In order to understand what the problems were with listing the Agency credit default swap futures contract on either the CBOT or CME, one has to appreciate that the two exchanges have been at each other’s throats for a hundred and fifty years. Neither exchange ever launched a contract without the other doing its best to cannibalize its rival’s volume. Traditionally the CME focused on short-dated commodities that were perishable, like eggs and butter, while the CBOT focused on commodities that could be stored. Incredibly enough, this thinking transferred to the financial futures markets, where the CBOT developed a niche in long-dated Ginnie Mae mortgage futures and the CME, on the other hand, focused on T-bill futures. Eventually the long- versus short-dated divergence translated into the modern set of products—the CBOT lists Treasury note and bond futures, while the CME has latched on to Libor futures with Eurodollars. As it turned out, the future of the Treasury market had in store for the CBOT sharp declines in the volume of debt that the U.S. Treasury would be issuing, and the long bond, which had been a staple of CBOT volume, was cancelled in 2001, only to be brought back five years later, in 2006, in a much smaller issuance program. Eurodollar futures, which can be used to create synthetic interest rate swap futures, turned out to enjoy much greater volumes than any of the CBOT products, and for many years they have enjoyed the title of not only the most liquid futures contract in the world but also the most liquid derivative of any kind. At times, $100 billion trades have gone through the exchange without even turning the market from the bid to the offer side, which is saying quite a bit for liquidity, since there aren’t too many traders capable of executing this large an order.

In late 2006, the CME announced a plan to take over the CBOT, and the deal was consummated in 2007, so that now there is only one dominant U.S. futures exchange, which is also the largest in the world along any dimension—volume, open interest, or margin deposits. The six-month battle to take over the CBOT was a bruising affair, during which time the CME had to raise its bid twice. As that campaign was waged, there was a strong market trend that the exchanges noticed, but had few unused resources to take advantage of: credit derivatives were becoming the fastest growing and most profitable market on Wall Street. Throughout their long history the exchanges have played into the prejudice of New York traders against the futures business, and in all likelihood lived up to the New Yorkers’ low expectations, because what the exchanges should have been doing for the past hundred years wasn’t fighting each other but rather cultivating relationships on Wall Street, so that they could benefit from the emergence of new ideas like credit derivatives. Rather than nurturing the development of new markets, both exchanges were often left to try to break into rapidly maturing products.

Dating back to the times when the exchanges were privately owned by their members, the boards of directors were into everything at their exchanges, and no one would have been surprised if the board chose to have a say about what kind of toilet paper would be stocked in the bathrooms. Floor traders elected the board members, and at the CBOT the grain traders in particular controlled an inordinate amount of voting interest, because it was grain trading that first brought the exchange to prominence. In order to launch a new contract, the board was required to vote on it and approve it. In the middle of a merger, how could the boards of either the CME or the CBOT be expected to vote to approve any new contract, no matter how potentially significant it might be? The simple truth is that these two exchanges were gridlocked while the merger discussions continued for almost a year.

Luckily, there was a third alternative to the CBOT and CME in the form of a start-up exchange that was backed by a different set of players in the market. Prior to the CBOT/CME merger, Eurex was the largest exchange in the world, and at one time it partnered with the CBOT to enter the U.S. market. Of course, the partnership broke down, and the former collaborators went about trying to cannibalize each other’s markets. What ended up happening was that Eurex listed rival contracts to the CBOT’s Treasury futures for a cut-rate price. The CBOT dropped its prices (temporarily) to drive Eurex US out of business. The same thing happened to the CBOT products in Europe. After several years of eating away at the other’s profits for no particular reason, both sides called a de facto truce. What was left in the U.S. was a Eurex-owned exchange that had all of the regulatory clearances to trade futures in the U.S., but no products to trade. Eventually a “supermajority” stake of 70 percent of the exchange was sold to Man Group, a publicly traded hedge fund.

