A genuine North Sea oil platform. Not shown: boats. Photographer: Chris Ratcliffe/Bloomberg
A genuine North Sea oil platform. Not shown: boats. Photographer: Chris Ratcliffe/Bloomberg

This big crude-oil price manipulation lawsuit is gibberish but that doesn't mean it's wrong. The gibberish might be inherent in how crude oil is priced. Which makes sense; I suppose that the harder it is to explain a market the easier it is to manipulate.

The story is that some guys who trade oil derivatives on New York Mercantile Exchange are suing a bunch of physical oil traders including Royal Dutch Shell, BP, Statoil, Morgan Stanley, Trafigura, Phibro and Vitol, claiming that those traders manipulated the Dated Brent physical oil price benchmark set by Platts, the price reporting agency. The Platts benchmark, which affects those Nymex etc. derivatives prices, is set in a semi-subjective way, based on transactions executed, and bids and offers made, between 4 and 4:30 p.m. London time each day for crude delivery 10 to 25 days out. The claim is that this window was pretty thinly traded and subject to manipulation, and that Shell and BP and the rest took advantage of the opportunity to manipulate the benchmark price in ways that helped their derivatives positions, and hurt those of the plaintiffs.*

It is tempting to start from the premise that the oil price benchmark is dumb and should have been less manipulable but let's resist that temptation. You know what is a great benchmark? The S&P 500. There are these 500 stocks, and they're a big chunk of the stock market, and they trade just a whole whole lot, and all the shares of each stock are perfectly identical, and all the trades are perfectly commensurable and instantly publicly reported, and you can add them up in a sensible market-cap-weighted way and then bang, you have an index. And I'm sure there are people with S&P 500 derivatives positions who are like "ooh what I will do is I will trade those 500 stocks uneconomically to make money on my derivatives positions" but this is not a thing people worry about for two reasons:

  1. The stocks trade so much that it'd cost you a ton of money to manipulate them, making it probably not worth it, and
  2. The derivatives trade so much that even if it did make sense to manipulate them by trading the stocks, there'd be someone on the other side of your derivatives trades manipulating against you.**

Then there's oil. This complaint is full of the technicalities of renting ships to pick up oil cargoes. Reading it you occasionally get a whiff of salt air, though I guess if you're on a boat in the North Sea loading oil it probably smells more like a gas station? Anyway, the point is that there is oil under the ocean and somebody sticks a tube way under the ocean and the oil flows up the tube into a big tank in the middle of the ocean and every couple of days a big boat comes up to the tank and you move the oil onto the boat and the boat goes on its merry way to deliver the oil to someone who will turn it into gasoline and eventually light it on fire.

There are important differences between this and a stock price, and they include the oil and the tanks and the boats. You don't need to rent a boat to buy shares of Microsoft. Your shares of Microsoft aren't waiting for you in a tank in the middle of the ocean. Stock trades are all very clean and computery and fungible. Nobody gets wet, or lit on fire.

Physical oil trades are different, and the job of the oil price reporting agencies, like Platts, is to compress all this messy nautical activity into a nice simple number and call that number a price. That is just hard in a way that the S&P isn't. So you have to make decisions and those decisions will be necessarily arbitrary and someone will necessarily game them a bit. Platts reports its Dated Brent price as of 4:30 p.m. London time, and it takes into account transactions and bids and offers during a window, called the "Platts window," that starts at Time X and ends at 4:30. Time X happens to be 4 p.m., meaning that trades at 3:59 p.m. are arbitrarily excluded, but I guess Time X has to be some time, meaning there's going to be some arbitrary exclusion.***

Similarly, Dated Brent takes into account trades in cargoes for delivery between 10 and Y days out. Y is 25, though it was 21 before 2012. This arbitrarily excludes cargoes for delivery Y+1 days out but what are you going to do, that is true for any Y.

Did the big traders game those things a bit? I mean, I don't know, probably, right? If nothing else your derivative positions will affect your timing decisions: If you are short derivatives -- that is, you want Dated Brent prices to go down -- and you are also selling physical oil, you will try to sell that oil at 4:01 and will be willing to take a lower price. If you are long derivatives, you'll try to sell the oil at 3:59, to keep the sale out of the Platts window. If you're short, and you're looking to buy oil in a few weeks, you might pick a delivery day 26 days out. If you're long, you might go with 25.

The complaint is one part making a lot of hay over those arbitrary distinctions,**** one part expressions of astonishment at sub-one-percent price moves that must have sounded better when they wrote them,***** one part complaining that the big oil companies would buy oil and rent tankers without telling everyone what they were up to, and like 10 parts quoting trader whining as evidence of manipulation. Here is a representative passage:

Demonstrating the artificiality of the pricing, traders commented that the level of backwardation in October and November was much higher than seen earlier this month even as the physical market had remained weak due to a lack of demand from European refiners amid the maintenance season. "Backwardation is picking up but it's all over the place," said one trader. "It's hard to know whether that's driven by physical market strength or not." A second trader said: "I can't read physical at the moment . . . it looks a touch heavy if anything, even with cargoes going East." To effect this manipulation, Shell fixed a VLCC [very large crude carrier], the Pu Tuo San, to leave Scotland between October 10 and 15. Statoil's VLCC was set to sail from Norway between October 14 and 15. Notably, trading sources at Shell and Statoil declined to comment on the cargo movements. Shell and Staoil's manipulative use of the VLCC had artificially influenced the market.

