Calendar Spread

Calendar Spread

1. Front and Back Contracts

Every financial derivative has a date on which the contract matures. On that day, the value of the deriva-
tive is the contingent payoff. Depending on the market conditions prevailing on the maturity date, the
payoff is fixed. As soon as the payoff is confirmed and processed for settlement, the contract expires and
exists no more. Futures contracts are no different.
In the futures markets, contracts on the same underlying asset but different maturity dates are offered
by the exchanges for traders to trade. The front-month futures contract refers to a contract for which the
expiration date is closest to the current date. Contracts that mature a month or so longer than the front
month contracts are known as back month contracts.
Many futures contracts do not mature on the monthly basis. For example, the Brazilian Ibovespa
index futures contract matures on the even-numbered months (February, April, June, August, October,
and December). Major index futures contracts, such as the Emini S&P 500 index futures, are on the
quarterly cycle; the maturity months are March, June, September, and December. It is more appropriate,
therefore, to refer to the futures contract that are nearest to maturity as the front contract, and the next
nearest as the back contract.
The calendar spread is simply the market price difference:

calendar spread := price of back contract − price of front contract.

You can see from this definition of calendar spread that it is market neutral. Whether the market is going
up or going down, the value of the calendar spread does not change. By trading calendar spread, you are
taking a neutral stance, as you don’t bet on the market direction. You don’t even have to have an outlook,
going forward.
What will happen to the calendar spread position when price moves? Suppose the price of front
contract moves up by 14 ticks, the back contract is most likely to move up 14 ticks as well. Since the
calendar spread is essentially a subtraction of one price from another price, the move of 14 ticks cancel
out. Hence, the price of the calendar spread remains unchanged. As an illustration, consider the MSCI
Singapore index futures, which is known as the Simsci futures on the street. Suppose the front contract is
traded at 350.10, and the back month’s market price is 349.00. The calendar spread price is therefore

349.00 − 350.10 = −1.10.

If the front and back markets move down by 14 ticks, the two futures prices will become, respectively,
349.40 and 348.30, because each tick is 0.05 index points. Despite the price moves, the calendar spread
is still −1.10(= 348.30 − 349.40). Therefore, with calendar spread, the risk of making a wrong bet in the
market direction is mitigated.
Everything else being equal, typically the front futures contract is the more liquid compared to the
back contract, and other longer-term futures contracts. So logically, to execute a calendar spread, it makes
sense to trade the back contract first, and then immediately execute a hedging trade of the front contract.
Following the market convention, buying the back futures contract in conjunction with selling the
front futures contract constitutes a purchase of the calendar spread. Conversely, selling the back contract
and buying the front contract is known as taking a short position in the calendar spread. It goes without
saying that to make a round trip of a calendar spread, you need four trades: two to open a calendar spread
position, and another two to close it.

c Christopher Ting Page 2 of 7
 Calendar Spread

What’s the point of trading a calendar spread? The point, really, is to earn the difference between
the bid and ask prices in a much less risky way. Every market maker posts bid and ask quotes at the
market to attract customers to trade with him. If the price moves ware mild, the market maker could
make money with little risk by adjusting his quotes and position size. The same market making strategy
can be implemented using limit orders at the bid and at the ask in the electronic market. But when the
market moves rapidly, if the market maker is not nimble in adjusting both the quotes and the position
size, he may lose substantially when the market direction runs against his position.
One of the ways to circumvent this price risk is to trade the calendar spread. If you sell the calendar
spread at −1.0 and manage to square off your short position at −0.9, you will make a profit of 0.1 index
points. The calendar spread trading is the round-about way of earning the bid-ask spread in a way that
is (almost) invariant to the market direction. In that sense, calendar spread traders provide liquidity to
the market. We shall see in the subsequent sections that the liquidity provision is indeed the positive
externality of calendar spread trading.
But like all strategies, calendar spread trading is also not immune to risks. Chief among the risks
is the sudden jump in both the quoting leg and the hedging leg. The jumps will lead to mis-hedges,
which mean losses, more often than not. When mis-hedge occurs, the spread is said to be legged and
a directional trade is said to have occurred In the age of high-frequency algorithmic trading, a sudden
intra-day plunge or surge in the electronic market appears to occur more frequently. These sudden jumps
are reinforced by the trigger of stop orders to cut loss. In other words, spread trading in general is subject
to the risk of mis-hedge when both legs jump.

