The TED Spread


Publication: Derivatives Quarterly

PDF Version

With the availability of alternative futures contracts based on related underlying instruments, it is often attractive to consider trading futures spreads, whereby similar futures are simultaneously bought and sold. One of the most common of these types of trades is the TED spread constructed by trading Treasury bill futures against Eurodollar futures.


In this case, both contracts relate to short-term (i.e., three-month) dollar-denominated interest rates. The TED spread trade thus reflects a view that these two underlying three-month rates will move differentially.


Over time, the concept of the TED spread has been extended, allowing for analogous transactions that bear on interest rate differentials pertaining to a range of maturities that extends beyond the three-month point on the underlying yield curves. This evolution has come to be known as “term TED spreads.” This article describes these various trades – their rationale, trade construction, and associated implementation issues.


TREASURY BILL AND EURODOLLAR FUTURES CONTRACTS

The Treasury bill futures contract is a price-fixing mechanism that locks in a rate on a three month U.S. Treasury bill with a deferred settlement date. Currently, expiration and delivery dates follow from the cycle of Treasury auctions. That is, each open Treasury bill futures contract is settled by physical delivery of $1 million par of U.S. Treasury bills. Deliverable bills must mature ninety-one days from the first of three allowable delivery days.


The Chicago Mercantile Exchange (CME) schedules quarterly expirations to maximize the supply of deliverable Treasury bills, so that the qualifying population includes old one-year bills and old six-month bills, each having three months of remaining life upon delivery, as well as new-issue ninety-one-day bills.


For example, the March 1997 Treasury bill futures contract expires (stops trading) on Wednesday, March 26, 1997. Bills that qualify for delivery must have ninety-one days of remaining life, counting from the next business day. So in this case, the required maturity date is June 26. Three different issues – an original one-year bill issued on June 27, 1996, an original six-month bill issued on December 26, 1996, and a new ninety-one-day bill issued on March 27, 1997 – satisfy the delivery requirements, as each has the requisite ninety-one days remaining until maturity.


The Eurodollar futures contract is similar to the Treasury bill futures contract. It too is traded on a quarterly cycle, although expiration days do not correspond to those of the Treasury bill contracts. For Eurodollars, expirations always fall two London business days prior to the third Wednesday of the month. This contract is said to be “cash-settled,” meaning that no physical delivery occurs. Instead, one last mark to market is made, where the final settlement price is assigned based on cash market yields, specifically reflecting the London Interbank Offered Rate (LIBOR) for three-month Eurodollar deposits, as quoted by the British Bankers Association.1 Each contract covers a national exposure of $1 million.2


Prices of both Eurodollar futures and Treasury bill futures are quoted on the basis of an International Monetary Market (IMM) index, where the associated rate reflected by the price index is found simply by subtracting that price from 100.3 For example, a price of 95.10 reflects an interest rate of 4.90% – a discount rate for the Treasury bill futures but an add-on money market yield for the Eurodollars.


THE BASIC TED SPREAD

Although three-month Treasury bill rates and three-month Eurodollar deposit interest rates generally move together – rising in times of monetary tightness and business cycle expansion and declining with monetary ease and cyclical weakness – the co-movement typically is not exactly equal. The TED spread, which reflects the difference between these two interest rates, may offer some attractive trading opportunities to those who can correctly anticipate such differential movement between the two rates.


Because Eurodollar deposits are direct obligations of private, offshore commercial banks, independent of the Federal Reserve System, investors believe them to be more risky and more likely to pay a higher interest rate than Treasury bills with a common maturity. As a consequence, the IMM price index for Eurodollar futures has been – and is likely to remain – lower than the price index for Treasury bill futures. At the root of any TED spread trade is the question of whether the present differential reflected in current futures prices is likely to change; and if so, will the differential increase or decrease?


One of the attractive aspects of the TED spread is its simplicity. An expectation that the spread will widen justifies buying the spread (i.e., buying Treasury bill futures and selling Eurodollar futures), while an expectation of a narrowing of the differential justifies selling the spread. The appropriate trade proportions are one-to-one.4


1 Eurodollar deposits are dollar-denominated bank deposits held by commercial banks outside the continental U.S.


2 A revision of the Treasury bill contract currently under consideration will result in analogous cash settlement processes for Treasury bill futures.


