$5 Gas? Oil Speculation and Politics

In the last 5 years, the volatility in the price of oil has led to debate, accusations, and a great deal of confusion about the role of the financial markets in determining the price of a gallon of gas. This has included congressional investigation, consumer group accusations, and finally new regulations for the commodity trading industry.

Brent Spot Price of Oil (EIA, 2011)

So what is the basic issue?  Critics allege that traders can artificially manipulate the price of crude by artificially creating demand through purchase of futures and other financial derivatives.  Once the price increases, these can be offloaded for a profit.  Last year the Commodity Futures Trading Commission (CTFC)—the primary government agency tasked with regulating the commodities market—filed a major suit against 5 traders alleging that they cornered the market on oil futures by obtaining such a dominant position that they owned 2/3rds of all the available oil at the Cushing, O.K. exchange.  Subsequently, the CTFC was given the authority through the Dodd-Frank financial reforms to limit the supply that any one trader can own, which it voted to enact beginning in 2012.

2 weeks ago, at the UT-Energy Forum, students heard from a panel of financial industry experts on the topic of movement in oil prices.  Panel member Jason Schenker, of Prestige Economics, strongly rebuffed the idea that financial speculation is a key driver of oil prices.  Instead, Schenker cited basic macroeconomic principles of supply and demand, which was supported by other panel members. This assumption is essentially the energy industry’s institutional answer, supported by major groups such as the American Petroleum Institute (API).  Research by the EIA in 2008 backed this finding, concluding that it was unlikely that inflows of investment or hedging in the commodities markets could increase prices.

Ultimately, this question remains unresolved. In the wake of the global financial crisis, it is unsurprising that the public and consumers are wary of traders and investment bankers.  While the congressional report states that “Addressing excessive speculation offers the single most significant opportunity to reduce the price of gas for American consumer,” it is too soon to tell gauge the impact of new regulations imposed by the CTFC.  However, given the amount of conflicting evidence and disagreement over the role of speculators in the oil market, regulators should proceed cautiously.

Advertisements

2 Comments

Filed under energy

2 responses to “$5 Gas? Oil Speculation and Politics

  1. danielnoll

    Good points Reed. I think it’s always useful to point out, as Niloy Shah from BP did on that panel, that the biggest drivers in oil price are supply chain constraints (#1) and geopolitical risk (#2). If you look at events in the Middle East, geopolitical crisis and oil price correlate pretty well, such as how the combattive rhetoric flying between Iran and Israel recently has kept the price per barrel above $100.

    Except for Saudi Arabia and maybe one or two national oil companies, market players don’t have the flexibility or size to influence price. What is necessary is to make sure that some regulatory limits prevent financial players from becoming “Saudi Arabias”. The best idea to do this has already been thought up. Dodd-Frank empowered the CTFC to cap the number of futures contracts a trader can hold. Unfortunately this rule has yet to be implemented. Our regulatory system is bloated with overly complex rules (as this week’s Economist cover article will explain), but the ones that count seem to fall through the cracks sometimes.

  2. mattawright

    I agree that in the long run, the price of oil can be thought of as primarily a function of demand, but predicting the short-run price movements of futures contracts seems to be less certain. This was illustrated this past week as Goldman Sachs issued a prediction that the price of West Texas Intermediate (WTI) would continue rising in price until at least September [1], narrowing the WTI-Brent futures spread. Goldman makes this prediction, as the linked article mentions, in a fairly contrarian environment. Many, including analysts at Citigroup, predict otherwise, or that the price of WTI will not rise relative to that of Brent and the spread will expand.

    The supply-and-demand and geopolitical motions here seem fairly clear. While both oil benchmarks have seen their prices increase, the price of WTI has fallen relative to Brent in recent months due to an expansion of North American production from oil sands and the resulting increase in supplies at Cushing. Brent has increased more than WTI recently due to still-reduced production capacity from the Libyan conflict and general market anxiety about supplies from the Middle East. Citigroup and others see these trends continuing more or less as they have been, but Goldman’s analysts think that a projected pipeline flow shift to move some Cushing oil to the Gulf will reduce the WTI stockpiles and bring prices closer to Brent. The takeaway here, as one of the analyst quoted in the Bloomberg piece notes, is that while both prices will likely still go up.

    The reason that the WTI and Brent prices are so important, by the way, is that most supplier-purchaser contracts are written such that payment is made at some benchmark price (WTI, Brent, etc) plus or minus an adjustment for the quality of the fuel in question [2]. What’s interesting here is that this means that the exact amount of money exchanged between the purchaser and supplier at the settlement of a private contract can be influenced by global market movements. This quirk of crude delivery prices being dependent on an external financial indicator is a unique historical legacy of the oil market specifically, and is not shared by most other commodities that are openly traded, such as agriculture products and other energy products like natural gas. Unfortunately, this is a often-overlooked fact by most financial professionals that have backgrounds in other commodity sectors when commenting on oil prices, and is likely a source of much of the confusion when discussing oil prices.

    ===========================

    Related to this, I wanted to take a look at how reliable oil futures were at predicting the actual spot prices at their settlement date, and performed a quick analysis using historical spot and NYMEX WTI futures contract prices available from the US Energy Information Administration’s (EIA) website [3, 4].

    The first graph shows the month-by-month history of the absolute pricing differentials between the WTI spot price on a contract settlement date and the associated futures contract price one, two, three, and four months before the settlement from 2000 on. As one would expect, the futures contracts were always off of their eventual closing price, and the more distant predictions (the four- and three-month futures prices) generally were the most erroneous. This confirms the intuitive thought that, although futures prices are an imperfect judgment of spot prices at settlement, they draw closer to the actual price as settlement nears (if I hadn’t observed this, I knew I would have done something wrong).

    I also made a few histograms of each error sequence linked above to see how the pricing differentials were distributed. If futures contracts are to be taken as an estimation of the eventual settlement price, then the pricing predictions should converge to the actual settlement price. The histograms for the results are linked below.



    As one would expect, the data does seem to be relatively normally distributed. The large outliers on the left side of the histograms correspond to pricing errors in contracts expiring in late 2008. These contracts, in August or September, were estimated to be worth a great deal more than they actually turned out to be worth at expiry in October or November. While the data does seem to be at least somewhat normally distributed, it turns out that the mean price differential is actually not zero. This is summarized below.

    Prediction n sample mean std dev
    1-month 144 0.5051 16.3698
    2-month 144 0.8990 13.5077
    3-month 144 1.1335 10.2137
    4-month 144 1.5033 7.3586

    The large standard deviations can be traced back to the huge 2008 outliers, but what’s interesting is that even by weighing the highly negative outliers with the same weights as any other sample (that is, not discarding them as aberrant), every pricing differential averages above zero. That is, the data seems to suggest that the market tends to predict that the price of oil at a future’s contract expiry will be higher than it actually turns out to be. Is the oil futures market suffering from irrational exuberance?

    This analysis is very rough and should not be taken as gospel. A more stringent analysis would include a test for statistical significance of the sample mean offset, a study for correlation with some other economic indicators such as the VIX, and especially an analysis of the first and second derivatives of the price, or how the motion itself of the futures price moves.

    [1]: http://www.bloomberg.com/news/2012-02-24/goldman-bets-on-rising-crude-oil-with-surging-u-s-supply-energy-markets.html
    [2]: http://www.eia.gov/pub/oil_gas/petroleum/analysis_publications/oil_market_basics/price_transactions.htm
    [3]: http://www.eia.gov/dnav/pet/pet_pri_spt_s1_d.htm
    [4]: http://www.eia.gov/dnav/pet/pet_pri_fut_s1_d.htm

Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s