So long LIBOR – 3 weeks to go

Speaker: Edwin Schooling Latter, Director of Markets and Wholesale Policy and Wholesale Supervision 
Location: Risk.net’s LIBOR telethon
Delivered on: 8 December 2021
Note: this is the speech as drafted and may differ from the delivered version

Highlights

• Sterling, Swiss franc, Japanese yen and euro LIBOR panels come to an end on 31 December this year. 
• Sterling interest rate markets have moved to SONIA, and most sterling legacy LIBOR contracts will have converted away from LIBOR by or at year end. 
• Some firms will need to take further steps to convert remaining legacy LIBOR contracts in 2022.

Panel bank sterling LIBOR, formalised 45 years ago, but with its origins going back to the 1960s, will make a final appearance on New Year’s Eve this year. So too will Swiss franc LIBOR, euro LIBOR and the panel bank rate for Japanese yen. New use of US dollar LIBOR is also to cease, a key landmark on the end to the US dollar panel just 18 months later.  

10s of thousands of firms have been preparing for this, with the encouragement, support – and occasional nagging – of authorities here and overseas. We have had the common aim of minimising disruption from the retiring of a benchmark rate referenced, even at the beginning of this year, in an estimated US$ 265 trillion of contracts.  

The largest slice of that is of course in interest rate derivatives, but it also includes trillion dollar sums of bonds and securitisations, loans, mortgages and other products.

Are markets ready? 

Data are more telling than a regulator’s words, so I will paint the picture in numbers. 

In the sterling swaps market, 89% of volume traded in November referenced SONIA. By end-November, cleared SONIA swaps outstanding, at 59% of the total, exceeded cleared LIBOR swaps outstanding, at 41%. In fact, cleared LIBOR notional outstanding more than halved between February 2021 and November 2021. The remaining cleared sterling LIBOR swaps outstanding are due to disappear over the weekend of 18/19 December, as LCH converts them all to SONIA. We estimate £3.1 trillion – 91% – of the remaining uncleared sterling LIBOR swaps outstanding will also convert to SONIA at year end, thanks to the application of ISDA’s fallbacks. 

In sterling futures, 58% of volume traded last week referenced SONIA. 42% referenced LIBOR. By end-last week, cleared SONIA outstanding had grown to just under 40% of the total, having climbed consistently week-by-week since 'SONIA first' for exchange traded derivatives in June 2021. When the December short sterling contract expires on 15 December, LIBOR’s share of outstanding futures will fall further. All the rest is due to disappear – converted to SONIA – on Sunday 19 December as ICE Clear Europe completes its conversion.  

As for bond markets, new issues referencing sterling LIBOR had already diminished to a trickle by autumn 2019, and to zero since October 2020. The greater challenge has been legacy bonds. But, by end-November, roughly two thirds of the sterling LIBOR bonds maturing after end-2021, by value, had been converted to SONIA.   

The commercial loans market was the last sterling market to move decisively from LIBOR to SONIA. But data collected from the banks show few cases of new use of Sterling LIBOR after end-March 2021. SONIA-linked lending has quickly surpassed £100 billion across a wide range of facilities being used by many types of businesses. The data we receive on progress on conversion of legacy loans are lagged, and collected through supervisory channels rather than automated reporting. So, we don’t have close to real time data in the way we have for derivatives and securities. But our supervisory information shows very substantial progress in converting outstanding sterling LIBOR loans to SONIA.  

In the UK mortgage market, there was already only a relatively small pocket of legacy LIBOR loans – about 1 in 100 residential mortgages, and 1 in 20 buy-to-let mortgages: around 200,000 contracts in total. Based on a membership survey conducted by UK Finance earlier this year, around two thirds of responding firms that hold sterling LIBOR mortgages were confident their entire legacy books would be transitioned to alternative rates – SONIA or Bank Rate – by end-2021, in both the retail and buy-to-let space.  

In other LIBOR currencies too, the change has been remarkable. ISDA’s weekly analysis based on DTCC figures for week ending 3 December showed SARON swaps volume accounting for 96% of the Swiss franc total. TONA swaps were 94% of Japanese yen totals. Last weekend, all outstanding Swiss franc, Japanese yen and euro LIBOR swaps at LCH were successfully converted to SARON, TONA and €STR – removing US$6 trillion of outstanding LIBOR notional in one strike.  

At some points in the past couple of years, we did have worries that the pace of transition in US dollar markets was slower than we hoped. But there too, as we approach year end, the change is dramatic. In the last week of November, 33% of new LCH-cleared US dollar swaps referenced SOFR.

All of those numbers are now better than I would dare to have predicted a year ago. They are a testament to the work that so many firms have done.  

In September we saw a successful international 'RFR first' initiative in the interbank cross-currency swap market. The market switched seamlessly from USD LIBOR versus sterling, Swiss franc and yen LIBOR to SOFR versus SONIA, SARON and TONA. Since then the interbank market has also migrated to SOFR versus €STR. Last week the CFTC’s Market Risk Advisory Committee announced Part II of 'SOFR first' for cross-currency swaps.

This next key step encourages all other cross-currency swaps with a dollar leg to switch trading to SOFR from 13 December, ahead of the end of year restriction on new use of USD LIBOR. Our March 2021 Dear CEO letter outlines our expectations for regulated firms to meet milestones and targets of relevant supervisory authorities. We encourage firms to take note and act accordingly.

But what about the firms who have been slow with their transition – is it now too late? 

There are still steps you can take. I have 3 key actions to flag. 

