I am grateful to Alastair Hughes for his assistance in drafting these remarks and Paul Brione, Andrew Hauser, Thomas Horn, Will Parry and Edwin Schooling Latter for their input.
## Libor: Entering the Endgame
Speech given by
Andrew Bailey
Governor of the Bank of England
Bloomberg Webinar, London and New York
13 July 2020
1
## Introduction
''The passing of summer,' I thought. 'There can be no summer in this land without cricket.'' So wrote the journalist and author Neville Cardus. 1 Covid has deprived us of many aspects of our communal lives this year. It has meant no cricket for us until last week, and in the US still no baseball. And to prove that we share an obsession with summer pastimes, a former colleague from many years ago when I was seconded to the New York Fed gave me a book to read called 'How Life Imitates the World Series'.
Sadly, my speech cannot match that for importance. But a more recent summer tradition has been speaking on Libor, a now annual event for the past few years - but one I do not expect to be a permanent fixture.
The time is fast approaching to draw stumps on the Libor benchmark, or for those to whom cricket is a mystery - after all not everyone appreciates a sport that can last five days and still end in a draw - we will soon be at the bottom of the ninth for Libor. Though from experience, I know the bottom of the ninth can last for eternity in a close baseball game.
## The continued importance of the transition away from Libor
Public authorities and market participants across jurisdictions have now been working together to transition away from reliance on Libor for a number of years. Its weaknesses were originally exposed during the global financial crisis and while reforms have improved the governance and administration of the rate, the markets on which it is based have long since shrunk to a trickle of transactions. We see little prospect of those markets returning.
Plans now need to be in place to transition from Libor to alternative reference rates by end 2021. John and I will speak about the progress that has been made. With 18 months to go these plans must now be acted upon in the time remaining.
With the authorities, businesses and households all having to adapt to the challenges of lockdown and new ways of working, there have been calls to step back from the Libor transition as a priority. We understand the level of disruption there has been and central banks have worked to ensure financial institutions have had the support and capacity to be able to ensure a continued flow of financing to support the economy. But in my view, what we saw in financial markets in March in response to the shock of Covid only reinforces the importance of removing the financial system's dependence on Libor in a timely way.
1 'Cardus on Cricket: A Selection from the Cricket Writing of Sir Neville Cardus' Souvenir Press
2
Fundamentally, this is due to the fact that Libor is trying to measure a market - the market for unsecured wholesale term lending to banks - that is no longer sufficiently active. The low levels of underlying activity make it fragile and more susceptible to liquidity and amplification effects in financial markets.
In the three-month sterling Libor rate, which is commonly referenced in many loan agreements in the UK, the proportion of 'transaction based' submissions only rose above 10% of relevant data supporting that rate once in the year up to March. 2 In the week of 16 March this year, as central bank policy rates were reduced to historically low levels, making cheaper funding available to banks, 'transaction based' submissions in three-month sterling Libor fell to zero. 3 Over half of the 35 published Libor rates across all currencies contained no transaction based submissions at all during that same week. At the same time Libor rates, and therefore costs for borrowers, spiked upwards based on firms' expert judgement.
By contrast, the value of transactions underpinning SONIA, the chosen alternative in sterling markets, increased from an average of around £40 billion per day to over £60 billion in April.
Now let me emphasise, as I have said before, we have no reason to believe there is misconduct related to the Libor rates today. However, the low level of transactions and heavy reliance on expert judgement does impact its robustness and sustainability. These are rates which directly impact the cash flows and values of an estimated $400trn of financial products globally. While most of this relates to derivative market activity, this is not an issue limited to professional investors - it impacts every company and household with a Libor linked product.
## The necessity for change
So the paradox of Libor is that if we are looking for a robust way to create transparency on bank funding costs, Libor is not that rate. Because the market it measures is a small to disappearing part of overall bank funding, it can no longer claim to accurately reflect the marginal cost of funds for banks, nor to provide end users with confidence their interest payments are directly linked to those costs.
It is a common human trait to state proudly 'I'm sticking to what I know', as though familiarity can itself outweigh the benefits of change or hide the weaknesses of the status quo. However, the market volatility we saw during March and April only served to underscore the long-standing weaknesses of the Libor
2 Libor submissions are classified in three levels. Level 1 is 'transaction-based' submissions - an average of transactions in unsecured deposits and primary issuances of commercial paper and certificates of deposit. Level 2 is 'transaction-derived' data, including information from historical transactions. Level 3 is 'expert judgement' - where a panel bank has insufficient Level 1 or 2 transactions, it estimates the rate at which it could fund itself in the unsecured wholesale funding market, based on an approved procedure. Further details are available at: www.theice.com/publicdocs/ICE\_LIBOR\_Methodology.pdf
3 Interim Financial Stability Report, May 2020 - Box 1 The continued importance of the transition away from Libor, https://www.bankofengland.co.uk/-/media/boe/files/financial-stability-report/2020/may-2020.pdf
3
benchmark. Libor rates - and hence costs for borrowers - rose as central bank policy rates fell to support the economy. 4
It is logical that credit spreads on bank debt (a measure of perceived risks in lending to banks) should increase during periods of stress. It is also unsurprising for banks to be attracted to the idea of being able to pass that cost directly onto their customers, but it does raise an important question:
Given this cost is passed directly onto borrowers through the interest rates they pay, how well does that rate actually reflect the cost of bank funding?
