Goldman On The Price Impact Of Physical Metal-Backed ETFs

In an interesting piece released last night, Goldman's Joshua Crumb, part of the same team which so far top-ticked gold to the millisecond, with its October 11 recommendation to buy gold at 1,364.2/oz, a call we suggested should be faded, looks at the role of suddenly popular physical metal-backed ETFs on actual price formation. The firm concludes that "we believe that under certain structures a financial product that takes physical delivery could impact short-term prices. However, we believe that any meaningful impact would not be sustained beyond the short-term as resulting market tightness would quickly generate incentives for investors to sell the metal back into the market." Whether this is a tongue-in-cheek rebuttal aimed at calls for acceleration physical delivery demands from various vaults, as an attempt to curb the same market manipulation that the CFTC disclosed yesterday as rampant in the silver market is unknown, but very likely. Of course, Goldman being part of the same syndicate, and now looking to offload gold courtesy of its repeated gold price target increase, looks at such generic interpretations of market tightness as contango curves to extrapolate that based on an expected supply/demand curve which is defined by the market itself (and thus provides absolutely no new information on what physical ETF price impact could be), that even a complete extraction of all underlying commodities via a physical ETF, the impact would be minimal. Obviously, this very theoretical interpretation is very much open to practical (in)validation, and should rampant money printing persist, Crumb's view will very likely be challenged and quite soon.

Going back to the Goldman analysis, here is how the firm views the market impact of physical ETFs:

Physical metal backed-ETFs continue to be a key topic of interest in base metal markets. Ultimately, many market participants see these products as “new” demand that could change the balance of supply, demand and inventory  fundamentals, and therefore influence prices. We have often argued that there is little evidence that broad investor flows into commodity futures drive price levels (see Speculators, Index Investors, and Commodity Prices, June 29, 2008), while we caution that under certain structures a financial product that takes physical delivery could impact short-term prices. However, we believe that it is important to differentiate product types for their economic impact, as well as assess the potential levels of “new demand” that would be needed to influence short-term price.

Ultimately, we don’t believe that these products will impact the market beyond the short term. This is because investors have no incentive to hold material once the forward curve flips into backwardation, unlike industrial users that will hold inventory even in a shortage because they derive benefit from using the material in their end products. Specifically, even if an ETF were to buy sufficient material to drive a market into shortage, it would then shift the forward curve into backwardation and there would be an arbitrage to sell the storage-paying ETF and buy lower priced futures, thus releasing the material right back into the market to ease the shortage. We believe such a product would have to assume  a large degree of irrational investor bias to ultimately disrupt the market and create significant “new demand”.

Oddly enough, it is precisely the same "irrational investor" bias that has been discounted ever since gold was at $300. Perhaps it is time to acknowledge investor bias as nothing more than a rational response to irrational central bank behavior? Just sayin'. Furthermore, we pretty much disagree across the board on the whole backwardation analysis, which is really only of benefit to industrial producers, whereas those who trade commodities (speculatively) in response to central bank actions, are far more concerned with spot. As for assorted interpretations of contango, we wonder just how much an efficient market may have priced in the possibility that China and now Korea are in fact minimizing PM output from their extraction facilities... If the recent Rare Earth Mineral scandal is any indication, look for China, and its immediate Asian sphere of influence which most certainly includes massive gold producer Indonesia once the OPEC cartel is recreated for gold, to not only collapse supply but to accelerate demand. But of course, the backwardation curve tells us all of that...

Here Goldman attempts to further attentuate the concerns of JPM's PM traders that Sprott-like ETFs can force remove tons of gold from circulation at a whim:

Specifically, in order for a physical ETF investment to materially impact prices on a sustained basis we believe two conditions would need to be met:

1. The ETF taking delivery would have to be large enough to tip the balance of spare capacity and inventory availability, which would be unlikely as investors do not have an incentive to hold metal once this “new” demand flips the curve into backwardation. This is because there would be an arbitrage to sell the storage-paying ETF and buy lower priced futures, unlike industrial consumers that derive a benefit from consuming the metal in their product;

2. The inventory accumulated would have to be permanently removed from the market, which may not be the case depending on how the ETF is structured, and this removal would have the most impact as it is occurring, as the shifting of physical material into the ETF would represent “new” demand for the metal, which would cease once the ETF was capitalized.

Addressing the second condition, we believe that the structure of a physical ETF would be important to short-term prices. In particular, if the position was built in a closed-end fund with no lending of metal accumulated, metal could be “permanently removed” from the market, as shares of the ETF would essentially be traded like shares in a corporation, with the net asset being the stock of metal. The only way this material might re-enter the market would be if an end user or government entity made a takeover offer for the stock of metal. Further, this removal would have the largest impact on the market as it is taking place, with the shift in physical material reflecting new demand that would disappear once the ETF was capitalized. Although the market would ultimately have less inventory to meet demand once the ETF had been capitalized, the market would quickly price this lower inventory level, similar to the way in which the market prices the loss of inventories as they flow into strategic reserves such as China SRB’s copper reserve or global oil SPRs. However, after an initial build, there would still be an arbitrage to sell the ETF and buy lower priced futures in a tight market, thus limiting the disruption of supply damage as the fund would theoretically have enough selling pressure to trade below NAV and inhibit the ability to issue new shares and accumulate more stock.