3 The International Monetary Market is the division of the Chicago Mercantile Exchange that lists interest rates and currency futures and options.


4 This trade proportion will foster zero gains and losses as long as the futures price (rate)


Generally, trades are executed with a single order to buy or sell the TED, and prices are assigned to the respective “legs” of the spread. Note that, as long as the differential is appropriately reflected by the assigned prices, the individual prices are irrelevant, because any “error” would be offsetting on the two legs.


The cash market TED spread (i.e., based on spot market interest rates, rather than futures prices) has shown fairly extended periods of relative stability, interrupted by sharp movements and higher volatility. The 1979-1982 period was one high-volatility time span, when the spread ranged from a high of about 400 basis points to a low of about 100. This episode was associated with the effects of Paul Volcker’s monetary policy and the Mexican debt crisis.


A few years later, another bout of volatility appeared in May 1984, in connection with the Continental Illinois National Bank crisis. At that time, it appeared that a major money center bank was on the threshold of failing. The entire banking system seemed vulnerable, and investors became considerably less certain of the safety of Eurodollar deposits. As a consequence, a “flight to quality” ensued, whereby investors shifted their holdings out of Eurodollar deposits and into U.S. Treasury bills. Banks were forced to counter this shift by bidding up yields on Eurodollar deposits.


In response to the prospect of a major bank failure, the U.S. government (the FDIC, the Federal Reserve System, and the Comptroller of the Currency) took measures to keep Continental Bank afloat by protecting its depositors (uninsured, as well as insured). The crisis was defused, confidence in the system was restored, and the TED spread worked its way from its high of about 200 basis points to its pre-crisis level of about 100, in a span of a few months.


Since the mid-1980s, the TED has been much more quiescent. In the most recent two years, for example, interest rates have been remarkably stable. Even so, during this time the TED spread has oscillated over a 40-basis point range, allowing ample opportunity for trading activity.


Besides the flight to quality issue, the TED is also traded as a surrogate for a play on the level of interest rates. Exhibit 1 shows a direct relationship between the level of interest rates and the size of the TED. That is, during the interest rate cycle shown, the TED generally narrowed when rates declined and widened when rates rose. This relationship is by no means certain in any given short-run situation.


Differential remains constant, independent of the fact that the two respective interest

rates are quoted using different rate quotation conventions (i.e., discount rates for

Treasury bills versus add-on yields for Eurodollars).

TRADE IMPLEMENTATION ISSUES

When trading futures contracts, it is often appropriate to consider the “basis,” the difference between the futures price and the spot price of the associated underlying instrument. Specifically, if the expected change in the spot price over the horizon to the futures expiration date is already embedded in the futures price, the rationale for initiating a position could be mitigated.


To clarify, first consider the issue in the context of an outright futures trade, rather than a spread trade. Suppose, for example, that the discount rate on the three-month (spot) Treasury bill is 5.00%, and you expect this rate to rise by, say, 25 basis points, to 5.25% by the next futures expiration. You would likely be tempted to sell the nearby Treasury bill futures contract.5


5 Recall that futures prices reflect 100 minus futures interest rates. An expectation that the interest rate will rise is equivalent to an expectation that the futures price will fall.


If the futures price were currently trading at 94.75 (already reflecting a rate of 5.25%), however, the trade would not be appropriate. That is, if expectations were realized, the futures price would remain at its current price of 94.75, so that no profit would result from the trade. In fact, if the current futures price were below 94.75, and you thought the increase in the spot rate would be limited to no more than 25 basis points (i.e., the futures price would go no lower than 94.75), a long futures position would be justified.


The same basis consideration applies when trading spreads. Prior to the expiration of one of the underlying futures contracts (i.e., the first of the two relevant expirations), the TED spread constructed with spot interest rates will likely differ from the spread constructed with futures prices, but spot and futures spreads will generally converge as the expirations approach. As long as the Treasury bill futures require physical delivery, convergence to the spot price will not necessarily be complete, as the Eurodollar and Treasury bill futures expirations do not occur simultaneously. An intended change to a cash-settled Treasury bill contract, however, will synchronize the expirations and foster complete convergence.


