The First of Many
Unintended consequences, that is.
ETFs have been a big deal for several years now, including commodity ETFs. Heretofore, the big commodity ETFs have been in energy, but now several firms have metals ETFs in the offing.
Most traditional ETFs have bought futures contracts. So the US Oil Fund, for instance, buys NYMEX light crude futures. As the contracts it holds near expiry, it rolls forward into deferred contracts, buying the deferred and selling the nearby.
But the new metals ETFs will invest in, and hold stocks of, physical metal rather than derivatives. FT Alphaville, quoting from a JP Morgan prospectus for its new copper ETF, has the details:
Firstly, being physically backed and not futures-based — the trust will escape all regulatory restrictions governing the size and scope of its speculative commodity positions (since there’s no exchange presence). No USO or UNG implosion danger here.
As the documentation states:
The Trust will take delivery of Physical Copper in the form of LME Copper Cathodes. Because the Trust will not trade in copper futures contracts on any futures exchange, the Trust will not be regulated by the CFTC under the Commodity Exchange Act as a “commodity pool,” and will not be operated by a CFTC-regulated commodity pool operator. Investors in the Trust will not receive the regulatory protections afforded to investors in regulated commodity pools, nor may the COMEX or any futures exchange enforce its rules with respect to the Trust’s activities. In addition, investors in the Trust will not benefit from the protections afforded to investors in copper futures contracts on regulated futures exchanges.
Note, too, that as a result it is likely that the ETF will not be subject to position limits. And I think this is a big part of the reason for going physical.
The construction of the fund to avoid regulation is clear as day. The effects are not likely to be good. The advantage of using derivatives (futures or swaps) to construct ETFs is that it makes it possible to achieve an arbitrarily large exposure to the price risk of the underlying (oil, or copper, or whatever) without ever handling the physical. That is, derivatives allow the unbundling of the price risk and the holding of the physical product. This lowers transactions costs, and crucially permits investors to achieve portfolio and risk objectives without being constrained by the size of the physical market.
In contrast, physical ETFs cannot exploit this unbundling. But apparently, the benefits of avoiding position limits and other regulatory costs are sufficiently great to justify the loss of efficiencies associated with this unbundling. (If the bundled risk-physical solution were better than the unbundled one, you would have seen physical ETFs when the regulatory constraint wasn’t a real threat.)
There is a perverse irony here. The whole rationale (supposedly) for position limits is that speculation somehow distorts physical markets. There is precious little evidence, outside of a few extreme examples (e.g., the Hunts) that this is a real problem. But by driving those that want exposure to metals prices, either for speculative reasons, or for portfolio balancing reasons, regulations are making it more likely that speculation will distort prices and the physical markets.
I’m not saying this will happen, or will be chronic. Only that distortions are more likely in the bundled world than the unbundled one in which ETFs trade in derivatives rather than the physical. ETFs that roll aren’t in the physical market, rolling out of expiring positions before they can go physical. This limits, and likely eliminates, their effect on prompt prices and consumption, production, and storage decisions.
In contrast, ETFs that hold the physical are by definition in the physical market, and there can be a conflict between the risk and profit objectives of physical ETF holders and the efficient allocation of physical supplies; those desiring exposure to metal price risk might hold onto stocks when it would be optimal to consume them instead. Conversely, in an unbundled world, traders can maintain exposure to the price risk without having any influence on the use of the metal.
I can also imagine some manipulation strategies that exploit the physical ETF. For instance, somebody could obtain a big physical position via the ETF, and use that to create or enhance market power in the derivatives market. That is, the ETF can be a way of “locking up” metal that would otherwise be available for delivery, thereby enhancing market power in the derivatives market.
I said this was a perverse irony. It’s actually more perverse than ironic. A policy intended to reduce (probably chimerical) distortions in the physical market actually increases the likelihood that such distortions will occur. That’s the kind of havoc that Sorcerers’ Apprentices wreak. Get used to a lot of that going around.
