “ruling elite” the onrushing Iran insanity has pretty much split it, and not right down the middle. No one with any rationality can see a good outcome from another war overlapping the Iraq debacle. But the word seems to be that 43 is not to be denied as the Cheney “force of will” continues to be irresistable to the little twerp.
    Which is worse…to lose the 15% rate on gains and dividends (with a Democratic house and/or Senate) or to suffer the results of an unprovoked and possibly nuclear attack on Iran (guaranteed in ’07 or ’08 with continued one-party rule)…that is the question on the minds of the cognac and cigar set.

  11. Joseph

    why do the Saudis want the Republicans in power?
    There is a long relationship between the Saudi royal family and the Bush family going back many years to their roots in the oil business. The Saudis were investors in Bush’s Arbusto and Harkin oil businesses. The Saudis are also investors in the Carlyle Group for whom Bush senior was a consultant and where James Baker, the family fixer, is a partner.
    But more directly, the Democrats made an anti-terrorist crackdown on Saudi Arabia a part of their platform in the 2004 election. They pointed out the coddling Saudia Arabia seemed to get compared to all other Gulf states even though Bin Laben came from Saudi Arabia, most of the 9/11 hijackers came from there, the 9/11 financing came from there and the bulk of radical Islamist financing even today still comes from there. Democratic politicians including Kerry, Dean, and Clark made very clear that they would have a much more demanding relationship than the cozy one between the Bush/Cheney oil men and Saudi Arabia. There is no doubt that Saudi interests are served by having Bush and the Republicans dominating American politics.

  12. My 1st Million At 33

    Carnival of the Capitalists – October 2nd, 2006

    Welcome to Carnival of the Capitalists – 10/2/06 Edition at Thanks to all those who participated, and thanks to Jay to give me this opportunity to host.
    If you’re on my site for the first time, I welcome you to take a look at my S…

  13. jfund

    There may have been a conspiracy but I doubt it. GS had little choice but to downweight unleaded. The unleaded contract was dying due to changes in gasoline regulations and the replacement RBOB contract had not caught on. You can’t roll large positions from an old liquid contract into a new illiquid contract without huge risk.

  14. JDH

    Joseph, I’m still not understanding exactly what you’re claiming. Which specific U.S. policies do you believe will change if the Democrats take control of the House of Representatives in the coming election? How much revenue do you believe the Saudis are willing to sacrifice in order to prevent this outcome?

  15. T.R. Elliott

    JDH: I’m not going to jump into support of the Saudi’s manipulating the oil markets to support Republicans. I have no idea what they want.
    Your counterargument based on economic reasons is certainly something to consider, but misses the big picture, the big picture that I sometimes think economists ought to be more honest about.
    The big picture is that money does not make the world go around. People very frequently do not act rationally. And the idea that actors operate purely according to economic stimuli just doesn’t seem true in my personal experiences nor in my reading of current and historical events.
    In particular, if you want to look at a good example of irrational uneconomic behavior, just check out the mess that is called the White House, Senate, and Congress today.
    In the present circumstances, we have no idea who is talking to who, what conversations are taking place, etc. I hear that Saudi Arabia is now building a wall along the border to Iraq to contain the violence. Could they Saudis possibly think that Republicans, though they’ve created a mess in Iraq, might at least keep troops in their longer. Might the Saudis be delusioned enough to think that stay the course a bit longer might work? They could be. At least 30% of the American electorate is similarly deluded.

  16. Adventure of Strategy

    “There’s Oil in Them Thar Mountains”

    Click on image to enlarge Heres a factoid from the Oxford Analytica conference that caught me a little by surprise: The USA is sitting on oil reserves more than 3 x the total of Saudi Arabia, in the form…

  17. Les

    The change in the Goldman Sachs Commodity Index wasn’t effective until August 9. It appears there was an initial dip of approximately 20 cents after the July 12 press release announcing the change. This was followed by a short-covering rally into the final announcement by Goldman on the 9th. The real damage to the price of gasoline, as you noted, occurred after this date.

