ICE swap rate adds safety net with Tradeweb quotes
This article originally appeared on Risk.net
A key benchmark used to price swaptions and some rate-linked structured products will move to a new methodology on May 29, after failing to print on more than a third of occasions this year.
The ICE swap rate, a daily measure of term-Ibor referencing swap rates from one to 30 years, vanished for several days in March as dealers pulled firm swap quotes from electronic order books – currently the only permissible inputs.
From Friday, ICE Benchmark Administration (IBA) will start using a waterfall methodology, with the addition of dealer-to-client quotes from Tradeweb as a safety net when executable prices from central limit order books (CLOBs) are unavailable for a currency or tenor pair.
“In normal times, this would definitely give enough data [to publish the ICE swap rate] and it might even cope with abnormal times,” says a benchmark specialist familiar with the methodology.
The period from January 2 to May 7 saw 1,914 of the scheduled 5,601 swap rates for euros, sterling and US dollars fail to publish. Test data from IBA suggests the addition of dealer-to-client quotes would have improved the publication rate for US dollar versions, which were the most acutely hit, from 22% to 97%, with only 51 incidences of non-publication, compared with 1,300 when relying on CLOB prices alone.
In the absence of CLOB prices and dealer-to-client quotes, a third layer of the waterfall methodology allows for the rate to be calculated using linear interpolation of daily rate moves in adjacent tenors. Interpolated rates can be calculated using CLOB or dealer-to-client prices.
The move to a waterfall methodology is being welcomed by swaptions traders, who struggled to value cash settled contracts when the ICE swap rate failed to publish in March.
“Particularly in the shorter end of sterling, we ended up with a fixing less often than we actually had one,” says a European swaptions trader at a US bank. “Nobody wants that ambiguity.”
Of the 1,320 sterling rates scheduled to publish during the January-to-May testing period, 246 failed to print – a 26% nonpublication rate. The breakdown saw rates vanish across all 15 tenors during the worst of the Covid-19 driven volatility in March.
When the rate fails to publish, most traders resort to dealer polling. Some use internal algorithms that correlate the rate to bonds or other fixed income markets, but these model-derived rates must then be agreed with counterparties.
“Most market participants do their best to get as flat as possible and completely insulated from any fixing, but in so much as we’re insulated, we usually find ourselves having to agree some kind of fixing,” says the swaptions trader.
Fixing problems
The inclusion of Tradeweb data takes the Ice swap rate into new territory. In contrast to fully firm and actionable CLOB prices, executable dealer-to-client swap quotes streamed on the platform are subject to last look, and therefore negotiable.
In both cases, IBA takes a snapshot of the last two minutes of trading, an hour before publication. Swap rates for euros, sterling and US dollars are published at 11am local time. Additional fixings for all 15 euro tenors are published at 12pm and the one-year dollar rate is fixed again at 3pm.
Analysis from IBA shows little difference in the average prices on CLOBs and Tradeweb’s dealer-to-client platform. For the US dollar 11am fixing, the gap was 0.07 basis points on average and 0.4bp at its widest point. The biggest difference was in the euro 12pm fixing, averaging 0.8bp with a maximum divergence of 1.8 basis points over the test period.
That’s good enough for some users of the rate. “As long as it’s representative, that’s all I’m really looking for,” says the swaptions trader.
The new waterfall methodology is the first major overhaul of the Ice swap rate since 2015, when it was known as the IsdaFix. Back then, the rate relied on submissions from panel banks, based on ‘expert judgement’. These were replaced by firm, executable prices from three swap execution facilities – BGC Trader, Trad-X and TP Icap’s i-Swap – to satisfy the benchmark principles of the International Organization of Securities Commissions.
However, these venues have failed to deliver the continuous liquidity required to calculate benchmark rates – especially for US dollar swaps, where a large portion of trades are executed bilaterally on request-for-quote platforms, rather than on exchange-like venues.
The first signs of problems emerged in February 2018, when dollar swap rates failed to publish following a spike in equity volatility. A benchmarks expert says dealer algorithms that rely on correlations were caught out by the wild swings in equity markets, causing them to stop quoting.
Since then, outages have intensified, prompting IBA to open a consultation on the inclusion of non-CLOB data in August 2019.
US dollar versions of the rate have continued to be sporadic even as volatility has calmed. On May 22, just two out of the 13 scheduled 11am dollar fixings published.
The benchmark firm is also developing a new swap rate referencing Sonia – Libor’s successor for sterling markets. Other versions for new risk-free rates, such as the secured overnight financing rate and euro short term rate are also in the works.