Hmm. Not sure really.
Taking physical delivery of LME copper is, well, not really taking physical delivery. The cathode stays in the LME warehouse, all that moves is the piece of paper stating who owns it….and the storage bill. So it seems like a reasonable enough method of speculating on copper prices.
The other thing is that while the trust may not be subject to US regulation on size or positions: assuming that it’s trading on the LME (reasonable enough) it will be subject to LME position limits.
Comment by Tim Worstall — October 27, 2010 @ 12:25 am
Tim–having been deeply involved in the design of 4 major commodity futures contracts, I’m intimately familiar with the mechanics of physical delivery and warehousing. When I say “physical delivery,” I mean taking ownership of a warehouse receipt (or a shipping certificate, as under the contracts I’ve helped design).
Under many circumstances, speculation via warehouse receipts is reasonable. But the important thing to note is that ownership of that piece of paper gives said owner control over the disposition of said physical copper. If an ETF owns the paper, that copper will not be available to be taken out of the warehouse to be consumed. That constraint does not exist if the ETF deals in futures or other derivatives. It’s econ 101 that adding constraints is costly, or certainly cannot reduce costs. In this context, it creates a potential conflict between use of that copper for consumption purposes, or use as an investment vehicle. There are circumstances in which that conflict could lead to a worse outcome than if the constraint did not exist.
There’s one possible way out of this. There are borrowing/lending markets for physical metal. These are, effectively, derivatives transactions, but constructed somewhat differently. I’ll have to read the JPM prospectus to see whether the ETF will engage in borrowing or lending.
Craig
There’s a fundamental issue here behind what I believe is a misconception by the man on the street, not to mention the regulator and politician on the street.
That is that the motive of a ‘speculator’ is transaction profit (aka greed is good), and he/she is agnostic as to market direction. To a speculator, volatility is good, and the only bad news is no news at all.
But at the zero bound in dollar rates, there are a gazillion dollars out there whose motive is not greed, but fear. These investors aim not to make transaction profits, but rather to avoid wealth losses.
Through participation in long only ETFs and the like they off-load dollar risk and take on eg commodity risk and they are the natural counterparty to a producer hedging. Like a producer hedging they hate volatility; and may get hit by adverse market structure (which if their size is big enough they bring on themselves – shades of MG) and they are always subject to the casino’s – sorry exchange’s – take on a wheel with maybe six zeroes, which is why they will cut out the middleman if they can. eg the bilateral arrangment between Shell and ETF Securities where Shell is essentially monetising oil in tank; in transit, and probably even in ‘reservoir storage’.
My take is that through the Brent/BFOE complex one or more producers are – with the connivance of investment banks – essentially borrowing dollars at zero % from ETFs and structured funds and to all intents and purposes conducting oil market repos like a Central Oil Bank conducting open market operations while keeping the price supported in the rather illiquid BFOE complex (it doesn’t actually take that much with the BFOE fields in secular decline). The effect currently is of a rough peg to keep the oil price at or near the upper bound where demand destruction sets in, rather than – as with the gas market – at the lower bound where production destruction sets in. Note how the historic relationship between crude and natural gas has broken down.
If the copper market manipulation by Sumitomo/Hamanaka showed us anything, it was to demonstrate once again (tin crisis anyone?) that if producers CAN manipulate prices, then they WILL, and also that manipulation can go on for a very long time before being exposed, and just as long again after the whistle is blown.
Comment by Chris Cook — October 27, 2010 @ 11:08 am
I read the JPM prospectus. It says explicitly it will not engage in borrowing transactions with “third parties.” It makes no mention of engaging in lending transactions. I presume that the first party is the ETF buyer and the second party is the ETF itself, which would also rule out engaging in borrowing/lending transactions with JPM or its affiliates, but I didn’t have time to read the prospectus closely enough to know whether that’s definitely true.