Unencumbered by a merger and with a fresh mandate from its hedge-fund ownership to be aggressive, Eurex US would be transformed into the U.S. Futures Exchange (USFE). An exchange with no products might seem like a strange choice to list Agency credit derivative futures, but there is more muscle behind this operation than it may seem at first. For one thing, USFE clears through what used to be the Chicago Board of Trade Clearing Corporation (BOTCC). In one of those inexplicable Chicago futures exchange management decisions, BOTCC decided to side with Eurex rather than the CBOT after the collaboration between those two exchanges ended. As a result, after more than a hundred years as a partner with the CBOT, the clearinghouse was dropped for CBOT trades, and the CBOT decided to clear its business with CME Clearing. Unlike at the CME, there has been physical settlement with contracts cleared at BOTCC for thirty years—although to be fair, clearing with the CME has been flawless since it took on the CBOT contracts. At any rate, USFE was able to line up substantial partners in the form of a large hedge fund and a large Chicago institution with BOTCC. The reality is that as long as everyone has the electronic connections to trade products on a certain exchange, it doesn’t really matter what contracts are listed on what exchange. Except for margin requirements, end users of the contracts should be indifferent to trading on USFE compared to Eurex or CME; the choice is little more than the click of a mouse to an end user.

Even after we had found a home for our contract, there were discussions about what the ultimate design should be, although by then it was clear that the concept of an Agency credit default swap index was viable and it would not change. One of the difficult aspects of dealing with the Chicago exchanges is their reliance on tradition. The CBOT and CME have had quite a bit of success with two distinct forward structures: cheapest-to-deliver note futures at the CBOT, and discount-style pricing of Eurodollars at the CME. Frankly, it’s difficult to get either of the exchanges to launch products that differ from their traditional offerings. The CBOT always wants to make every instrument a forward contract so that there are no intermediate cash flows and the accounting is simpler. CME designers enjoy finding new ways to apply the same discount-style pricing that works so well with Eurodollars. The problem with credit derivatives is that they don’t shoehorn very well into either of the favorite styles of the exchanges. Nothing about credit derivatives has to do with a forward rate; that is solely the domain of the interest rate markets. As the explosive growth of the default swap market shows, the complications of forward pricing are completely unnecessary. The very low fixed coupons also don’t mesh well with the discount-style pricing of the CME. Moreover, the conditional nature of default swaps throws everybody for a loop.

Default swaps are actually options on a credit event, we can see that the cash flows could be quite complicated if there were a default during the life of the contract. Trying to avoid cash flows with forward contracts works with instruments that aren’t conditional in nature. The whole point of default swaps, however, is to capture deteriorating credit perceptions. If there is a default, the buyers of protection need to receive some lump sum, presumably to make up for the loss on their note holdings (if they happen to be hedgers and own a note from the issuer). This “square peg in a round hole” problem seems insurmountable if one approaches the problem from the perspective of fitting the traditional Chicago exchange styles into the new product, but we need to be careful about the assumptions and prejudices we bring to the table in the first place. Beyond meaningless tradition, why is it necessary to make default swaps look anything like the existing futures products? There is no reason, and in fact the easiest way to get default swaps onto an exchange is to offer all of the benefits of a futures contract—centralized trading, pricing, and clearing—without any of the drawbacks. Matching the existing over-the-counter default swaps in a futures contract brings all of the benefits of exchange-traded products and none of the headaches of having to convince traders of the merits of a new design.

The CME proposed a design for a single-name default swap that works identically to an option, and although the economics match the over-the-counter market, the specification is quite different. Instead of paying for the default swap premium over time, in a type of very small annual yield, the very first CME default swap was priced as the sum total of the entire cost of the option. For example, a default swap that is priced at a premium of 20 basis points per year would cost 100 basis points over five years. In the over-the-counter market the quotation for this swap would be 20, while in the proposed CME contract the quote would be 100. Why the change? It happens that an option specified this way is easier for the exchange to program into its systems, since the mechanics match that of a typical option. However, there has never been a mechanism that allows an exchange to automatically deduct a fixed amount from any futures contract, and until now the exchanges have relied on the “simple” operation of the market to price all of their contracts. For example, what would happen to an option worth 100 basis points in one year? In the over-the-counter market there would be an automatic and fixed quarterly payment of 5 basis points, for a total cost in the first year of 20 basis points—nothing complicated about that. The problem is, no futures contract in the world has ever worked like that, until now.