To demonstrate, consider the conditions shown in Exhibit 1, reflecting prices (rates) some months prior to the expiration of the Eurodollar futures contract.

In this case, the futures TED is trading at a 35-basis point premium to the spot TED. As a consequence, if spot rates remain unchanged over the next two months, the futures spread would have to adjust downward, generating a gain for the short TED spread position. Put another way, for a long futures TED spread to be profitable, the spot spread would have to rise by more than 35 basis points by the time the first component futures contract expires. Thus, the spread basis fosters an edge to one side of the trade and somewhat of a penalty to the other. In this case, the edge accrues for the short TED position, while the penalty applies to the long.


The basis consideration is particularly important in the context of using a TED spread as a surrogate for an outright interest rate position — particularly when yield curves are steeply sloped. When yield curves are steeply upward-sloping, outright futures prices generally trade at a large discount to spot prices; when yield curves are downward-sloping, futures trade at a large premium.


As noted earlier, to the extent that basis conditions reflect expectations of a prospective market move, the rationale for trading is mitigated. If the basis conditions were roughly equivalent for both Treasury bill and Eurodollar futures, however, the TED basis would be close to zero (i.e., the two basis effects are offsetting). As a consequence, no basis edge or penalty would apply to the TED. Thus, when basis conditions appear to be unattractive for outright futures, the TED spread may not suffer the same unattractive characteristics.


THE TERM TED SPREAD

Beginning in the early to middle 1990s, the market took trading basic TED spreads a step further. Participants started trading longer-term Treasury notes against strips of Eurodollar futures.6 For example, buying (selling) a two-year term TED would involve buying (selling) a two-year cash Treasury note in the spot market and selling (buying) a two-year Eurodollar futures strip (i.e., a strip composed of eight contracts).


6 Strips are a construction that employs a sequence of successive Eurodollar expirations (e.g., March, June, September, etc.).


The decision to enter a term TED spread (rather than the traditional TED) is justified by the same kind of interest rate expectations that motivate the traditional TED spread, albeit for a different point on the yield curve. The earlier example, for instance, focuses on the two-year points on the Treasury and Eurodollar yield curves. Just as before, a judgment that the spread will widen would justify buying the Treasury instrument and selling the strip, and vice versa for the reverse expectation.


The starting point for assessing whether a term TED is an attractive trade requires evaluating the strip yield from a set of component Eurodollar futures prices and then comparing this calculated yield to the yield on a government obligation of similar maturity. A rigorous calculation of strip yield, which reflects the compounding of interest, requires a two-step process.


First, explicitly using the days between successive Eurodollar futures value dates in the calculation, we find the terminal value of a dollar by assuming a quarterly reinvestment/refunding rate based on successive futures rates, as per Equation (1):

Given this value of Tn, the yield to maturity of the n-quarter strip is found by solving Equation (2) for R*:


R* therefore pertains to the yield to maturity of a zero-coupon instrument with quarterly compounding.


Strictly speaking, this rate calculation is not comparable to rates quoted on Treasury notes and bonds for two reasons:


1. While the strip yield pertains to a zero-coupon instrument, the Treasury is a coupon-bearing security with semiannual compounding.


2. The strip corresponds to a forward time interval beginning with the first futures value date and extending three months beyond the last futures value date, while the Treasury pertains to an imminent time span from today’s spot value date to the maturity date.7


7 Some analysts deal with the latter concern by adjusting the strip yield calculation to include a “stub” – a multiplicative term [1 + Rs(ds/360)] to reflect the period preceding the strip, where Rs is the stub rate and ds is the number of days from the spot value date to the first futures value date. Additionally, the weight given to the last futures contract in the yield calculation would typically be adjusted to make the coverage of the combined stub plus strip correspond to the remaining life of the bond or note.

PREDOMINANT TRADE CONSTRUCTION METHODOLOGY

Although no unique method for arranging a term TED is universally accepted, a number of analytical subscription services offer algorithms that are widely used. While the precise Eurodollar futures positioning may differ somewhat from model to model, most of these services recommend a disproportionate weighting of successive futures contracts, with declining numbers of futures for more distant expirations. This outcome follows if one conceptualizes the Treasury security as