There have been significant changes to the way banks fund themselves since the global financial crisis, making them more robust. Regulatory changes have significantly reduced dependence on the fickle shortterm unsecured markets that Libor measures. 5 We have seen the benefits of those reforms through the first half of this year as banks have continued to lend and markets have remained open. And central bank facilities such as the Bank of England's Term Funding Scheme have offered a cost effective source of funding to support additional lending to the real economy, providing insurance against adverse conditions in bank funding markets.
These changes have led many banks to choose simply not to borrow in unsecured markets as the costs increase. Overall, this leaves Libor rates capturing (at best) only a tiny fraction of overall funding costs even in benign market conditions. As transactions become scarce during periods of stress, this reduces yet further. For example, during March and April Libor rates were impacted by liquidity mismatches in money market activity. Money Market Funds, that are important investors in unsecured bank funding markets, experienced large outflows, the so called 'dash for cash'. This meant Libor linked borrowers were immediately exposed to a sharp increase in rates driven by these more esoteric liquidity factors in addition to any perceived changes in bank credit risk.
When the banks that provide the data on which the Libor rates are based can't reference actual transactions that they have undertaken, they are asked to provide Libor's administrator with a submission based on transactions in other markets, where they are able. More commonly though, those data aren't available either, so they have to rely on 'expert judgement' which is an estimate of the rates that the firm would expect to pay in those wholesale unsecured funding markets.
It is simply not desirable for spikes in Libor rates, which banks are largely insulated from in terms of their own cost of funding, to nevertheless result in those costs being passed directly onto businesses and households.
4 Interim Financial Stability Report, May 2020 - Box 1 The continued importance of the transition away from Libor, https://www.bankofengland.co.uk/-/media/boe/files/financial-stability-report/2020/may-2020.pdf
5 For example see How Correlated is Libor with bank Funding Costs? https://www.federalreserve.gov/econres/notes/feds-notes/howcorrelated-is-libor-with-bank-funding-costs-20200629.htm
4
The factors impacting Libor can be far removed from the risks reflected in the borrower's own credit worthiness and can introduce significant, and for many impenetrable, volatility in interest rates. Financial markets should not amplify these risks and pass them through to the real economy.
## So what's the alternative?
So if Libor isn't a robust and transparent basis for determining funding costs for firms and households - is it possible to create such a benchmark that is sufficiently robust?
Well, some years ago now, working groups made up of industry participants in Libor jurisdictions recommended near risk-free rates as the alternative. And more recently we have seen the market develop products using these rates. It is a fact that homes and businesses with loans linked to these rates have seen greater direct benefit from the action taken by central banks during the Covid disruption and have not been as exposed to movements in Libor rates impacted by the scarcity of underlying transactions. Many more deserve the same protection.
There is no one size fits all approach, and while there is a common view across jurisdictions that liquidity should move to overnight near risk-free rates where appropriate, this is not to exclude the use of other robust benchmarks.
There is a workstream under the US working group - the ARRC 6 - that is currently looking at the potential to develop alternative dollar reference rates that include a credit-sensitive spread element, perhaps as an alternative or add-on to the new near risk-free rate across a subset of products.
But market demand for credit-sensitive rates that seek to reflect changes in bank funding costs and directly pass these on to borrowers cannot come at the expense of ensuring those rates are sufficiently robust. Introducing new approaches that simply replicate the inherent weaknesses of Libor would be a failure.
Would-be producers of such rates face the challenge of doing so in markets where liquidity may be thin or variable. Those choosing to use such rates would need to understand the exposure they were taking on to the liquidity conditions in those markets, and the volatility in the benchmark rate this could cause. We will continue to work domestically and internationally to prevent financial markets being materially exposed to risks from fragile benchmarks in terms of both the integrity of those rates and ensuring continuity of provision.
6 The Alternative Reference Rate Committee is a group of private market participants convened by the Federal Reserve Board and New York fed to help ensure transition from U.S. dollar Libor.