Futures options pricers might very well have reached the same conclusion, that the annual cost of the default swap was 20 basis points and the new remaining premium was 80 basis points (still quoted as 20 basis points per year in the over-the-counter market), but futures exchanges would have relied on the market to make the 20-basis-point deduction. Rather than create a formal mechanism to drop the option price by 20 basis points, the exchanges have relied on arbitrage to set prices. If by some accident the price settled at 80 basis points on the appropriate day, that would be fine with the exchanges, but until now they have never said anything about what the prices of their products should be. The distinction is a subtle one, but if there had never been a mechanism for the transfer of a fixed coupon, if the market price of the option in question turned out to be 70 basis point after a year, the exchanges would simply have said, “the market is the market,” and an inappropriate amount of premium would have changed hands. In the over-the-counter market, the operation of a fixed coupon and a price that changes depending on the prevailing market value is as simple as trading a coupon bond—save for the fact that the dealer is making the coupon calculation and transfer rather than an independent third party like a calculation agent.

While the USFE wasn’t shackled by tradition, it turned out to be not entirely agnostic about what the best design of the contract should be. The very first contract listed by USFE was a binary event option. The name is complicated, but the idea is very simple. The payout of this type of option is fixed, and the price of the option represents the probability that it will go in the money. Suppose the price of a binary event option (called a “binary event future” or BEF by the USFE) was $750, the payoff would be $1,000, and the profit would be $250 if the event occurred. The first BEF contracts were on the merger of the CBOT and the CME. If there were a merger the contract holders would make $250, but they stood to lose $750 on a wrong-way bet. This contract structure has many of the features necessary to list a credit derivative. For one thing, the payoff is contingent, which is probably the most important aspect of a default swap. The recovery rate could surely be set by physical delivery of the underlying notes or reference to an International Swaps and Derivatives Association (ISDA) recovery rate. One glaring problem is that the structure doesn’t allow for a coupon of any sort. The conditional nature is there, but nothing references a fixed coupon that represents the “pay-as-you-go” nature of credit options.

Although we have examined the details of pricing default swaps at length in previous chapters, they are quite ill-labeled as “swaps”—in fact, they resemble a traditional interest rate option, but there are differences in the way the premium is paid. When one buys or sells an interest rate option, the total premium is paid or received up front. While time decay surely eats away at this premium over the life of the option, effectively transferring time value from the buyer to the seller of the option, there is nothing in the contract specifications that require this to be so—it is merely an aspect of the economics of an option, one that can be described using the Black-Scholes formula to formalize the decay in value of the instrument. If the premium for an option is $100, nothing in the contract specification requires that option to be worth $90 in a week and then $10 just prior to expiration if the option remains out of the money. It is merely the market’s perception of the value of the instrument that determines its value.

On the other hand, credit derivatives have adopted a “pay-as-you-go” style that stretches the option’s premium payments out in equal amounts over the life of the instrument. It is a misnomer to call this payment a coupon, but it does share many of the same functions as a coupon in a traditional bond. For example, the payment on a default swap is fixed at the birth of the instrument and, at least when included as part of an index, the price fluctuates to match the present value of the instrument with a new par contract. Just as with a Treasury note, the price changes so that the yield of the instrument represents prevailing rates. In pay-as-you-go (PAYG) structures, the premium for the option is fixed when the swap is created, but the values of the fixed payments change. The difference is a subtle one, but it is worth making a distinction that there is no return of principal with a default swap, as there would be when a coupon bond matured. How might one represent this same structure in a binary event option? The payoff from a binary event option is fixed, so how could it represent either the PAYG structure or the appropriate recovery rate from a typical default swap?

One way to get around the problem of the PAYG structure with a binary option is to set the initial price above the maximum payout of the option. For example, if the option would either pay off $1,000 or $0, then the initial price might be set at $1,100, which could represent a $20 fixed payment for five years. In fact, the $100 in premium could show up anywhere, depending on how the contract was written. Rather than add it to the maximum payout, one could simply list the initial price of the option at $100. If there were no credit event, it would expire worthless at $0, and if there were a credit event, it would expire at $1,000. In solving the PAYG problem by allowing a fixed recovery rate, we have introduced another problem. The very essence of a default swap contract is that the severity of the loss is unknown when the contract is initiated, and fixing the payout in a credit event creates quite a different dynamic. How flexible is the binary event futures design? The contract writers can do anything they want, but the question remains: What is the market’s appetite to trade what they come up with? Rather than fix the payout rate on the contract at either $0 or $100 in a credit event, why not require physical delivery of the note, thereby introducing an unknown recovery rate and matching the economics of the default swap market perfectly?

