Exchange of Futures for Physical (EFP) Explained - Part One
NOTE: I’ve been having some problems accessing the SILVERAXIS blog but hopefully it should be working now and comments should be enabled on this latest post. I’ll be providing some market observations shortly as well. This post was made available to Metal Augmentor subscribers on July 16, 2009 at 6:47 p.m. EDT.
Adrian Douglas, a Director of GATA, wrote a piece recently called The Alchemist in which he pointed out that the “Exchange of Futures for Physical” (”EFP”) mechanism of the gold commodity market allows ETF shares such as IAU and GLD in the definition of “Physical Products”. While this is true, we shall see by the end of my rebuttal that it is also irrelevant to the physical gold market. Furthermore, Mr. Douglas alleges that the regulators are rigging the market in favor of the manipulative shorts since ETFs are “paper” and not the same as physical gold — indeed ETFs may not hold any bullion at all! Alas, I’ve addressed the “empty ETF” issue many times before so I will not spend much more time on it, but I will demonstrate in detail why it doesn’t really matter if ETFs are really “paper” or “Physical Product” when it comes to EFP transactions. In fact, it doesn’t really matter to market participants what is exchanged on the physical side of an EFP, be it even toilet paper, as long as EFPs remain only a minor component of overall trading volume on the exchange.
Unfortunately, Mr. Douglas has done a disservice to the gold community by not explaining how the EFP works and thus I shall try to correct this first. Doing so, however, will take a bit of effort and so I have dedicated the first part of this rebuttal to providing an overview of the EFP transaction. The second part will consist of looking at Mr. Douglas’ allegations and refuting them point-by-point.
Every Exchange of Futures for Physical (or Product) involves two parties that wish to swap Futures and Physical positions at the same time. In all cases, one party already has Physical exposure to a Product that it wishes to sell and/or to convert to a price exposure via Futures. The reasons for this are myriad and I’ll have specific examples in a moment, but first it is important to note that the two parties in an EFP transaction will always first come to a private agreement off-exchange (also called ex-pit) on the number of positions, timing, price, and form of the Product being exchanged. The parties may use a reference price for the EFP transaction that is based on Futures trading on the commodity exchange or they may use whatever reference price they wish, as long as both agree. Once an agreement is reached, the parties’ brokers are informed of the number and price of the Futures positions being exchanged as part of the EFP and this information is then transmitted to the exchange. Importantly, price data is NOT transmitted as a trade, only quantity, and therefore EFP transactions are not part of the futures market pricing mechanism.
The following illustrates how an EFP in COMEX gold basically works. There are other ways to do an EFP, including ways to close out Futures positions, but they are all variations of this basic transaction.
You’ll note in the above graphic that the Futures position isn’t simply moved over from Trader B to Trader A, but rather there is actually 10 new long Futures entered into the account of Trader A and 10 new short Futures entered into the account of Trader B. Since Trader B is already long 10 Futures, the 10 new short Futures are essentially an offset. The net effect of the EFP transaction is to move the Physical gold from Trader A to Trader B and the 10 long gold Futures from Trader B to Trader A. That is essentially why it is called an exchange.
But, why use an EFP in the first place? Well, it turns out there are several reasons why two parties would wish to engage in an EFP instead of separate transactions involving a purchase or sale of Physical Product and the purchase or sale of Futures contracts. The key consideration, however, is usually transaction efficiency. Without an EFP, the sale or purchase of Physical Product and the separate entry of a Futures order may result in “price slippage” due to market volatility or fluctuating trading volume in the spot and/or Futures markets. EFP transactions are often for hundreds if not thousands of contracts and typically there is not sufficient depth in the bids and asks to fill such a large order at the same price.
The risk of adverse price movements can be substantial, and given that commodity transactions often seek to take advantage of profits of 1% or less, price certainty before a transaction commences is crucial. An EFP can avoid price slippage since it may be transacted outside the regular trading session. More importantly, the EFP “price” itself is quoted in terms of “basis”, being the difference between the Futures price and the spot price for the Physical Product. In effect, you know what you are getting with an EFP transaction ahead of placing the order. Another way to think about it is that an EFP is essentially a spread order but instead of the legs consisting of two different Futures, one leg is a Futures and the other leg is a transaction for Physical Product in the cash market.
Perhaps the best way to answer the “why” question is to provide an example of a possible EFP transaction in the gold market. We’ll refer to the above illustration but replace the abstraction with a situation that could actually occur in the real world. Indeed, this particular type of transaction does apparently occur in crude oil trading, which is the only description of an EFP transaction from a credible source that I was able to find online. Most of the following example has been adapted directly from the document Exchange Futures for Physical (EFPs) for ICE WTI Crude Futures. This document has additional EFP nuances that I will not mention here and so the serious student of the commodity markets may want to study it as well.
Trader A: The trading arm of a metal refining company, which has just finished a production run of 1,000 ounces of COMEX-approved gold bars refined from concentrates that were purchased from mining companies. Note that Trader A is long 1,000 ounces of gold once the concentrates have been refined to bullion. [I could complicate this example by having Trader A hedge the long Physical position, but let's not go there in this simple example.] Let’s just say that Trader A naturally wishes to sell the recently-refined 1,000 ounces of gold since he is in the refining business, not the gold hoarding business. Trader A has a policy, however, of selling in installments throughout the year with the majority of gold sales being made in months with favorable seasonals for gold prices. July and August are not favorable seasonally but September and October are. Moreover, Trader A wishes to receive the proceeds from selling the 1,000 ounces of gold soon so that he can finance the purchase of additional concentrates from mining companies.
