The United States Oil Fund, currently the world’s most controversial exchange-traded product, has just filed a staggering update to the SEC. (h/t to the FT’s Phil Stafford.)

It notes (our emphasis):

RBC and other market participants , including other FCMs, have taken risk mitigation measures that constrain USO’s ability to invest in the Benchmark Futures Contract and other Oil Futures Contracts. RBC, currently USO’s only FCM, has expressly informed USO that, until further notice, USO may not hold positions in the Benchmark Futures Oil Futures Contract and that it may not purchase any other Oil Futures Contracts for USO’s portfolio through RBC whether or not such purchases would be within the limits permitted by the exchanges. RBC has indicated that such limitation on USO is a result of RBC’s own internal risk management requirements and directions it has received from other regulators in the United States, Canada and the United Kingdom. USO cannot predict with any certainty when and whether RBC will remove these limitations. USO has been engaging in efforts to enter into additional FCM agreements for the purpose of the purchase and sale of Benchmark Oil Futures Contracts as well as other Oil Futures Contracts. To date, USO has not entered into an agreement with any FCM other than RBC and it cannot predict with any certainty when it will do so.

RBC Capital was the USO’s only clearer. RBC is no longer prepared to clear USO’s portfolio. It is specifically not prepared to clear USO holdings in the benchmark front month but also, rather staggeringly, in any other months either.

It also appears USO is having trouble finding an alternative clearer to take its place.

What this means for the oil market at this point is unclear. In the strictest literal interpretation of the filing, it could mean that unless it finds an alternative clearer pronto, the USO will be forced to liquidate all its holdings before the next roll period. A looser interpretation might be that it is allowed to hold onto existing positions but not be allowed to add new ones. In a continuing contango market this would see it run down its overall position over time.

As to RBC’s own internal risk issues, it’s worth noting that RBC is not only USO’s clearer but also an authorised participant to the fund.

This is important because as John Hyland, the fund’s former chief investment officer, explained to us a few weeks ago by email, how the USO operates in practice differs to how it operates in theory, according to its prospectus.

The nature of the beast is such that, for practical purposes, authorised participants tend to become the counterparts of the derivatives the USO acquires. So the long positions the fund holds are offset by short positions taken on by the fund’s authorised participants.

As Hyland explained to us (our emphasis):

Creations and redemptions for USO were for cash, not “in-kind” like most stock ETFs, but with a twist.

I assume because technically a future is a contract, and not a security, you cannot transfer ownership from one party to the other like you could with 1,000 shares of APPL [sic]. So when I created the 2nd commodity futures-based ETF/4th commodity ETF with USO in 2006, we could ONLY handle creations and redemptions by having the AP send us cash in exchange for shares for a creation, or receive cash in exchange for shares on a redemption. That is still how the AP documents read as far as I know.

This of course was not ideal, but the fairly Dickisonian nature of futures exchanges at the time left us with no alternative. I mean you could not even trade WTI contracts electronically in 2006 or early 2007! Eventually, with some pushing by me, not only did NYMEX go electronic with WTI but they also introduced the ability to do block trades “off exchange”. Once that happened we largely moved to a work around to the creation/redemption process. What then mostly happened is an AP orders say a 1 million shares creation which is equal to 80 WTI contracts. At the same time they enter into a block trade with USO to sell the ETF 80 WTI contracts that day and tied to the end of the day settlement price (a trade order known as TAS, meaning Trade At Settlement). A redemption is just the reverse. So instead of a single transaction, as with an in-kind creation/redemption, USO actually had two transactions. This was still not quite as efficient or cheap as doing true in-kind, but it did largely eliminate tracking error from USO having to buy or sell its own contracts in the market to offset a creation or redemption.

For an AP taking the offsetting short to the USO’s long, it is essential to have the capacity to offset that position fluidly and cheaply so as to remain market neutral. That, however, becomes more difficult if you are forced to operate further down the curve where liquidity is far lower.

Which introduces the question: just what are RBC Capital’s exposures to the USO at this point?