That manipulation is: Shell and Statoil put some boats in the North Sea to move oil, which remember is their business. But they wouldn't tell everyone what they were doing with the boats. Also some traders were confused. Manipulation! I mean, maybe? But it's not evidence.

I am no expert but there's nothing here that looks like a smoking gun. There are claims that the defendant oil companies did uneconomic physical transactions, presumably to make derivatives profits, but they're pretty vague.****** There are some accusations -- like one about "the large physical cargo position accumulated by Shell as it manipulated prices lower" -- that seem to defy economic sense,******* though again that doesn't mean they're wrong. It means either that they're wrong or that this market defies economic sense, which I guess is sort of the point of the lawsuit.

More generally, you can't confuse the weakness of this case with wrongness. These guys don't seem to have any proof that anyone manipulated oil prices, but they have reason to hope that they'll happen upon some. The European Commission, the U.K. Serious Fraud Office, and the U.S. Federal Trade Commission are all looking into allegations of manipulation by some of the same oil traders here. Maybe they'll find something! These plaintiffs sure hope so; they say "The foregoing investigations are expected to yield information from Defendants' internal records (e.g., instant messages, e-mails, telephone records, Brent Crude oil trading data, etc.) that provide further support for Plaintiffs' claims." At which point they'll be all, "we told you so! Also, pay us."

Like the plaintiffs' lawyers here, I will look forward to the results of those investigations, and like them I will hope that those results include lots of dumb IM conversations to the effect of "yo bro nice job manipulating oil prices thx!!!1!," because like them I more or less make my living off those dumb IM conversations. In that capacity I am saddened by the fact that this lawsuit contains no real proof of manipulation. I guess they are too.

It's a disappointing lawsuit, but there's no reason why it wouldn't be. These plaintiffs were big market participants -- one of them was on Nymex's board of directors and executive committee -- but that doesn't seem to have gotten them any privileged understanding of how the market worked, or didn't work. After all, they were the guys on the other side of BP's and Shell's and Statoil's and so forth's allegedly manipulative trades. They were the ones who got fooled. It'd be too much to hope that they'd be the ones to provide the evidence of how they were fooled.

* The derivatives here seem to be mostly short-term contracts for differences but also forwards and futures, though it's a bit vague.

** How important is #2 compared to #1? I don't know. It's sort of my pet theory of Libor; Libor had no #1 at all -- it is basically not a traded thing -- but arguably has some #2, in that there are big banks on both sides of Libor swaps.

*** I suppose it could be, like, 4:31 p.m. the previous day, so every trade counts, but then your price starts looking a little stale.

**** There are, for instance, claims that Platts used, say, a too-low price from a trade by one of the defendants during the 10-to-25-day window, but ignored another much higher price from an independent trade for delivery in 26 days or whatever. But that's just how the benchmark works! A trade in the window counts, a trade outside the window doesn't, even if someone after the fact claims that the outside-the-window trade was more economical.

***** "Dated Brent, which priced off the Forties, fell an astounding $0.375 on that day [September 24, 2012] to minus $0.325 under North Sea Dated Strip. This was the largest one-day fall in the Forties crude prices since June 12, 2012," or in about three months, which, like, the biggest thing in three months happens roughly every three months. Also that's $0.375 on a price of over $100 a barrel so it's not that astounding.

****** For instance, BP sold a bunch of cargoes in the Platts window one time, and the plaintiffs say "BP's conduct was not economically rational for a company that had a large number of unsold cargoes," since such a company would actually sell over a full trading day rather than selling all at once in the half-hour Platts window. Which sounds plausible, but you can't really hold BP to the standard of "trade in a way that some other guys on Nymex think in hindsight would be most efficient," even if their argument for its efficiency seems pretty good.

Also, even if they're right, and the BP trader came in and knew he had to sell 6 cargoes that day for good fundamental reasons but decided to do it inefficiently at the close because that would be better for his derivatives positions -- is that manipulation? If the trader was like "I'm going to do uneconomic trades solely to help my derivatives," sure, manipulation; but if he was like "I gotta do these trades anyway and they're going to move the price against me on physical, so I might as well at least do that in a way that helps my other-way derivatives positions," that seems kind of okay to me? I'm open to persuading on this point.

******* I mean. If you buy a lot of stuff that tends to push the price up, no? "We will buy all the oil in the world in order to push its price down" is not a great cunning plan.