Professional Oil Investors Crushed By Duplicitous Nature Of Oil, Mathematical Ignorance

by: Charles Armstrong

A few days ago, the Wall Street Journal reported that "some of the crude [oil] market's most high-profile investment firms, including hedge funds run by former Goldman Sachs Group Inc. traders" lost big on a wrong-way bet on the NYMEX WTI / Brent crude oil spread.

In the past, I've written extensively about the differences between Brent and West Texas Intermediate, and back in April of this year, I advocated a strategy for using the two blends' representative ETFs (USO and BNO) to take advantage of futures chain inefficiencies, but to give you some brief background: the American crude oil standard is "West Texas Intermediate". The name (abbreviated WTI) refers to a very specific type of crude, one of many blends bought and sold in the physical US oil market. However, for whatever reason, WTI is also the benchmark for US crude prices in general, to the point that the front-month futures contract is typically what is being reported when a news outlet lists the price of "oil". In this way, WTI (traded on the New York Mercantile Exchange and so also called "NYMEX Crude") is both the global benchmark for US oil, AND a specific physical commodity, which happens to be delivered only to Cushing, Oklahoma (for an interesting conversation on what that means to the broader US oil market, see the discussion I had with commenter "Alex_G" in the comments section of the first aforementioned linked articles). Brent oil, by contrast, is a British blend of crude sourced from the North Sea and historically considered the European benchmark. Technically, WTI is a higher quality oil than Brent, which meant (again, historically) that it traded at a premium. Since 2008, however, for myriad reasons outside the scope of this post, that relationship changed, and Brent has become the global standard and the pricier of the crude blends. Still, both of these things are oil, so it makes sense that their prices would move together, even if one is viewed worthy of a premium.

You can read the WSJ article for the play by play on what happened, but the gist is that several energy hedge funds sought to squeeze blood from the stone of the languid oil markets by playing the Brent/WTI spread. More specifically, the funds bought NYMEX crude contracts, while selling short Brent contracts. Because the two oils are supposed to move in a tight correlation (they are the same thing, after all), this was in theory neither a bullish, nor bearish bet on oil in general; rather, it was simply a bet that American crude prices will rise relative to European (Brent) crude prices. And, on the surface, that's not so crazy a bet. WTI is, by definition, a superior grade of crude; its lower sulfur content and density make it the easier of the two to refine. So, in making this investment, the hedge fund managers were essentially betting that, regardless of whatever happened in the oil market at large, WTI and Brent would move to (or toward) their historic relationship with American crude on top. That I am writing this piece should tell you this did not happen.

While the strategy seems sound, unfortunately, the assumptions behind the strategy were not; despite their both being crude oil, WTI and Brent no longer behave as two flavors of the same commodity, but rather as two different things entirely. Allow me to demonstrate; here is a chart showing Brent and WTI prices from November 2007 until now (so, almost exactly six years):

Brent and WTI, 11/2007 - 11/2013
There are a few things to notice here. First of all, initially, the graph is basically a single purple line until Brent pulls significantly away at the end of 2010. Still, despite the bigger spread, the two oils continue moving in more or less lock-step until the end of 2012, when each seems to take on a mind of her own. Curious about their correlations in each of the time frames, I pulled some data and ran a linear regression. Here is a scatter plot of continuous front-month futures prices for WTI and Brent crude from January 1st, 2008 through December 31st, 2010:

Crude vs Brent, 2008 - 2010

That is a near perfect correlation. The linear regression model gives me an r-squared of .992, which is just about as highly as two supposedly different things can possible be correlated.

Next, a scatter plot of WTI and Brent from January 1st, 2011 through December 31st, 2012:

Crude vs Brent, Jan 2011 - Dec 2012

The correlation is still there, but the dots are moving further from the line of best fit. My linear regression model gives me an r-squared of 0.633, meaning about 60% of the variation in the price of one type of oil can be explained by its relationship with the other. An r squared of .633 is nothing to scoff at, but it is certainly much lower than we'd expect for two slightly different versions of the same thing.

