I am wondering if anyone out there is familiar with a trading opportunity called by some, the Goldman Roll. As it has been explained to me, there is a large numbers of long-only commodity funds. As a given contract that they hold long, say oil is coming due to expire they need to sell that one and then roll into a long position in a further out contract. This creates a very definite trend in the spread that can be exploited. Sell the near one short and buy the next one out. As the roll transactions are executed the Far minus the Near spread has a very predictable and smooth rise. It is claimed that this phenomenon has not be widely recognized and thus remains in existence thus far. Any comments out there on this claim would be appreciated.

Dr. Aronson is author of Evidence-Based Technical Analysis, Wiley, 2006

Nick White comments:

Goldman Roll? More like market roll!

This has been around as long as futures have existed and is nothing sinister. However, as some here were actually around when modern exch. traded futures began, I shall defer to them.

You can maybe get some clue as to roll direction by looking at open interest depending on the contract, but it's not always a good guide. Worth bearing in mind that people hold offsetting positions and much also depends on commercials vs specs etc.Also, if it were that easy to make money, it wouldn't exist…

Michael Cohn writes:

There is index money invested in commodity Indices and a plethora of ETFs. For example, USO or UNG. These commodity ETFs hold futures and there is a need to roll the contracts in a somewhat predictable way although there is now more flexibility as to day. This long exposure always has to sell the near and buy the far contracts. It is fairly easy to see the amounts involved…

David Aronson replies:

Goldman Sachs

Yes, I am on the lookout for all of these creatures. But kidding aside for the moment, are you saying that the claim that such an opportunity exists is on par with sightings of Big Foot? i.e., it's nonsense?

Russ Herrold writes:

It is a safe statement that there are and will always be 'unknowable unknowns' out there in the woods, and that the 'Absence of evidence is not evidence of absence' (but rather sometimes, just a statement that we cannot prove a hypothesis with our current tests and tools)

If I had a Bigfoot in my basement that laid gold bars, I would never reveal that secret, and take great pains to keep that 'trade secret'.

If I had engineered a winning strategy, I would certainly consider sowing disinformation and negative results and disinformation, to lead people seeking to reverse engineer my results, down into blind allies.

I think as a careful investigator, all we can say is: We do not know of a public proof that such exist.

By co-incidence, I am wearing a tee shirt today of a Unicorn, feasting on roast leprechaun, and as she takes knife and fork to her meal, the magic rainbows are let out.

Ken Drees adds:

The idea of taking advantage of a robotic function (mindless ETF doing its monthly maintenance) makes sense; once you notice the ripoff, wouldn't a hunter now wait for the fox?

Tom Printon writes:

I used to fill the GS roll in the coffee pit. Locals typically positioned themselves one to two days ahead of GS. When and if profitable was usually good for few tics, but one had to have size on to be worth while. Off the floor trader's vig would be difficult to overcome.

Paolo Pezzutti adds:

This reminds of "The Night Of The Long Knives" also called Operation Hummingbird. It is interesting how the market was "prepared" for this event that occurs after an impressive up leg. We will see if the event will be able to trigger more volatility. It will say a lot about this market.





Speak your mind


Resources & Links