  18. Joseph

    How much revenue do you believe the Saudis are willing to sacrifice in order to prevent this outcome?
    I have no opinion on that subject and never said anything about it. I only answered the simple question you asked about why the Saudis prefer Republicans. The Democrats have a stated policy of being less favorable to the Saudis. Things like publicly denouncing the Saudis for their tardy cooperation in efforts against radical islamism would be embarrassing. Congressional investigations into the Saudi connections to terrorist groups. Restrictions on Saudi investments like the Dubai port deal. Restrictions on military aid and arms deals. For example, the U.S. has supplied their entire national defense infrastructure including command, control and communications systems, missile defense systems, and AWACS radar aircraft. A Democratic congress would weaken Bush and his ability to push his foreign policies.
    To deny the Saudi’s motivation to favor a Republican administration in the face of Democrats public declaration to get tough on Saudi Arabia is just absurd.
    As to whether they would actually try to manipulate prices to affect an election I will leave to others and have no opinion. Their willingness and ability to manipulate prices, we know to be true. In the 80’s they started cutting prices and increasing production when they realized that conservation efforts were beginning to reduce demand. It’s a simple cost/benefit analysis.
    If there are efforts to influence the election, I suspect those to be occurring on the domestic wholesale gasoline level, not the international level. It is not so hard to believe that wholesalers would be just a little less aggressive in pricing for a month or so. This might amount to only a nickel a gallon or so, but every bit helps the cause. As to how much they are willing to spend, we know that the oil industry has collectively spent hundreds of millions of dollars in direct contributions to Republicans presumably after making the requisite cost/benefit calculations. It is not beyond the realm of possibility that they would be just as willing to make indirect contributions. I don’t buy into the idea that most of the price decline is due to manipulation, but you can’t deny that the oil companies are extremely motivated to get Republicans re-elected.

  19. Hal

    Two other problems with the conspiracy theory. First, this is not the only widely used commodity index out there. I don’t know how many funds follow it versus the others, but AFAIK this was the only one to make such a change at this time.
    Second, even if some traders closed their gasoline positions in response to this change, that would not necessarily drive the price down. The futures market is zero sum, with as many traders short as long. When some segment of the market closes their positions the price could either go up or down.
    Now, large investors do get their positions tracked, and last I saw they were on net more long than short, but it was only about 60-40. So this could create some downward pressure when they sell, but it would be a relatively moderate effect and unlikely to produce such a dramatic move.

  20. Russ Winter

    In the interest of fair play, an actual reading my blog on GS gaming makes absolutely no reference or even an inference to a political motivation or to the elections. I simply refer to an event (suddenly revamping an important index) that has a rancid odor to it. In fact I blog on similiar rancidiness all the time. Basically, you had a situation where speculators were leaning too heavy long, and GS used the situation to game them, blow them out. Money is the motivation here. Happens all the time in these markets. Finally, speculative money plays a huge role in today’s markets, and in fact has come to dominate them, that’s my key point. I refer to it as synthetic economics, or the “Land of Oz”, and it’s a dangerous condition.

  21. JDH

    Fair enough, Russ. I’ve reworded the post to try to draw a distance between what you concluded in your post and what others are inferring from these same facts. Did want to give you a link here, though, since your post is where I started from.

  22. Russ Winter

    Of course there very well could have also been a political motive or favor involved? I don’t know? But my feelings on the matter lean more towards tawdry market gaming and trading profits. These Boyz are really addicted to that kind of stuff, and have few scrumples, especially in a Bubble economy. And never underestimate the impact of hedge fund or speculative capital either now days. An otherwise quiet appearing market can suddenly explode because of it.

  23. Thomas James

    The world is sooo complicated, ey? This just in from the WSJ:
    WASHINGTON — The Energy Department will hold off purchases of oil for the government’s emergency reserve through the upcoming winter, the department said Monday.
    The department said it will not try to recover 1.7 million barrels of crude oil that it loaned refineries a year ago after Hurricane Katrina, nor would it seek to purchase additional oil for the reserve until next spring so as not to take oil off the market during the heating season.

  24. Alan Greenspend

    It appears that the NY Times also picked up on this cause and effect from Secretary Paulson’s old company, notably after Russ did.

    “Change in Goldman Index Played Role in Gasoline Price Drop”

    Coincidence is not causality, of course, no matter how happy the coincidence. I’m sure this will be just another forgotten footnote without an investigation.

  25. Rich Berger

    I know you mean well, but remember Swift: It is useless to attempt to reason a man out of a thing he was never reasoned into.