2. Fair Value of the Calendar Spread

You may wonder why should the calendar spread be of a particular value, such as −1.10 in the earlier
example. Futures is a simple linear-payoff derivative, its fair value can be easily derived. In the case of
calendar spread, its fair value is simply the difference between the fair value of the back contract and that
of the front contract:
theoretical calendar spreadt = Fback,t − Ffront,t .

Here, Fback,t is the fair value of the theoretical price of the back contract while Ffront,t is the fair value of the
front contract at any given time t.
Recall that the fair value of a futures contract is primarily determined by the cash index. When ex-
pressed in index points, the fair value is

cash index + costs of holding the underlying asset − benefits of holding the underlying asset.

In a calendar spread, the cash index is common to both the front and the back contracts. Being a difference,
the cash index cancels out. Therefore the value of the theoretical calendar spread is marginally dependent
of the cash index value.
What about the costs of holding the underlying assets? Suppose the only cost is the interest amount.
The front month contract matures at time T1 and the interest at time t is

cash index × r1,t × ( T1 − t).

where r1,t is the benchmark risk-free rate corresponding to the tenor of T1 − t years.

c Christopher Ting Page 3 of 7
 Calendar Spread

Likewise, for the back month contract with the maturity date T2 , the interest amount is

cash index × r2,t × ( T2 − t).

Analogously, r2,t is the benchmark risk-free rate at time t for tenor of T2 − t years. The fair value of the
calendar spread attributable to the interest rate is


cash index × r1,t ( T1 − t) − r2,t ( T2 − t) .

As an illustration, suppose the front contract has 15 days to maturity and the back contract matures in
45 days. From the swap rate curve, you find that the annualized interest rates are 1.0% and 1.05%, respec-
tively, for these two tenors. Suppose further that there is no benefit accrual from holding the underlying
asset. Moreover, the cash index level is 340.44, and that a year has 365 days. The theoretical value of the
calendar spread is, in index points,
 
45 15
340.44 × 0.0105 × − ×0.01 × = 0.30.
365 365

This differential in the interest amount, which is the fair value of the calendar spread in this case, is
0.088356% of the cash index level. In terms of ticks, the calendar spread is 6 ticks (since every tick is 0.05
index points).
What about if a component stock is paying dividend? Suppose the dividend payment is $0.40 per
share and the ex-date is 30 days from time t. Moreover, suppose the (free-float) shares outstanding is 100
million shares. The dividend payment works out to be 40 million dollars. To convert the dollar amount
to index points, simply divide by the cash index’s divisor.
Using the same example of the Simsci futures, since the front contract expires in 15 days’ time, the
front contract does not get to receive the dividend. The back contract expiring in 45 days will have this
dividend benefit. For simplicity, suppose the divisor is 25 million. Accordingly, the dividend payment of
40 million dollars is equivalent to 1.6 index points, which correspond to 32 ticks.
When combined with the earlier analysis involving the interest amount, the fair or theoretical value
of the calendar spread is, in index points,

0.30 − 1.60 = −1.30.

Notice that this fair value of the calendar spread is different from −1.10 traded in the market. It should not
come as a surprise to you that the market price of spread is different from the fair value. The difference
of 0.20 or 4 ticks does not necessary mean that the best bid and best ask is 4 ticks apart. Neither does it
imply that there is an arbitrage opportunity. To traders, the disparity represents the market expectation
of changes in costs and/or benefits. The expected changes are not equal for the front and back contract.
In this particular example, the market expects the cost of interest rate to rise.

3. Outright Calendar Spread

Futures contracts entail obligations on both the buyer and the seller. Holding a futures position through
expiration legally binds the holder of the contract to settle the position in accordance to the futures’ terms
of delivery. Most traders usually square off their futures position by going into an equal and opposite