In contrast, if a potential physical fund had a structure that both bought and sold metal with flows into and out of the product, we believe that the fund would be just another financing vehicle for inventory accumulation.  Ultimately, such a fund would be expected to buy low and sell high, with their capital paying to accumulate and store inventory in time of surpluses (low prices), and release stock with capital appreciation in times of shortages (high prices). However, such a fund could also build inventories that might be a little more “sticky” than traditional working capital management, and also potentially add some initial disruption as a new pipeline of working capital builds in the market, though this remains to be seen.

What Goldman fails to realize in its dogged pursuit of Gold ETFs (which is what this ditty is targeted at) is that the equilibrium of market entrance and exit is terminally offset by such monetary factors as a few trillion of pieces of linen printed at a whim by the central banking syndicate. As such the entire argument pretty much falls flat on its face. But, once again, Goldman's greatest flaw is assuming that the market is responding in an inefficient way to efficient monetary stimuli, when it is in fact precisely the opposite. We are confident Crumb will figure it out shortly.

And here is again how Goldman rationalizes irrational behavior in rational times, once again completely fudging cause and effect:

Ultimately, a physical base metal ETF would be a product that allows those without the ability to buy futures to take a metal position in a time of surplus with forward views of metal appreciation, thus bringing new capital to the market and financing the storage of metal for future use. In order for the product to be seen as store of value, the buyer of the ETF would have to believe that there is sufficient shortage in the future at the entry price as to earn a return great enough to overcome the opportunity cost of forgoing a dividend or yield on that capital, on top of paying the fees for storage. Fundamentally, we don’t see a physical ETF as anything different than the contango in a futures curve, and not “new demand”, but just another player in the fundamental role of financing the storage of a commodity needed in the future.

Yet even Crumb acknowledges that in tight markets (and somehow spot gold, which now has $10 bid/ask jumps is not considered tight), the price impact of physical ETFs is quite palpable (for now short-term; when this report is revised in one year, we will also see the long-term optionality):

However, what can’t totally be discounted is that a closed-end product launches anyway in a tight market like copper or tin, therefore exacerbating short-term price movements by deploying committed capital into a deficit market, thus bringing forward critical shortages by removing inventory or competing for material in a deficit market. Further, even in an open fund that buys and sells material, launching an initial position in a tight market would likely bring short-term volatility to the market, much as a large new consumer entrant would by building an initial working capital. It is important to distinguish base metals from gold in this context, as base metals have a large end-use consumer base that require the material to produce product as opposed to gold which has little industrial usage competing with monetary demand.

Looking across metals, we don’t see potential for a fundamental disruption from a physical ETF in aluminum, nickel or zinc, as these markets are expected to remain in surplus over the coming year and have ample inventory to handle the initial shock of a new storagemarket entrant, even if there were some very short-term disruptions upon launch depending on the size and structure of the fund. However, a physical ETF in copper might exacerbate short-term volatility and perhaps bring forward our forecasted deficit and extremely tight, backwardated market. Based on our balance tables and current exchange inventories, we believe an initial aluminum ETF would have to be well over a million tonnes to disrupt the market, over half a million for zinc, and over 50 thousand tonnes for nickel and copper. By value, this would equate to over 2 billion US dollars worth of aluminum, over 1 billion USD in zinc or nickel, and nearly half a billion USD for copper. Though these are not large numbers compared to the size of other asset markets, or even the roughly USD75 billion in physical gold ETF holdings, collectively or individually they are large amounts for an initial position to launch, which is where we see the most volatility created from these products. Further, in copper, where there is currently a backwardation looking out past three months, that initial position would have to be purely coming from investors without the ability to purchase lower priced futures.

Overall, it is great that the topic of physical ETFs is starting to gain prominent attention at firms like Goldman. To be sure, just as in most other client-facing discourses (unlike open-ended ones such as those by the firm's economic team), the call is to be faded. We eagerly await a reevaluation which accounts not only for the contango implications on surplus markets, but takes into account the fundamental issue associated with the current reserve capital system, namely that monetarism no longer represents a one for one equivalent in pursuing price parity w/r/t futures price expectations. Hopefully soon someone, and certainly not Goldman, will redo the analysis to account for central bank behavior which is making all commodities, not just gold, be so mispriced that the traditional trader mindset is unable to properly capture pricing expectations along the futures pricing curve.

Full Goldman report link