Understand how implied liquidity is generated in CME Globex implied inter-commodity spread (ICS) markets (instrument type: IV), which include spreads for Treasury, Credit, UMBS TBA, and Eris swap futures.
For more information, see: Interest Rate futures Inter-Commodity Spreads
Introduction
An ICS is a listed spread between two legs of pre-defined quantities. Buying (selling) the spread involves buying (selling) a given quantity in the first leg and simultaneously selling (buying) a given quantity in the second leg. Since an ICS is a listed spread, it has its own separate central limit order book.
For example, the NON is a listed ICS that results in a long position in one 10-Year Treasury Note futures (ZN) contract and a short position in one Ultra 10-Year Treasury Note futures (TN) contract. Figure 1 shows live markets for ZN, TN, and NON as shown on CME Direct.
Figure 1: Markets for ZN, TN, and NON
For implied ICS, CME Globex generates implied orders in one market based on orders that exist in related markets. These implied orders can increase market depth and improve prices. There are two different types of implied orders: implied in orders and implied out orders. Implied in orders exist in the ICS market based on orders in the two outright leg markets. Implied out orders exist in an outright leg market based on orders in ICS markets and other related outright leg markets. It is important to note that many of the Treasury futures contracts are included in multiple ICS contracts, e.g., 2-Year Treasury Note futures are included in the TUF, TUT, and TUX ICS contracts, as well as many others. In addition, in some cases, CME Globex offers multiple ICS markets between the same pair of futures with different ratios.
Implied orders are subject to the following rules:
- Implication requires a minimum of two orders in related markets in the proper combination.
- Implied bids do not trade against implied offers.
- Implied bids can exist at the same or inverted price levels with implied offers and although these implied bids are calculated and can be traded against, CME Globex will not disseminate them.
- Implied out orders will not be disseminated when the spread ratio is greater than one; however, implied out prices and quantities are still calculated and can be traded against.
- Implied orders in futures markets do not have time priority:
- Direct spread orders will take precedence over implied orders at a given price.
- Implied outrights created by ICS are filled according to a predefined priority.
- Implied quantity is not available when:
- The market is not in a matching state (e.g., Pre-Open).
- Implied calculation has been suspended (e.g., CME Globex detects a trade occurring outside of limits because of implication).
In figure 1, the 1 “Imp” (abbreviation for “implied”) columns represent the disseminated quantity of implied orders in each market – the left column for implied bids and the right column for implied offers. Note that CME Globex does not disseminate all calculated implied orders, meaning that hidden implied liquidity may exist that can be traded against. For example, hidden implied out liquidity is generated from ICS markets where the spread ratio is not one-to-one.
How implied liquidity is generated
The following example demonstrates the mechanics of implied pricing using the markets shown in figure 1.
The NON was chosen in this example for two reasons:
- Both legs share the same minimum price increment (MPI). Calculating implied prices involving contracts with different MPIs requires rounding to ensure that implied prices match the relevant MPI (this is discussed later).
- This ICS has a 1:1 spread ratio, which allows us to validate calculated implied quantities against implied quantities disseminated to the order book. This is because implied orders involving ICS with spread ratios other than 1:1 may not be disseminated.
Figure 2: NON parameters
Implied ICS follow a net change on day pricing convention, such that:
- PICS is the ICS price.
- P1 is the price of the first leg.
- P2 is the price of the second leg.
- S1 is the previous daily settlement price of the first leg.
- S2 is the previous daily settlement price of the second leg.
- r is the price ratio and is defined as:
For example, using the information in figure 2, the price ratio for NON is calculated as:
Note that the price ratio may not be equal to one if the spread ratio is not 1:1 or if the legs have different notional values. For example, the price ratio for the IQT ICS between IQB ($30 notional value per point) and ZN ($1,000 notional value per point) can be calculated as:
Implied in prices
To calculate prices that are implied into the ICS market, we need to calculate the prices that an aggressor would face if they replicated an ICS trade by legging the two outright markets. Selling the ICS at bid is equivalent to selling leg 1 at bid and buying leg 2 at ask. Using the relationship between ICS prices and leg prices, this implies:
- BICS is the ICS bid price.
- B1 is the leg 1 bid price.
- A2 is the leg 2 ask price.
Similarly, buying the ICS (at ask) is equivalent to buying leg 1 at ask and selling leg 2 at bid:
- AICS is the ICS ask price.
- A1 is the leg 1 ask price.
- B2 is the leg 2 bid price.
Using the markets for ZN and TN shown in Figure 1, this implies a market for the NON shown in Figure 3.
Figure 3: NON implied in market
Observe that the implied in market has a bid-ask spread of 1/32nd, which is equal to the sum of the bid-ask spreads in the two outright markets. This is intuitive since the implied market is generated by crossing the spread in each of the two outright markets. However, it is important to note that this relationship may not hold if the outright legs follow different pricing conventions, due to rounding (this is discussed later).
Comparing the implied in market shown in Figure 3 with the live market shown in Figure 1, we can see that the market bid of -3/32nd is higher than the implied in bid of -3.5/32nd. The market is tighter because of direct bids in the ICS for 2 lots at -3/32nd. However, the top-of-book NON ask (-2.5/32nd) is the same as the implied ask shown in Figure 3 (-2.5/32nd).