c Christopher Ting Page 4 of 7
 Calendar Spread

transaction before the last trading day of the contract. The intention is to avoid delivery and its accompa-
nying uncertainties altogether.
Another way to avoid delivery is to roll forward by closing off the expiring position (font contract)
and going into a similar futures transaction for a later expiration (back contract). Managed futures hedge
funds run by commodity trading Advisors (CTA) need to roll over their long or short positions regularly.
To continue in a long position, CTA will buy the calendar spread. By so doing, effectively the CTA closes
the long position in the front contract expiring soon, and simultaneously opens a new long position in the
back contract.
To facilitate the roll, exchanges offer a product called the outright calendar spread. By trading the
outright spread, CTA can achieve their roll objective with absolute certainty. In other words, there is
no possibility of missing the boat when prices in the near and back contracts start to shift rapidly. This
certainty provides CTA the peace of mind in managing their myriad futures positions. Being important
clients as their trade sizes are large, the outright calendar spread is a great innovation that serves them
well.
But more importantly, the outright calendar spread, by design, ensures that the demand to roll a large
position will not create havoc on both the front and back futures themselves. If there was no calendar
spread and if a CTA tried to buy and sell directly on the front and back contracts, then their trades would
shock the liquidity providers (limit-order traders), leaving them no alternative but to pull their offer prices
higher when CTA was buying on one calendar leg, and to lower their bid prices on the other calendar leg
as CTA was selling heavily. Due to the large price impact, the calendar spread would tear wider. This
distortion is extremely detrimental to CTAs, and to the integrity of the futures market as a whole. The
same tear will happen not just for CTAs but also any institutional investors who need to roll over a large
position.
Nearer to the expiration, the limit orders on the outright calendar spread increase substantially. Many
prop traders, CTAs, and other institutional traders strive to get in front of the queue to buy and to sell
the outright calendar spread. Prop traders who usually do not carry an inventory of calendar spreads
overnight become active. By getting in front of the queues and placing big chunk of limit orders at
the market, they effectively become market makers, providing liquidity to CTAs and calendar rollers
in general.

4. Synthetic Calendar Spread

These days, professional trading software allows prop traders to configure a strategy to trade the calendar
spread. The resulting price ladder is known as the synthetic calendar spread. By “synthetic”, a distinction
is made from the outright calendar spread, which is a derivative product offered by the futures exchange.
A synthetic calendar spread is made up of two legs: quoting leg and hedging leg. Traders prefer to
call the futures contracts as legs when they open or close a calendar spread position. The difference in
prices of the quoting and the hedging legs is the spread.
The quoting leg is the bait. As a matter of practice, the bait should be placed on the leg that is less
liquid The idea is that as soon as the bait is taken, you can hedge at the hedging leg without difficulty. By
hedging, you establish a spread position. Given the higher liquidity at the hedging leg, it is more likely
(but not always!) for you to hedge at the more liquid leg whenever the bait is taken at the less liquid leg.
Specifically, an at-the-market limit order will be submitted to the hedging leg’s limit order book as soon

c Christopher Ting Page 5 of 7
 Calendar Spread

as the bait is taken. If the bait is a sell limit order, a limit price at the best offer price will be submitted to
the hedging leg to buy as quickly as possible. Conversely, if the bait is a buy limit order, then the synthetic
calendar spread requires the trader to sell, as soon as possible, at the best bid price of the hedging leg.
The bait has to be dynamic. Its limit price must change in tandem with the market price of the less
liquid quoting leg. The reason is that the synthetic spread price is a target price, and to achieve the target
price, the limit price of the bait has to adjust according to the market prices at both legs. For example,
suppose you want to buy the spread at −1.10. Your quoting leg is the back contract, which is at the best
bid and ask of 360.10 and 360.15. Your bait is a sell limit order at 360.15, as the front contract serving as
the hedging leg is traded at the best bid of 361.20 and the best ask of 361.25. If the bait is taken at the offer
price of 361.15, then a buy limit order will be submitted immediately to the hedging leg to attempt to buy
at the offer price of 361.25. The difference of these two prices is 360.15 − 361.25 = −1.10, and you have
achieved the goal of selling the calendar spread at −1.10.
Essentially, a dynamic limit order is really a series of cancelation orders and limit orders. A simple
algorithm is to submit a cancelation order to cancel the current limit order, and simultaneously submit a
new limit order with a different limit price. The limit price difference is usually a tick. When the market
moves up by a tick, the bait will need to move up by a tick as well. Likewise, the bait’s limit price must
move down a tick whenever the market moves down by a tick.