Trader B: A gold fabricator needs 1,000 ounces of gold by late September in order to manufacture 50,000 linear feet of 18K gold chain to supply her jewelry manufacturing clients in time for the holidays. Trader B, like Trader A, expects the price of gold to rise between now and the end of September, and she therefore wishes to “lock in” the gold price. She has provisionally done so by going long 10 contracts of October 2009 COMEX gold Futures, but she will not be able to use such Futures contracts for her supply of Physical gold since it would be too late to take delivery on the October 2009 Futures. Trader B would like to have the gold available sooner than later so that there is no need to scramble for supply at the last moment. Late July would be an ideal timeframe to switch from long gold Futures to long Physical gold.
Note that both participants are long in a market where they expect the price to rise. Trader A, however, has not secured a buyer for his gold whereas Trader B wants to be able to secure supply of the quality and delivery timing she needs.
It turns out that Trader A and Trader B have done business together before. And so they agree to exchange their respective positions in order to meet their respective needs — the seller (Trader A) wants to remain long the market as he thinks the price is going up. The buyer (Trader B) wants to secure a price and the quality and delivery timing she needs.
So how exactly does the Exchange of Futures for Physical work? Sometime in late July of 2009, Trader A agrees to sell to Trader B 1,000 ounces of gold for cash at the London afternoon fixing price for that day’s trading (let’s say it is $940/ounce). By agreement between the parties, the 1,000 ounces of gold will be delivered to Trader B within 3 days and Trader A will be paid $940,000. Concurrently, Trader B agrees to sell to Trader A ten (1o) October 2009 COMEX gold Futures contracts at the settlement price for that day’s trading*. Once an agreement has been reached, the two parties inform their respective brokers of the terms of the EFP. The brokers then contact each other and register with the Exchange that the EFP has been agreed and the price. After the EFP transaction is registered with the exchange and cleared, Trader B is able to use her freed-up margin equity as well as accumulated Futures profits from her account as partial payment for the Physical gold.
* The 10 Futures contracts aren’t exactly being transferred from Trader B to Trader A. Rather, what is happening is that Trader A is credited 10 new long Futures contracts against 10 new short Futures contracts for Trader B. The Futures contracts in Trader B’s account are then offset so that she ends up with a neutral position. The net effect is that Trader B’s long Futures position is exchanged for Physical Product consisting of gold bullion.
Because both participants believe the gold price is going to increase, the EFP has suited both their needs, enabling security of supply for Trader B without commitment to a price on behalf of Trader A.
Before we go on, it is important to reiterate that typically the biggest benefit to EFP participants is that the transaction is able to bypass the exchange or more specifically the trading pit (or its electronic equivalent). Being able to make large trades at a certain price, especially outside of regular market hours, can be advantageous for market makers, hedgers, price scalpers and scalliwags of various stripes. EFPs therefore encourage the use of Futures by parties that would otherwise not consider them and probably for that reason many exchanges allow them. At the same time, however, EFPs also deprive the market of trading volume. It is primarily for this latter reason that regulators have placed rules and limits on how EFP trades can be made.
Still, the advantages of EFP transactions can be such that historically many trades were not “bona fide”, meaning there was no physical leg to the transaction at all. No doubt a certain percentage of EFP transactions in every market, including COMEX gold, are nothing more than a sham seeking to take advantage of ex-pit transfers of positions between two parties, which would not be otherwise possible save for EFPs.
If Mr. Douglas is worried about the use of ETFs in Exchange of Futures for Physical transactions, I wonder how he would feel if he found out that many EFPs are backed by nothing at all, not even paper? Fortunately, the main damage that results from such transactions is that the exchange is deprived of minor amounts of trading volume. Yet if left unchecked, the price discovery process could eventually be impacted by runaway sham EFP trading. The ultimate progression of this would be a subversion of the exchange as various entities would set up their own “trading circles” and report their trades as EFPs.
Now that many of you hopefully have a rudimentary understanding of how EFPs work and why they are used (and abused), let’s now examine why somebody would wish to exchange ETF shares for Futures. In terms of how such a transaction would look, we can actually just substitute ETF shares for the 1,000 ounces of gold in the above graphic. But let’s consider a slightly different format to explain a new scenario using our Trader A and Trader B.
Trader A: Arbitrage trader who buys the ETF when the shares are trading at par or discount and sells when the shares are trading at premium. Those familiar with Metal Augmentor’s ETF Basis charts will know that fluctuations between discount, par and premium do exist. Since Trader A is an arbitrageur, he hedges the accumulated long ETF position in GLD using short gold Futures.
Trader B: Authorized Participant of the ETF who officially accumulates or distributes shares based on premium or discount to spot prices. The Authorized Participant can deliver Physical metal to the ETF in order to create new shares, or remove Physical metal from the ETF after redeeming existing shares.
In effect, Trader A is a subcontractor of Trader B, accumulating ETF shares until there is a sufficient quantity for Trader B to bother with. An additional incentive for Trader A other than getting a cut of the total arbitrage profits is that Trader B may actually fund a portion of the cash to buy the ETF shares. In other words, a portion of the $920,000 cash shown in the illustration above may have been advanced by Trader B to Trader A in the form of a low-interest loan so that Trader A will have the funds to start accumulating ETF shares.
The above examples are two of literally dozens that can be used to transact EFPs on a bona fide basis. My examples actually have a valid business purpose but it is entirely possible and permissible to also use EFPs for purely speculative purposes. All the exchange cares about is that the physical component of the transaction is real and not a sham and of course also that position limits and other exchange rules are not circumvented by the EFP.
Based on the foregoing, it should be easy to see how Adrian Douglas’ concerns about the EFP mechanism are misplaced if not entirely unfounded. Nevertheless, I will address his points in Part Two of this rebuttal, to be available in a few days.