For one thing, we are miles past leaving the convenient result that the price of a binary option is the probability that it will go in the money. In fact, we lost that result as soon as we integrated the PAYG feature into the structure. In theory, we have all of the pieces we need to match the economics of the over-the-counter default swap. In fact, these same discussions went on with USFE on how exactly to structure the Agency credit derivative index, but ultimately we came to a single conclusion. We had already decided on Agency credit derivatives because they are the only instrument that is allowed to be traded on an exchange in the exact format as the over-the-counter market, so why diverge from that path? If we could also match the over-the-counter market with the new futures contract, then that’s exactly what we should do. Rather than getting bogged down in explaining why some complex and brilliant alternative to the existing structure should be traded, we could spend our time getting the word out on the new contract. In the evolution of the product, it seems as if this was simply one step too far. Although it would have been convenient for the USFE to list this product immediately, since they already had binary event futures trading on their systems, it seemed worth the wait to create a futures contract that would exactly match the over-the-counter market. Having settled on the design of the contract, the USFE needed to invent a way to handle the cash flows of such an instrument, because it had never been tried before.

The futures market evolved in response to a need from agricultural hedgers who were eager to eliminate the uncertainty of price fluctuations between the time they sowed a crop and the time it was harvested. In another example of the inertia that is traditionally a characteristic of the markets, the financial futures contracts that were first traded at the CBOT and CME adopted this same forward design. There is nothing inherently necessary about creating a forward contract in order to hedge financial risk. In fact, there are few products in the cash markets that trade on a forward basis. The only reason futures have been so successful is that either they can be adapted to replicate the economics of a spot instrument (as a Eurodollar strip replicated the risk of a spot interest rate swap) or, as in the case of Treasury futures, the market itself can adapt to the relatively small risk of hedging notes in the cash markets with forward instruments. There happens to be a significant advantage to forward contracts: there are no cash flows to deal with, an advantage over coupon or pay-as-you-go option premiums. For example, a Treasury futures contract represents the price of a note that is financed in term repo for a number of weeks or months, but when it is time for that term to be over, the forward contract would turn into a cash note whereas the futures contract expires. Another example of the simplicity of eliminating coupon payments can be found in Eurodollar futures, which are contracts based on forward Libor covering the three-month period beginning with the expiration of the contract. Eurodollar contracts expire when the three-month period begins, so there is never a transfer of cash that has to do with an accrual, or anything resembling a coupon for that matter.

The reality is that daily transfers having to do with time value happen all the time in futures contracts. In fact, fair value for every financial contract has to do with ticking away the days of the year in a somewhat mechanical fashion. The difference between seeing a forward price roll down to the spot price is that there is nothing in the contract specifications in a forward contract that dictates the transfer of money on a daily basis. Memorializing the transfer of money from a coupon or other form of time decay is what makes the USFE approach so unique, and its name for its instrument is the “spot-equivalent futures contract,” which is a somewhat cumbersome but accurate description of the mechanics. With the contract’s set of legal documents and the appropriate code in the computer systems, the USFE is able to replicate the transfer of cash from one side of the trade to the other in a way that mimics a coupon or PAYG option premium. The reality is that this should be quite a simple project, since the USFE merely has to take a fixed coupon, apply it to some notional value, and then move the money from the seller of the coupon to the buyer. Like all good ideas, this one smacks of the “why didn’t anyone bother to work this out before?” Married to their agricultural origins, the CBOT and CME haven’t made any efforts to deviate from listing forward contracts. While futures contracts have all been forward contracts, the introduction of the Agency credit derivative index emphasizes the fact that anything traded on an exchange is an exchange-traded derivative, and it need not be a forward contract to be a “futures” contract.

There was one inspired design alternative put forward by the CBOT to avoid the problem of referencing an SEC-registered security in the settlement process, which is the idea of a “maximum running spread.” The concept of capping the loss of a default swap isn’t new, but the maximum running spread calculation allows for a flexible default price without fixing something like a 40 percent recovery rate, as the CME chose to do with its contract. The mechanics of the contract are relatively simple: fifty of the most liquid default swap contracts settle to an index of yield spreads published by a reputable data provider. If there is a default, near default, or general financial stress on a firm, then we can presume that its premium will increase and the price for that portion of the index will drop. The maximum running spread, however, is a cap on how high the premium, and how low the price, can go. The contract doesn’t mandate that an issuer in default use the maximum recovery rate; this calculation is simply a cap on yields and a floor on prices. Allowing the value of the contract to fluctuate as long as it is within the cap allows the futures contract to match the economics of the over-the-counter default swap without creating any basis risk between the two, so long as we believe that the data provider used to settle the contracts is doing a reasonable job. The fact that the initial contract focuses on the most liquid fifty names in the market probably makes the job of observing their prices in the over-the-counter market quite simple. Unfortunately, there are some drawbacks to this design choice.

