Speech by Edwin Schooling Latter, Director Markets and Wholesale Policy at the FCA, delivered at City & Financial's Managing LIBOR transition event on 26 January 2021.
Speaker: Edwin Schooling Latter, Director of Markets and Wholesale Policy
Location: City & Financial's Managing LIBOR transition event
Delivered on: 26 January 2021
Highlights:
- 85% of uncleared UK derivatives market ready for the end of LIBOR as 12,500 firms sign the ISDA protocol
- IBA consultation on proposed end-dates for LIBOR has now closed, opening the way to determining and announcing the future path for all 5 LIBOR currencies simultaneously
- users of LIBOR should press ahead with transition plans – in their new business and their legacy LIBOR books
Note: this is the speech as drafted and may differ from delivered version.
In July 2017 the FCA made public its plans to ensure the London InterBank Offered Rate (LIBOR) was maintained until the end of 2021, but not to use its regulatory muscle to prolong LIBOR’s life beyond that point. We set this end-2021 date to give sufficient time for a complex transition that would need to span multiple markets and multiple jurisdictions.
Here in the UK, with the LIBOR panel banks, and with the Sterling Risk-Free-Rate Working Group, we and the Bank of England in effect set up a public-private sector partnership to facilitate and deliver this transition. Similar working groups were set up in the other major LIBOR jurisdictions.
End-2021 seemed a long time in the future. Though we knew from the beginning that a transition of this size would be challenging.
Now – it’s less than 12 months away.
In December last year, LIBOR’s administrator, ICE Benchmark Administration (IBA), set out for consultation its proposal for an end to all 35 of the LIBOR currency-tenor settings. 30 of them, including all the sterling, yen, Swiss franc and euro settings would cease at end-2021. In its consultation, IBA proposed that the remaining 5 US dollar tenors would continue to be published for a further 1 and a half years, before also coming to an end. In some ways that difference in timing overstates the difference in the proposed future path for US dollar LIBOR. As the US authorities have made clear, the 5 US dollar LIBOR settings proposed to continue beyond end-2021 would be for use in legacy but not new transactions after end-2021.
Targets to stop new LIBOR lending, deliberately set to be achieved comfortably before the proposed end date, will bite sooner than end-2021. In the case of sterling, for example, the date to stop new LIBOR lending is less than 10 weeks away – at end Q1 this year.
The good news is that the industry – at least most of it – has risen to the challenge of this transition. Regulators, supervisors and public authorities across many jurisdictions have worked to support those industry efforts through co-ordination, guidance and indeed legislative proposals.
That progress made on transition is of course interlinked with IBA’s December proposals on cessation dates, as panel banks, themselves among the biggest users of LIBOR, have become increasingly confident that end-December 2021 is an appropriate date to draw their participation in 4 of the 5 panels to a close, and end-June 2023 appropriate for the fifth.
So I hope today’s audience does not include any market participants who haven’t yet realised the importance of ensuring they are ready for life without LIBOR from end-2021.
I’d like to use my speech today to flag firstly some of the major milestones that have been achieved in the past weeks, and secondly, what you can expect next.
ISDA protocol
Let me start with yesterday’s milestone. Monday 25 January was a landmark day in the LIBOR journey. It was not only the day that IBA’s consultation on its cessation proposals closed. It is also the day that new International Swaps and Derivatives Association (ISDA) language on interbank offered rate (IBOR) fallbacks became effective. These now apply as the standard arrangement in all new ISDA interest rate derivatives referencing LIBOR. They also take effect in all outstanding covered ISDA contracts where both parties have signed ISDA’s protocol. They mean these contracts will convert from LIBOR to a fallback rate based on the chosen risk-free rates (RFRs), ie Secured Overnight Financing Rate (SOFR), Sterling Overnight Index Average (SONIA), Tokyo Overnight Average Rate (TONAR), Swiss Average Rate Overnight (SARON) and Euro Short-Term Rate (€STR), at the point LIBOR ceases or becomes 'unrepresentative'.
I have talked previously about the importance of putting in place fallback arrangements for LIBOR’s cessation (or loss of representativeness) in derivative contracts, by signing the protocol, or, if not taking that easy route, bilaterally negotiating fallback arrangements, or simply removing your exposure to LIBOR-referencing financial instruments. For many firms, it is a regulatory obligation to have fallback arrangements.
As of this morning over 12,500 firms had signed the protocol.
Firmly in the lead on a jurisdictional basis is the United States, accounting for a little over 10,000 of those signatories. In second place is the UK, with over 450. Also well into treble figures are Singapore (234), Thailand (191), Japan (149), India (147) and Switzerland, with 136. I can’t see any major jurisdiction missing from the list. For example we counted 115 signatories from Canada, 57 from France, 34 from Germany, and 22 from Italy.
The really important point for firms who have not been so quick off the mark, is this. The protocol is still open for signature. The importance of signing the protocol is unchanged.