First, if you have outstanding uncleared LIBOR swaps, and have not yet done so, it is not too late to sign the ISDA protocol. Assuming your counterparty has also signed – which is highly likely to be the case – this will ensure your outstanding uncleared LIBOR contracts are converted to the new market standards of overnight SONIA, SOFR and their peers when the relevant LIBOR panel rate ends. Close to 15,000 firms, and counting, have already signed. For sterling markets, where our data are most complete, we calculate this means 99.8% of contracts have one-sided adherence, and 91% now have two-sided. If you are in that remaining 9%, but want to join the pack, then you can do so by completing a few simple steps set out and accessed via www.isda.org/protocols[2]. The agreement to include the fallback language comes into effect immediately upon your adherence, and, assuming your counterparty has done the same, your contracts would reference the fallback rates on the first date after the end of this year when LIBOR in the relevant currency and tenor is not available or representative.  

Second, if you are an issuer of a LIBOR bond that’s not yet been converted, then you do still need to act. The good news is that the 90 successful consent solicitations to date should give you confidence that this process works for bonds issued in sterling, and prospectively other LIBOR currencies, under UK law. I am not suggesting you could complete that by this New Year’s Eve if you haven’t yet started. But the FCA decision to require publication of synthetic sterling LIBOR for the duration of 2022 gives you time to catch up. Choosing to convert to SONIA, over the relevant term, plus ISDA spreads would not change the expected value of your interest payments. SONIA and ISDA spreads are anyway the two components of synthetic LIBOR. But synthetic LIBOR will not last. SONIA will.  

Third, there may never be as many experts in LIBOR transition as there are right now. And awareness of the need to transition may also be at its peak. Synthetic LIBOR is a bridge to RFRs. It is not a permanent solution. So do not rely on synthetic LIBOR lasting. We strongly encourage firms to complete the transition while expertise and awareness are at their greatest.

Synthetic LIBOR is not forever 

The publication of the synthetic LIBOR rates for 1-month, 3-month, 6-month sterling and yen LIBOR settings, for the duration of 2022, coupled with wide permission to use it, should substantially mitigate the risk of widespread disruption to the significant number of legacy LIBOR contracts which have not converted by or at end-2021. But synthetic LIBOR is temporary. We have already made clear that synthetic yen LIBOR will be for one year only, ceasing at end-2022.

We have not yet decided whether we should extend the sterling synthetic rates beyond end-2022. Based on the data for this year, we concluded the case for continuing all three of these sterling tenors for 2022 was made. But it is worth noting that the case for 3-month sterling LIBOR was stronger than for 1-month and 6-month. When outstanding contracts that still reference a particular LIBOR setting have reduced significantly, it may no longer be proportionate for the FCA to require continued publication of that setting on a synthetic basis. All the synthetic rates will cease in due course, and none will be continued simply for the convenience of those who could take action to convert, but have not bothered to do so.

What if my contract still refers to one of the LIBOR settings that are due to stop at end-2021? 

24 of the current 35 LIBOR settings will stop altogether after publication on 31 December. Counterparties to contracts referencing these settings have had a long time to prepare for this, to convert, agree replacements, or ensure there are fallbacks that will work. This is not the first time that, generally less widely-used, LIBOR settings have ceased altogether. Hopefully it is a good sign that our FCA inboxes are not full of known problems that will emerge in January.  

There is one issue on which we have had incoming queries. This relates to so-called 'dealer polls'. These are fallback arrangements, designed in times rather different from today, and designed more for a temporary rather than a permanent interruption to LIBOR publication. In essence, they ask a counterparty or calculation agent to collect quotes for borrowing rates from a range of dealers, in lieu of LIBOR itself. They may be the main fallback, or a step in a waterfall of fallbacks.

As we noted 2 years ago, it seems rather optimistic to think these will work given that a principal reason for the end of LIBOR is banks’ unwillingness to continue to provide such quotes for LIBOR itself. And we are now hearing exactly that prediction from various market participants – ie they won’t work because quotes will not in practice be offered. Although we have asked for any information that suggests otherwise, we are not aware at this point of any firm that has confirmed a willingness to provide rates in response to such a poll after the relevant LIBOR setting is no longer published, other than where they have a contractual commitment to do so.   

We do not think it would be appropriate or reasonable for us to put regulatory pressure on firms to respond to such polls. We understand that this would create a variety of conduct and other risks.  

However, it is helpful if market participants are able to assess and conclude at appropriate speed whether such dealer polls will work, and thus whether they must proceed to the next step in the waterfall, or contact the counterparty to agree alternative arrangements. Banks that might receive such requests may wish to consider setting up a centralised point to receive and make clear if any response will be provided to such requests. They may wish to consider being clear in their client or other communications where they have policies to decline to respond.

We will of course be monitoring for evidence of unexpected problems emerging. Experience in the first quarter of 2022 may yield useful lessons for the future cessation of the remaining synthetic sterling and yen LIBOR rates, and the US dollar LIBOR rates, each of which will follow in due course. 

Together with the PRA, we will also be continuing to monitor UK-supervised firms’ wind-down of their legacy LIBOR books, the end to their new use of US dollar LIBOR, and their preparation for the end of the US dollar LIBOR panel. As we have noted before, firms shouldn’t be relying on a synthetic solution or the same legacy use permissions for a synthetic US dollar LIBOR as we have given for sterling and yen. We will stand ready to use our powers where it is feasible and desirable to do so, but firms should not plan on the basis of an assumption that this will be the case.

Adieu 

Goodbye then, panel bank sterling, Swiss franc, yen and euro LIBOR – after your final appearance just in time for those New Year’s Eve celebrations and parties. Many for whom LIBOR was previously something of a staple, have managed successfully to cut down, or cut out, their consumption of this particular financial system lubricant. While it is probably realistic to expect a bit of clearing up to be completed when the New Year dawns – notably of those legacy LIBOR contracts – the many years of preparations that industry has made will have much reduced the prospect of too big a hangover.