Implied in quantities
In Figure 1, we can see that the top-of-book NON market has 2 lots on the bid and 375 lots on the offer. In general, if the implied bid/ask is top-of-book:
ICS implied in bid quantity = min(Leg 1 bid quantity, Leg 2 ask quantity)
ICS implied in ask quantity = min(Leg 1 ask quantity, Leg 2 bid quantity)
Since the implied in bid is not top-of-book, the “Imp” bid column is blank and the 2 lots on the bid are direct quotes. Since the implied in offer is top-of-book, an implied in quantity of min(1344, 330) = 330 lots on the offer is shown in the NON order book. The other 45 lots offered at that price are direct orders in the ICS.
Implied out prices
To calculate prices that are implied out of the ICS market into one of the related outright markets, we need to calculate the prices that an aggressor would face if they replicated an outright trade by legging the ICS market and the other related outright market. Without loss of generality, we will calculate the implied out leg 2 (TN) prices using the ICS (NON) and leg 1 (ZN) markets, but the same process can be followed to calculate the implied out leg 1 prices. Selling leg 2 at bid is equivalent to buying the ICS at ask and selling leg 1 at bid. Substituting the correct prices for the ICS and each of the two legs into the pricing relationship introduced earlier:
Rearranging this equation to make the leg 2 bid the subject:
Similarly, buying leg 2 at ask is equivalent to selling the ICS at bid and buying leg 1 at ask:
Using the markets for NON and ZN shown in Figure 1, this implies a market for the TN shown in Figure 4.
Figure 4: TN implied out market
Again, notice that the implied out market has a bid-ask spread of 1/32nd, which is equal to the sum of the bid-ask spreads in the two markets where prices are implied from.
Comparing the implied in market shown in Figure 4 with the live market shown in Figure 1, we can see that the market bid of 113’10.5 is higher than the implied in bid of 113’10. The market is tighter because of direct bids for 330 lots at 113’10. However, the top-of-book TN ask (113’11) is the same as the implied ask shown in Figure 4 (113’11).
Implied out quantities
In figure 1, the top-of-book TN market has 330 lots on the bid and 230 lots on the offer. In general, if the implied bid/ask is top-of-book:
Leg 2 implied out bid quantity = min(Leg 1 bid quantity, ICS ask quantity)
Leg 2 implied out ask quantity = min(Leg 1 ask quantity, ICS bid quantity)
Since the implied out bid is not top-of-book, the “Imp” bid column is blank and the 330 lots on the bid are direct quotes. Since the implied out offer is top-of-book, an implied out quantity of min(1322, 2) = 2 lots on the offer is shown in the TN order book. The other 228 lots offered at that price are direct quotes.
ICS involving different MPIs
Earlier it was mentioned that there are complications involved when calculating implied prices involving two contracts with different MPIs, for example:
- The FYN ICS between 5-Year Treasury Note futures, which have an MPI of ¼ of 1/32, and 10-Year Treasury Note futures, which have an MPI of ½ of 1/32
- ICS between 30-Year UMBS TBA futures, which have an MPI of ¼ of 1/32, and 10-Year Treasury Note futures, which have an MPI of ½ of 1/32
ICS between Eris swap futures, which use decimal pricing, and Treasury futures, which use fractional pricing.
When this is the case, implied bids and asks may not be eligible prices, i.e., not multiples of the MPI. This can also occur if the MPIs are the same for each leg, but the spread ratio is not 1:1. To ensure that the implied prices are eligible, or “on tick”, the bid is rounded down to the nearest eligible price and the ask is rounded up to the nearest eligible price.
Price improvement from implied price rounding
An important question is who benefits from the price improvement caused by implied price rounding?
For implied in orders, the price improvement benefits the ICS aggressor. Meanwhile, for implied out orders, the price improvement benefits the passive ICS order.
To see how this works in practice, consider the markets for 10-Year Eris Swap futures (YIY), Ultra 10-Year Eris swap futures (TN), and the ICS between them (EYT) shown in Figure 5. The EYT is a listed ICS that results in a long position in one YIY contract and a short position in one TN contract, with parameters shown in Figure 6.
Figure 5: Markets for YIY, TN, and EYT
Figure 6: EYT parameters
Using the markets for YIY and TN shown in Figure 5, we can calculate the implied in EYT bid as:
However, this bid is not a multiple of 0.01 (the EYT MPI) and is not an eligible price. Because of this, Globex rounds out this bid to -2.18, as shown in Figure 5.
While an aggressor who hits this bid sees an on-screen price of -2.18, they will effectively face the true unrounded price of -2.17125. Meanwhile, the resting orders in YIY and TN will be filled at their respective bid and ask prices.
Suppose that a direct bid order is entered in the EYT market at a price of -2.12, improving the EYT market. To calculate the bid price that is implied out to the YIY market, we need to calculate the prices that an aggressor would face if they replicated selling YIY by legging the two outright markets. Selling leg 1 (YIY) at bid is equivalent to selling the ICS at bid and selling leg 2 (TN) at bid. Using the relationship between ICS prices and leg prices, this implies:
However, this bid is not a multiple of 0.02 (the YIY MPI) and is not an eligible price. Because of this, Globex rounds out this bid to 92.04.
The aggressor who hits the implied out YIY bid will face the bid price of 92.04 and the passive bid order in TN will be filled at its bid price. However, the passive bid order in the EYT will benefit from the price improvement caused by rounding and will be filled at a price of:
All examples in this report are hypothetical interpretations of situations and are used for explanation purposes only. The views in this report reflect solely those of the author and not necessarily those of CME Group or its affiliated institutions. This report and the information herein should not be considered investment advice or the results of actual market experience.