But let me bring out more clearly how important this protocol is to the overall LIBOR transition.
Reports describing the challenge of LIBOR transition have referred to several hundred trillion dollars in LIBOR-referencing contracts. Our current estimate is US$260 trillion. About 80% of this total is accounted for by cleared interest rate swaps and exchange-traded derivatives. Cleared swaps will have, in effect, the same triggers and fallback arrangements as in the ISDA protocol. So in economic terms, they will become RFR-based contracts from the point LIBOR ceases or becomes unrepresentative. Moreover, central counterparties like LCH may seek to convert them somewhat earlier than this through rulebook-based processes or supported by multilateral conversion exercises.
A further 9% or so of this total is accounted for by uncleared derivatives, most of them covered by ISDA language.
For example, the total amount of uncleared interest rate swaps referencing sterling LIBOR is around £4 trillion. Thanks to the reporting required under the European Market Infrastructure Regulation (EMIR) we have a good view of the counterparties to these uncleared derivatives. Based on who has signed the protocol already, we estimate just over 85% of, for example, uncleared sterling LIBOR-referencing swaps now have effective fallbacks in place because of dual-sided adherence. 99.7% have at least one-sided adherence.
Adding that 85% coverage to cleared swaps and futures means 97% of sterling interest rate derivatives are covered by fallbacks. If you assume similar levels of protocol coverage across the other LIBOR currencies, and adjust for contracts maturing naturally, it takes out, by the time the relevant LIBOR panels are proposed to cease, around US$245 trillion from that US$ 260 trillion total of outstanding LIBOR-linked contracts.
We should not – and do not – underestimate the challenges associated with managing other LIBOR outstandings, notably in bond, securitisation, loan and mortgage markets. But, clearly, the success of this protocol means a major reduction in risk faced by the 12,000 plus firms who have already signed, and for the system as a whole.
If you have not signed it, and your LIBOR derivative positions are still exposed to risk of LIBOR ceasing or becoming unrepresentative, it is not too late to do so.
What happens next?
As you know, IBA’s consultation on cessation dates for the 35 LIBOR currency tenor settings closed yesterday.
I have no doubt that IBA will move with its usual speed and diligence to assess the feedback to the consultation, complete the necessary internal governance and then advise the FCA of its determinations on how it intends to proceed, having taken those responses into account.
Notwithstanding the difference in proposed end dates for US dollar LIBOR, the single consultation covering all 5 LIBOR currencies opens the way to determining and making announcements on the future path for all five simultaneously. Consistent with the intent of the provisions and processes set out in the Benchmarks Regulation, giving clarity to the market on all 35 settings as soon as practicable helps maximise time and prospects for an orderly transition.
The FCA made clear back in 2017 that it did not intend to compel panel bank participation beyond end-2021. In approximate parallel to IBA’s consultation, the FCA consulted on a framework under which, and the circumstances in which, we might consider it desirable and feasible to require continued publication of any LIBOR currency tenors on the basis of a changed methodology. This is what the market generally refers to as requiring publication of a ‘synthetic’ LIBOR. As set out in that consultation, the changed methodology we have tabled is also based on the RFRs, combined with a fixed spread that is identical to the spread in ISDA’s fallbacks.
We said that the implications of the proposed framework and methodology for synthetic LIBOR were that we didn’t think it would be desirable and feasible to compel continuation on a synthetic basis for euro or Swiss franc LIBOR, or indeed lesser-used tenors in any currencies. We said it did appear, in contrast, to be appropriate for the more commonly used sterling settings, and we would continue to assess whether this might also be the case for more commonly used yen and US dollar settings.
It is important to understand how IBA’s determinations following its consultation, and the policy framework that FCA has proposed for future decisions on whether to require publication of a synthetic LIBOR, could interact.
If IBA confirms to the FCA that following its consultation it intends to cease LIBOR settings, and the FCA is satisfied that the benchmark can be ceased in an orderly fashion, and the FCA confirms a policy that would not envisage compelling continued production on a changed methodology basis of a relevant setting, then we could announce that these settings will cease.
Another possibility for some settings is that it is clear the panel will end, but we do envisage consulting on requiring continued publication on a synthetic basis under proposed new powers set out in the Financial Services Bill, or are continuing to assess whether to do so. In that case, it would be clear that the rate will no longer be representative beyond the relevant panel end date. The only way it could continue would be on an unrepresentative, synthetic basis. For these settings we could make a so-called 'pre-cessation announcement' in terms of ISDA documentation.
I can’t give you a specific date for when these post-consultation announcements will be made. And I am not today pre-judging IBA’s consultation or precisely what FCA or IBA announcements will be. But we see no case for delaying decisions or announcements beyond the time necessary properly to assess the consultation responses that have now been received. Whether the relevant currency-tenor setting is subject to a cessation announcement, a pre-cessation announcement, or indeed both, announcements covering all settings could be made on the same day. Spread calculations would therefore be fixed on the same day. Of course, these spreads would not actually be applied until after the last date of proposed panel bank publication, as the LIBOR rates are expected to remain representative until that point.