Primary among the drawbacks to the CBOT contract is that an issuer in severe distress, whose recovery rate will clearly be far below the one envisioned by the maximum running spread, will diverge from the CBOT contract. For example, the maximum running spread for the September 2007 Liquid 50 futures contract at the CBOT is calculated in Table 9.1.

The calculations take some explaining, because for one thing, an “index DV01” is pulled out of thin air without any explanation. Credit traders may recognize this value as the credit DV01 of a 5-year default swap, but it may not be as obvious to the new user. Also part of the interesting twist of this contract is the fact that it uses a fixed recovery rate of 40 percent in the calculation of what will eventually be the fixed recovery rate, i.e., the maximum running spread. The whole point of a maximum running spread is to put a floor on how low the price of any individual contract can trade (or conversely, a cap on how high the premium can trade), but this calculated value doesn’t have anything directly to do with what the CBOT calls the “standard recovery rate,” which is 40 percent in this example. Use the recovery rate to get the recovery rate? In this case, what the CBOT calls the “standard recovery rate” is actually just a plug used to calibrate the maximum running spread.

The idea of a maximum running spread is an ingenious way to get around the fact that the CBOT can’t reference corporate-security prices in its settlement algorithm. In essence what happens with this contract is that there is a floor on how low the price of an issue in distress can trade, but there is no requirement that an issue in default hit this low price. The contract settles to an index-pricing service that serves as a sort of neutral third party resource for credit default swap prices. If the index begins life with an average spread of 50 basis points and the index service publishes a set of premiums so that the ending spread is also 50 basis points, there will be no profit or loss from the contract save for the fact that it is a forward instrument. However, suppose that the index begins life with spreads comfortably around 50 basis points, but then something goes wrong with one of the companies in the index. Ever day it’s apparent that the spread of this distressed company is moving higher. How high can that particular name trade? Put another way, at how low a price can that security trade? It’s all determined by the maximum running spread. Rather than simply fixing a recovery rate to be applied to all contracts, the CBOT essentially comes up with this value through a formula. This formula can be changed for each expiration month, but once it is fixed for a particular month, it can’t ever change during the life of the contract. If there is a complete catastrophe and the company goes bankrupt with a zero recovery rate, then the maximum running spread puts a floor on how low that default swap can trade in the index. If the default swap premium hits the maximum running spread, or the pricing service fails to publish a price for that name, then the 1,324.50 basis point spread is used for the index calculation for the remaining life of the contract. If by some miracle the company pulls itself back together and its default swap spreads fall, then it is those new spreads that are published by the pricing service and that are used in the calculation of the index. The important thing to realize is that with the CBOT contract, no name is ever dropped from the index because of a credit event: it simply gets priced with the maximum running spread. There is no reference to ISDA language regarding defaults or what constitutes a credit event, and the contract is really quite elegant in how it handles pricing and settlement.

Of course, every yield translates into a price, and a maximum running spread of 1,324.50 in our example above translates into a dollar price for the contract of around $64 per $100 face. This is the beginning of the problems for the contract, since it’s at this point that the CBOT contract will begin to diverge from the over-the-counter market and create a basis that few traders are interested in dealing with. If there is a calamity, and it becomes clear that the recovery rate is going to be quite a bit lower than what it will be allowed with the maximum running spread, which is somewhere around $64 per $100, then the CBOT has essentially lopped off one of the tails from the loss distribution. There have been plenty of instances when the recovery rate is below 60-ish percent, and in fact the standard recovery rate used in modeling is 40 percent. While the design of the CBOT is quite elegant, again it errs in favor of the sellers of protection, since it limits their downside loss should there be problems with a particular credit. What if it becomes clear that the recovery rate is just twenty cents on the dollar? An implied floor of 1300 basis points forces the futures contract to be much more valuable than the over-the-counter contract. Why artificially inflate the value of default swap simply because it solves a regulatory problem for a futures exchange?