And finally, here's WTI and Brent from January 2013 through last week (11/13/2013):

Crude vs Brent, Jan 2013 - Nov 2013

Now we're basically looking at schematics drawn up by Dr. Ian Malcolm. Since the beginning of this year, the commodities markets have treated Brent and WTI as two completely different, unrelated things. With an r-squared of .118, implying little to no correlation, you might as well have been going long Palladium and shorting Distillers' Dried Grain. Until I began researching this piece, I had not realized how large the gulf in price correlation had become, and in my previous articles about WTI and Brent, I kept trumpeting that they were still the same thing and, therefore, should move in synch regardless of spread size. I see now, much to my honest surprise, that assumption was wrong, as the two oils seem to behave entirely independently. I would like to sincerely apologize if my prior analysis set anyone on the wrong course. Of course, I didn't have millions of dollars of other people's money on the line, otherwise I might have undertaken this relatively straightforward statistical analysis sooner, but hey, we can't all be "professional" hedge fund managers, right?

It would be next to impossible to point to a single reason for the two crudes' divergence, so complex is the global oil market. However, there are a few broad trends that go a long way to explaining what may be happening here. Remember that, while indeed a "pricing" benchmark for the global economy, unlike stocks on the Dow, oil eventually gets burned, literally. As such, being an actual, real world, useful good, speculation can only drive the price so far before the immutable laws of supply and demand take effect. (Incidentally, this is the very reason the commodities markets are to me the most fascinating, but I digress.)

The first broad trend I'd consider when trying to explain petroleum's newly duplicitous nature is the major shift in domestic petrol-politics that's taken place in the last 16 months or so. You may remember that last year it was announced the US is on pace to overtake Saudi Arabia as the world's leading producer of crude, and yet, it is still functionally illegal to export crude, a restriction stemming from an antiquated, protectionist, cold war mindset. Meanwhile, America's appetite for oil has eased over the past decade. So, in the US demand has been falling, supply has been rising, and exports are illegal. (And the aforementioned pros were betting the price of oil would go up.) We also continue importing tremendous amounts of foreign oil, a fact which is somewhat insane considering we now, for the first time since 1995, produce more crude domestically than we import. Meanwhile, the market for a slightly inferior grade of crude oil across the pond is open to the world, and as more nations continue developing, so grows their thirst for the black stuff. The head of any multinational business using oil futures as a hedge (as they were meant to be used) would obviously want to buy or sell the contract that reflects the global price of oil. Think of it this way: say you own a long-haul trucking company in India. Since you're able to anticipate about how much oil your fleet will use in a given year, you hedge your costs and eliminate uncertainty by taking positions in the oil futures market. Now, which contract are you going to buy? The one that represents the price of crude delivered to Cushing, Oklahoma, which happens to be completely inaccessible to you both geographically and politically? Or the futures contract that everybody else in the world is using, reflecting the actual global price of oil?

The other broad trend that comes to mind has to do with WTI's strange existence as both pricing benchmark for American oil in general, and one of several specific types of crude bought and sold in domestic markets by US oil producers and consumers. The price for WTI may have just as much (if not more) to do with accessibility and price compared to other domestic blends for which no futures contracts are traded on major global commodities exchanges (e.g. Louisiana Light Sweet Crude). It would, for example, be possible for WTI's price to move dramatically relative to other blends if, say, the pipeline infrastructure in Cushing, Oklahoma were inundated with months of horrendous tornadoes, even if the rest of the country's oil market were stable and healthy. I must step back here and admit that I know very little about intra-nation domestic oil trading, and so am not qualified to comment on the extent to which pressures upon the various US blends are influencing WTI prices globally (indeed, the aforementioned comments-section discussion with "Alex_G" is a testament to just such an ignorance on my part). However, my general understanding is that among crude oils domestically, demand for WTI has been falling in favor of LLS, as was explained very well in a report from the CME website a few months back. If LLS has indeed become the new American standard for domestic oil, we would expect to see an increasing correlation between the price of LLS and Brent globally, as WTI falls off the international stage. Unfortunately, not trading as a straightforward commodity future, I could find no historic LLS price data easily accessible with which to perform this analysis. If someone knows how to come by it, please let me know in the comments below.

Whether WTI's diminishing correlation has to do with political factors influencing economic activity, or the rise of LLS in domestic oil markets, or neither, or both, the overarching point is the same: the price of West Texas Intermediate is no longer correlated with the price of Brent crude. It's just not. The tremendous historical correlation that existed just three years ago has quickly decayed into meaninglessness. Any time before 2011, taking an opposing spread position in WTI and Brent (that is, buying one and shorting the other) would have been one of the safest bets in the energy markets. Since then, the correlation between the two has diminished into nothingness; buying one and shorting the other is no more meaningful than taking two opposite positions in securities chosen at random.

If the hard-hit fund managers were really trying to make a conservative bet on the spread, then clearly, they didn't do their math homework.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.