The setting of the spread is itself an important landmark that facilitates transition. It means that LIBOR contracts covered by fallbacks that use that spread, or by a synthetic LIBOR rate, where we’ve also suggested we would use the same spread, become, in economic terms, SONIA, SOFR, TONAR contracts from the date publication of a representative LIBOR ceases. It means that those managing conduct risk in transition can be confident that planning transition using RFRs plus those spreads, from the point that panel bank LIBOR publication ceases, will not result in a worse expected economic outcome for one party than waiting or hoping for a synthetic LIBOR. If you haven’t seen them already, the conduct Q&As on our website address this and other conduct issues.
Key policy decisions ahead
Where we do envisage requiring continued publication of a LIBOR setting on a synthetic basis, this will be subject to a further consultation and decision on the specific settings. If IBA confirms it has determined to bring the relevant currency panel to a stop at end-2021, we would want to launch such consultations in a timely way so as to minimise unnecessary uncertainty. Such consultations could therefore be the next step after relevant pre-cessation announcements. Decisions could then be made once the new powers were on the statute book – perhaps in summer 2021 subject to Parliamentary approvals – so that markets had clarity as early as possible.
But a key point to understand is that synthetic LIBOR is not for use in new contracts. It is intended to help the problem of genuine tough legacy contracts, i.e. those that it is not practicable to convert before publication of a representative LIBOR ceases.
Even for the US dollar LIBOR settings which IBA has proposed to cease later than the others, prospectively continuing to publish on a representative panel bank basis until end-June 2023, there will be restrictions on new use after end-2021. US regulators have already set out limitations – ie to use only in legacy contracts, and defined categories of risk-management transactions used to manage that legacy exposure.
In the UK, the Financial Services Bill – Article 21A of the revised Benchmark Regulation – proposes FCA powers to restrict use of a critical benchmark when we know it is going to cease, as would be the case for US dollar LIBOR under IBA’s proposals. That can help reduce risk of misuse, or unnecessary risk to market integrity from the creation of new exposures.
The Bill also proposes powers for the FCA to define which legacy contracts will be allowed to use synthetic LIBOR – Article 23C – to help an orderly wind down.
So, expect consultation proposals from us in the spring on a framework for using both powers. How we decide to do so, drawing on input from market participants, will be important to managing LIBOR’s wind down successfully. Our decisions will be subject to consultation. But it may be helpful if I give a very high level indication now of ideas we are likely to consult on.
In terms of restricting new use of a benchmark that we know will cease, and taking the hypothesis that this will be relevant to use by UK regulated firms of US dollar LIBOR after end-2021, then we will probably consult on taking a similar approach to that already set out by US regulators – ie allowing new use only in risk-management transactions. We see merit in co-ordinating with other overseas authorities too. Use of US dollar LIBOR is, after all, international.
In terms of what legacy use of synthetic LIBOR to allow in the UK, we will need to strike a careful balance in terms of where this is necessary and desirable.
At one end of the spectrum, for centrally cleared LIBOR derivatives, there seems to be no need or material benefit to allow use of synthetic LIBOR. Central counterparties have rules that allow them to change LIBOR contracts to work on the better alternative of compounded RFRs plus relevant spreads.
At the other end of that spectrum are contracts which fall unambiguously into the 'tough legacy' bucket. For example, take a retail mortgage where the lender must have borrower consent to change, but can’t get the borrower to respond to change proposals. The best way to protect this consumer may be for the mortgage to continue using a synthetic LIBOR that offers a fair approximation of what panel bank LIBOR might have been, rather than find their mortgage contract has become inoperable. Or take a bond where the issuer offers conversion to compounded SONIA plus a credit adjustment spread calculated on the same basis as in ISDA documentation, in line with market consensus on a fair fallback, and the bondholders don’t reply or withhold their consent in an effort to push for terms that are out of line with these market standards. Use of synthetic LIBOR may help protect market integrity by ensuring these securities continue to function in the way intended.
But where parties can practicably agree to convert on the fair terms that have now become standard across derivatives, securities and loan markets, they should do so. This takes away the uncertainty about whether the FCA will or won’t conclude it should require publication of a synthetic LIBOR. It takes away the uncertainty of how long that synthetic LIBOR will continue. It enables conversion to a stronger and more liquid underlying benchmark – the compounded RFR – which synthetic LIBOR will not. It will allow you and your counterparties to keep control over the economics of your contracts.
Concluding remarks
I hope the conclusion of these remarks is already obvious. The need to transition is clear. The economic terms of fair transition have been worked out, they stretch across markets and across jurisdictions. Fair spreads for conversion will be locked in when cessation and pre-cessation announcements are made. Press on with your transition.