The irony is that the maximum running spread calculation is flexible enough to accommodate any inputs, so why not put the floor for prices lower than the mid 60s? If the CBOT had wanted to more faithfully represent the modeling assumptions of the over-the-counter market, they could have matched a recovery rate of $40 per $100 quite simply, by making the maximum running spread:

This line of criticism may well damn the CBOT contract. Another problem is the fact that the Liquid 50 index is a little too narrow to replicate the CDX index, the most popular and widely followed credit index, while at the same time it is too broad to be considered an index subcategory, like the “high grade” or “high volatility” components of the CDX. The potential for a substantial divergence between the futures and over-the-counter contracts in instances of severe distress, as well as the somewhat awkward positioning of the fifty names in the index, cast doubts on the CBOT’s contract. There is a final issue to this contract, in that it trades on a yield basis, rather than a price basis, and the seller of protection is the buyer of this contract, rather than the seller, which is different than in the typical price/ yield situation. The profit and loss of the contract are literally paid by the change in basis points, which means that to profit from deteriorating credit conditions, one has to get long the contract. However, whatever the foibles, the idea of a maximum running spread is a clever one, and it may show up again in a slightly different format, with a slightly different calibration, down the road.

As was our original contention, a successful product always feels as if it was inevitable, and Agency credit derivative futures are no exception. However, even our brief history of the design alternatives for the contracts highlights just how much work and attention went into their specific design. Originally conceived as an entirely new approach to credit derivative structuring relying on cash settlement to published market rates, the contract ended up matching what already trades in the over-the-counter market. In between, we considered the possibilities for replicating the same economics with different structures, such as binary event futures. Alternatively, we also considered adding another underling risk that had never been formalized before, namely the handling of credit risk for the joint and several liabilities of the Federal Home Loan Bank system. While these design alternatives may hint at the path of future development for Agency credit derivatives, only the simplest elements seemed appropriate for the introduction of the product. Even if the format for the contracts remains the same, there is still quite a bit of innovation that might be accomplished by combining already-existing pieces into something that has never been offered before on a futures exchange.

One way to combine existing pieces is to list two versions of the same contract to mirror the “funded” and “unfunded” contracts that trade over the counter. As with all derivatives, there is no cash exchanged to initiate a position, except for margin, which usually runs around 1 percent of the notional value in financial futures. What happens when investors actually want to put dollars to work the same way that they do when they buy notes in the cash market, and not simply replicate the economics of a long position? In that case, the over-the-counter contracts bundle an interest rate swap with the credit derivative in order to expose the position to interest rate risk along with the credit risk inherent in the credit derivative. This approach takes advantage of the market-making commitments of Citibank and Goldman Sachs in the interest rate swap contract at the CBOT, since the spread between the funded and unfunded contracts is simply an interest rate swap. Suppose that someone put a bid into the funded contracts, and there was an offer that matched the bid in the unfunded market. Since the spread between the two contracts nearly matches the interest rate swap contract at the CBOT (save for the fact that the CBOT contract is a forward instrument and the contracts at the USFE are all spot instruments not representing forward values), savvy arbitrageurs could hit the bid in the funded contract and lift the offer in the unfunded contract, effectively crossing the two positions. What the arbitrageurs are left with is the risk of an interest rate swap, which in this case is equivalent to paying fixed on an interest rate swap (getting short the swap rate), since they hit the bid on the funded contract, which is equivalent to selling the swap rate. The two credit-derivative legs of the trade wash, and all that is necessary to eliminate the residual risk of having sold short the interest rate swap is to buy the CBOT interest rate swap futures contract in the right proportion. Conveniently the arbitrageurs can enter the trade knowing exactly what kind of liquidity they will see in the interest rate swap futures contract, since Goldman Sachs and Citibank have already committed to the size and price spread they are willing to make in the CBOT swap futures contract. As an aside, the ironic part of this structure is that end users are often happy to accept relatively wide spreads in the interest rate swap component of the funded trades in the over-the-counter market, and the quite narrow price spread of the CBOT market makers was intended to compete with the naked interest rate swap market, rather than to facilitate the combining of these securities into a package. Whether or not there are official market makers in the spread between the two contracts, it’s obvious just how someone could take on the job, by making slightly wider markets in the interest rate swap leg of the funded spread than he or she is seeing in the CBOT contract, which would be used as a risk offset.