  26. Thomas James

    The following quote is evidently taken from Bob Woodward’s new book:
    “During a meeting in the Oval Office, according to Woodward, Bush personally thanked Bandar because the Saudis had flooded the world oil market and kept prices down in the run-up to the 2004 general election.”
    Looking to the numbers, Saudi crude production in the five months Apr-Oct was an average 750,000 bpd greater in 2004 than 2003.
    But then again, the world is just soooo complicated, ey?

  27. Jon

    Anyone else remember that is was Goldman who was the standard bearer for last year’s “Crude is going to $100” hype?

  28. Rich Berger

    The price of oil continued to rise in 2004, right through the election. If the Saudis were trying to help Bush, their help was ineffective.

  29. Tim


    On your four points above:

    1. The signficant event was August 9th and the October roll of unleaded gas futures (see Goldman Sachs Press Release of August 9th).
    2. The NY Times says there were $6 billion in unleaded gas futures affected (see
    NYT: Change in Goldman Index Played Role in Gasoline Price Drop). Don’t know how that was calculated, but $6 billion sounds like a lot of gas.
    3. Movement down in the price of unleaded gas preceded the move down in crude oil since Aug 9th (see TMTGM: Friends in High Places chart at bottom).
    4. Fundamentals of supply and demand for oil? What are they? Oil prices are all about risk and spculation.

    – Tim

  30. Tim

    [Second try – the first one looked fine in the preview …]
    On your four points above:
    1. The signficant event was August 9th and the October roll of unleaded gas futures (see Goldman Sachs Press Release of August 9th).
    2. The NY Times says there were $6 billion in unleaded gas futures affected (see
    NYT: Change in Goldman Index Played Role in Gasoline Price Drop). Don’t know how that was calculated, but $6 billion sounds like a lot of gas.
    3. Movement down in the price of unleaded gas preceded the move down in crude oil since Aug 9th (see TMTGM: Friends in High Places? chart at bottom).
    4. Fundamentals of supply and demand for oil? What are they? Oil prices are all about risk and spculation.
    – Tim

  31. Jon

    Tim…nice links
    Buried in the gas article, #2, is this:
    ….” in August, the bank said it would not roll over any more positions into gasoline and would redistribute the weighting into other petroleum products….”….
    Which would make sense from an indexing standpoint given concerns over OI and tradability of the reformulated contract….

  32. Russ Winter

    Gasoline demand (four week average, see last chart) is running 5.21% ahead of last year. That’s 458.000 barrels per day of extra useage.
    The US flush with oil and distillate inventory, but the rest of the world is extremely low, see days of worldwide demand.….
    Basically the US has sucked all the product in, and hoarded. Maybe to prepare for an Iran action, last man standing?
    Friday I’ll bet we see specs short oil. They already ahave been short gasoline and HO.

  33. JDH

    I see that Gumby Fresh has dug up some further interesting evidence on this hypothesis. Gumby asks what happened to the price of the component whose weighting by Goldman Sachs increased. The result? Down 27%.

  34. Gari N. Corp

    This Gumby Fresh guy, he is I. And I’m not sure I’ve read the right thing into the futures prices, which, by the way, I just clicked through from the Slate article that references this discussion, and I haven’t followed carefully enough what sort of lag period we should be looking at. But I do think it’s worth taking a look at markets that GS could influence with the move. Ethanol likely is one of them.

  35. Graham Zabel

    No one’s mentioned the very convenient timing of the Jack 2 oil find in the Gulf of Mexico on September 6th.
    Hard to believe that the CEO’s of Chevron and Devon wouldn’t have considered what effect this announcement would have on oil prices and as a consequence gasoline prices, consumer confidence and the November election. I doubt very much they’re Democrats.

  36. davy

    To take another angle on this, suppose we accept Treasuries manipulation of the gasoline markets, and then ask ourselves, was it justified? Most traders agree there was a war premium in crude, the result primarily of belligerent statements in regard to Iran. Is crude more fairly valued at current levels? Evidently the Saudis think so, and why would they give Bush a nickle when he denied them state of the art weapons systems? But then I have the same problem with the so called LLTC conspiracy,in which private bankers agreed to bail out one of their own kind. In a free economy you have a right to ask, what’s in it for me? And when someone is taking on floundering assets, the way Citigroup is taking on those of Amaranth, we have a right to ask, what’s in it for them? Until proven otherwise the conspiracy theorists deserve the benefit of the doubt.

Comments are closed.