5. How Is the Bait Taken?

As mentioned earlier, the limit order as a bait in the quoting leg is dynamic in nature. When the market is
moving up by one tick, the bait will also move up by a tick. Likewise, the bait will move down by a tick
when the market moves down. So how it is possible for the bait to be taken or hit?
It turns out that a major factor leading to the bait being traded against an incoming market or mar-
ketable limit order depends on how far the bait is from the market (best bid or best ask). If the bait is at
or one tick behind the market, a BIG market or marketable limit order will consume the liquidity in the
quoting leg, including the limit order of your quoting leg. In fact, by posing a limit order, albeit a dynamic
one, calendar spread trader is providing liquidity to the market participants for the quoting leg, which
lacks liquidity.
Now, as soon as a BIG liquidity consuming order arrives at the quoting leg, it will surely impact the
limit order book by taking out a few levels up if it is a buy order or down if a sell order.
As discussed earlier, to prevent not achieving the target spread price, the default order is a limit order
with the limit price set at the price by which the target spread price is achieved. In the case of bait taken
at the market, the limit price is the best offer price immediately before the bait is taken. But things won’t
happen exactly to your preference. If the best offer price at the hedging leg has already been bought
by other traders, your limit order will not get executed, and a mis-hedge is said to have occurred. To
absolutely prevent mis-hedges, of course you can also use the market order to buy. But the danger is that
you might not be able to sell the calendar spread at the spread price you have set out to achieve.
As an example, consider 6E futures for EUR/USD. The outright spread is traded at 30.0 bid and 30.5
offer. Suppose the contract at the quoting leg (6E Jun16) is your bait at 10934. If it is taken and if the sell
limit orders at the hedging leg (6E Mar16) are still there, you should be able to hedge by buying at the
price of 10903. With these two trades, you have opened a short spread position of 31. Note that this is
higher than the best offer price of 30.5 at the outright calendar spread.

c Christopher Ting Page 6 of 7
 Calendar Spread

6E Jun16 6E Mar16 6E Jun16 - 6E Mar16

Bid Qty Price Ask Qty Bid Qty Price Ask Qty Bid Qty Price Ask Qty
10938 12 10907 54 32.5 450
10937 12 10906 64 32.0 350
10936 14 10905 49 31.5 100
10935 7 10904 45 31.0 25
10934 1 10903 10 30.5 32
1 10931 8 10902 119 30.0
15 10930 62 10901 418 29.5
14 10929 46 10900 450 29.0
12 10928 55 10899 50 28.5
52 10927 38 10898 56 28.0

Table 1: The back and front contracts, and the outright spread of EUR/USD futures on December 17, 2015
7 PM EST.

If the hedging leg moves up by one tick, your current bait should be canceled and a new bait is
submitted at a tick higher, i.e., 10935 is the new limit price. Conversely, the new bait will be a tick lower
at 10930 if the hedging leg moves down by a tick.
Now, what if you want to short the calendar spread at a higher price of 32? In this case, your bait
should be at the price of 10935, which is behind the market. For your bait to be taken, a buying order
must arrive, take out the one contract offered at 10934 and then take some liquidity at 10935. Your hedging
order will be sent out as fast as possible to try to buy at the price of 10903. But it is by no mean an easy
task, because given a strong buying pressure at the quoting leg, the hedging leg will also move up. Either
the 10 contracts at 10903 are canceled, or they are taken by other traders. If your system can beat them and
buy before these 10 contracts disappear, then you are in a very good position to profit from this statistical
arbitrage. Otherwise, you will surely suffer many mis-hedges.
One of the solutions is to make your hedging a litter bit smarter. If mis-hedge occurs, cancel the mis-
hedged limit order and simultaneously submit a marketable limit order to buy. If you are lucky, you could
buy at the price of 10904, in which case you manage to short a calendar spread at the price of 31. This
calendar spread price of 31 is the same price as the earlier scenario when your bait is at the market.
In any case, it is very important to ascertain the liquidity of the hedging leg before putting the bait. If
you have historical tick-by-tick data you can back-test to examine the jump frequencies, the quantum of
price changes (one tick, two ticks, and so on) and their occurrence frequencies. Also useful is to examine
how the lead-lag relationship between these two calendar months.

6. Quoting on Both Legs

When the front contract is about to expire and the back contract is about to take over as the front contract,
liquidity on both legs tends to become more or less the same. In other words, on the last trading day of
the front contract and a few days prior to that day, you can put baits on both legs. The objective is to
increase the number of calendar spread trades.

c Christopher Ting Page 7 of 7