Whether traders in credit derivatives choose a funded or unfunded contract, they still have not used any cash beyond margin to initiate the position. The funded contract represents the economics of a long-note position in the cash market, save for the very important use of cash and what can be earned on that money market investment. In order to use the cash, dealers often sell guaranteed investment contracts or “GICs,” which specify that the buyer will hand money over to a dealer to earn a guaranteed return, and the dealer will accept the risk of earning the required return, often quoted as a spread to Libor. A synthetic note in the over-the-counter market includes selling protection on a credit default swap, receiving the fixed-rate side of an interest rate swap, and then buying a GIC. In the case of synthetic Agency notes, the pricing might be something like 7 basis points to sell protection on an Agency, while of course an interest rate swap would earn the Libor flat, and the GIC might cost something like Libor minus 2 basis points, for a total yield compared to Libor at a positive 2 basis points.

There is one important distinction between the funded versions of the Agency credit derivative in the over-the-counter and futures markets: one is actually what it says it is, and one is a fraud! Or, if not a fraud, the futures contract only appears to be a copy of the over-the-counter contract, because there is no accounting for the cash in the futures position as there is with the GIC in the synthetic over-the-counter note. Rather than asking users to invest the money in a GIC, a futures contract requires them to post margin, but the balance of the position is left unaccounted for. Margin accounts typically earn the fed funds rate plus some spread in basis points. Since the margin for futures changes every day, the benchmark is an overnight rate, rather than Libor, but often the effect of earning fed funds plus some spread is roughly equivalent to Libor. It wouldn’t be a surprise if the reason the GICs offer Libor less 2 basis points is that they’re pulling off of the same managed money market account that futures margin is deposited into. The point is that the return on the cash portion of the investment is normally comparable for a GIC and a futures margin account. Aside from the required margin for a futures position, the rest of the cash is free to invest as the trader wishes, and in fact, it would be appropriate to buy a GIC if money needed to be put to work.

The advantage of a futures contract that includes an interest rate swap is that it mimics all of the mechanics of an Agency note, save for the investment of the cash. If there is a change in Treasury or swap rates, the yield of the interest rate swap will change. Of course, if there is a change in the credit condition of the Agencies, the default swap will change in price. Still, it’s worth noting that the DV01 of the synthetic note is not exactly the same as the cash issue, as illustrated by Table 9.2.

In order to judge the interest rate risk of the synthetic Agency note, we have to add the DV01 of the interest rate swap to the interest rate DV01 of the credit default swap, for a total of $432.22. This value is slightly larger than the DV01 of the cash note, by just under 3 percent. A grand total of $11.26 per million face doesn’t sound like much, but what about a $1 billion portfolio experiencing a 25 basis point move, which is a relatively small readjustment in rates? All of a sudden this 3 percent difference has a quarter-million-dollar value attached to it. In the scheme of things, this dollar value isn’t a major consideration, but it is slippage that could be easily avoided by careful measurement.

Good ideas seem inevitable, but formalizing an idea depends on making design tradeoffs. In the case of credit derivative design, the market is growing so quickly and is so profitable that every Wall Street dealer and all of the largest futures exchanges in the world are either heavily invested in or trying to be heavily invested in the market. Why spend time wondering about evolutionary dead-ends in derivative design? Understanding design elements that were not chosen can help to deepen one’s understanding of the parts that remain. In the case of Agency credit derivative futures, there were substantial hurdles to overcome from both the regulatory and design standpoints. All of the design decisions in credit derivatives have been made in a relatively benign environment, with robust economic expansion and few corporate defaults. Remembering that the value of all derivatives stems from settlement, and that there have been few defaults, one can argue that the value of the entire credit derivative market is not well understood. How does a default swap settled via an ISDA auction differ in valuation from one settled by physical delivery? How much cheaper, if at all, should a synthetic corporate bond be relative to the cash note? As of this writing, no one on Wall Street is prepared to answer these questions, but they will be on the mind of every derivative trader at some point in the future, when the economy enters a recession and the market faces defaults on a large scale. The first step to answering these questions is to understand what design choices were made, and what the alternatives could have been.