CME to reinforce term SOFR with swap inputs

Inclusion would leapfrog a 25% OTC liquidity threshold embedded in methodology


CME Group is in discussions with data providers about adding over-the-counter swaps to the pool of inputs for calculating its term version of the secured overnight financing rate, or SOFR.

The benchmark administrator, which currently publishes the rate only from its own futures data, expects to incorporate swap inputs well before OTC instruments hit a liquidity threshold that would compel the group to consider adoption under its own methodology.

“We are already working on getting ready to include swaps in CME term SOFR. We have some sources already lined up for providing swaps data and there are other providers we are engaged with,” says Agha Mirza, global head of OTC products at CME.

CME term SOFR is calculated from $1 trillion-a-day of futures traded on the Chicago bourse. Under the published methodology, SOFR overnight index swaps (OIS) with tenors below one year would be “considered for inclusion” in the event transacted monthly volume exceeds 25% of SOFR futures volume over a consecutive six-month period.

While SOFR futures volumes have grown fourfold over the past year, OTC swaps have also grown exponentially, albeit from a lower base. Data from the International Swaps and Derivatives Association show SOFR swaps traded $12.7 trillion notional during the first quarter of 2022, averaging around $200 billion per day. Contracts with maturities of up to one year account for less than a third, leaving relevant swaps volumes around 5% of equivalent futures tenors.

“We fully expect to include swaps prior to hitting the qualifying 25% threshold,” says Mirza. “What we’ve seen in recent months is that swaps have fallen to 5% of futures due to an exponential rise in futures trading, but we are preparing to include swaps data because more transactions are better.”

With an endorsement from the Alternative Reference Rates Committee, the Federal Reserve-backed group tasked with ushering US markets off Libor, CME term SOFR has emerged as a preferred Libor successor for US lending.

Data from Refinitiv shows around 1,000 lending facilities worth a combined $1.3 trillion are now pegged to term SOFR. An additional $150 billion of derivatives, including swaps, caps and floors, are also linked to the rate, trade repository data show.

“We’ve seen very significant market engagement on term SOFR,” says Mirza. “We’ve had a massive amount of inbound client enquiries and have made a lot of progress on the licensing front.”

The exchange group has issued 4,500 licences for the rate to more than 1,000 firms across 70 countries.

Futures and swaps in the dollar market trade nearly all the time on a no-arbitrage basis. The interest rate information should be the same in both markets
Agha Mirza, CME

Yet CME is not the only term SOFR in town.

The rate now faces competition from Ice Benchmark Administration, which launched its own term SOFR rate in March. This alternative rate is based on committed swap quotes from electronic central limit order books, aligning with Ice’s term Sonia rate, which was selected by UK regulators as the basis of a ‘synthetic Libor’ to mop up so-called tough legacy sterling contracts.

A debate over the relative merits of futures-based and swaps-based methodologies has raged since global regulators first invited proposals on term rates back in 2019.

Forward-looking risk-free rates represent market expectations of overnight rates in the future. In swap markets, dealers use their own internal models to derive rate expectations from futures and other data, meaning assumptions are baked into their price streams. With futures, benchmark administrators must apply their own set of assumptions to infer term rates from contract prices.

One rates trader warns the standardised nature of futures means that term rates derived from the instruments can be incorrect around rate moves. If a Federal Open Market Committee meeting falls within the contract, any hike would be averaged across the duration of that contract. In OIS markets, this hike would be represented as a single-day jump.

“Now we face multiple hikes, I guess the issue has been compounded and they need to address it,” the rates trader says.

CME’s futures-based methodology follows a blueprint laid down in a 2019 Federal Reserve working paper by Erik Heitfield and Yang-Ho Park. Their analysis reveals that while futures and swaps-derived term rates track closely, they can diverge during times of stress, particularly in shorter tenors.

The paper shows the mean difference in one-month rates implied from fed funds futures and fed funds OIS is just 1.5 basis points over the long term – in this case from January 2000 to January 2019. But that figure rose to more than 10bp between July 2007 and December 2010 during the global financial crisis.

Mirza says published rates are unlikely to be affected by the inclusion of swaps as the depth and liquidity of US markets causes the two sets of instruments to trade largely on a “no-arbitrage” basis.

“The US dollar market is very big, actively traded and efficient. Futures and swaps in the dollar market trade nearly all the time on a no-arbitrage basis,” says Mirza. “The interest rate information should be the same in both markets, so it will just be added volume with the calculation not getting impacted in most cases.”

Combining both futures and OIS into a single benchmark may prove tricky, some participants say.

CME’s rate is built on a full trading day and published with a one-day lag. Trading is divided into 14 observation intervals of 30 minutes each, with a volume-weighted average calculated for each observation. These are optimised and aggregated to calculate the final rates.

Swap-based term rates, such as those published by Ice Benchmark Administration and Refinitiv, typically take shorter snapshots of committed swap quotes on electronic central limit order books during morning trading for same-day publication. Other data sources, such as negotiable dealer-to-client quotes from Tradeweb and listed futures, provide additional support in a waterfall of fallbacks.

Mirza says CME is taking a different approach and will incorporate transaction data from dealer-to-dealer and dealer-to-client venues, with a focus on real transactions rather than quotes.

“The distinguishing feature of our methodology is that it is entirely based on transactions,” says Mirza. “Each instrument’s contribution is weighted by transaction volumes. It is the key in creating a transaction-based interest rate benchmark. If there is $100 billion of OIS swaps that go into the inputs, that will be 10% of our current transaction volumes.”

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact or view our subscription options here:

You are currently unable to copy this content. Please contact to find out more.

You need to sign in to use this feature. If you don’t have a account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here