5
## Adapting to the impact of Covid
Despite Covid, work on transition in sterling markets has continued. Indeed, there has been considerable progress since John and I spoke on this topic last year - both here in the UK and the US.
We have not been blind to the disruption to businesses and the challenges for banks in communicating with clients caused by Covid. In co-operation with the Working Group on Sterling Risk-Free Rates, in the UK we have worked to change the phasing of some of the key milestones that were put in place for transition at the start of this year.
Communication and education for those households and businesses exposed to Libor is a huge challenge. The early focus has rightly been on raising awareness across those in the financial industry that generate Libor linked exposure that will need to support their clients through transition. Increasingly though, regulators' and markets' efforts need to turn to much broader communication to ensure the end users of Libor understand the weaknesses with the rates, why they need to move to alternatives and what the timeline is for the action they will need to take to prepare for that move.
This remains a difficult balance given the challenges faced by many. We are working to make this process as accessible and straight forward as possible, cooperating with bodies such as the Association of Corporate Treasurers and the Confederation of British Industry. And from next month, as previously announced, the Bank of England will begin publication of a freely available compounded SONIA index to support use of the rate across a wide range of sterling products.
Avoiding the problem won't make it smaller or change the need to transition. The answer is for us collectively to confront the challenge and make the necessary progress over the next 18 months. In September, the Working Group on Sterling Risk-Free Reference Rates will begin a further program of public communications to help those with Libor linked exposures navigate onto more robust alternatives.
But well done is better than well said, and from October UK banks should all be offering alternatives to Libor. Indeed, a number of the largest UK lenders are already doing this. I'd suggest if you are borrowing past the end of 2021 you consider seriously the greater certainty that those alternatives offer. If you do decide to borrow linked to Libor then your bank should discuss how that contract will change from Libor to an alternative rate ahead of the end of 2021. We would not expect to see any further sterling Libor linked lending after the end of March 2021. Regulated firms in the UK should expect their supervisors to monitor and discuss their progress with these important milestones.
6
## The state of play in sterling so far
I've highlighted the problems with continuing to rely on Libor, but the positive side of this is the progress that continues to be made in creating deep and liquid markets in robust alternatives.
In sterling, the chosen alternative is SONIA, produced by the Bank of England. The market for SONIA linked derivatives is well established. The share of swaps traded using SONIA is broadly equivalent to that linked to Libor, with continued growth at longer maturities. Volumes in Libor linked options are decreasing and there are promising signs of a SONIA options market developing; we have also seen growth in SONIA futures during the first half of 2020.
There is now a fully functioning SONIA linked bond market with 40 Floating Rate Notes issued, totalling over £16bn so far this year. In total, there has now been issuance of over £185bn of bonds referencing compounded SONIA. There has also been in excess of £20bn of publicly distributed SONIA linked securitisations since April 2019. We have seen businesses such as Shell, South West Water, Associated British Ports, National Express Group and Riverside Group all move borrowing on to near risk-free rates. March saw the first multi-currency revolving credit facility linked to both SONIA and SOFR. The industry is making progress on the conventions underlying these products and infrastructure vendors are ensuring their products can support SONIA linked products.
## The beginning of the end
I will now turn to legacy business and what has come to be referred to as the 'endgame'.
After extensive consultation by ISDA across markets we will shortly have a robust and trusted fallback for trillions of dollars of derivative contracts. We encourage early sign-up to the protocol, but that is not a substitute for continuing to move business onto near risk-free rates. We expect to see more of the sterling swap market to move towards SONIA in the UK given the high likelihood that contracts will be linked to this rate after 2021. We are working with the market on fallbacks for cash instruments, looking to ensure hedge positions convert at the same time as underlying exposures, to minimise the introduction of any new basis risk.
We have continued to see firms start to convert Libor linked instruments to alternative rates, for example through consent solicitation processes where over £10bn of legacy Floating Rate Notes, securitisations and covered bonds have converted to SONIA.
We have been clear and consistent in the importance of transitioning away from Libor and that this transition should be a market driven, but we have also acknowledged the authorities have an important role to play. We remain committed about both the need for the market to continue to drive transition forward moving new
7
issuance onto alternative rates and converting legacy exposure where possible; and also to ensuring there is a solution for those contracts that legitimately cannot be converted from Libor.
That is why the UK government announced last month its intention to legislate to provide the FCA with increased powers to deliver an orderly wind-down of critical benchmarks, such as Libor. 7 It remains in the interests of financial markets and their customers that the pool of contracts referencing LIBOR is shrunk to an irreducible minimum ahead of LIBOR's expected cessation, leaving behind only those contracts that genuinely have no or inappropriate alternatives and no realistic ability to be renegotiated or amended. Those that decide to rely on regulatory action, enabled by the legislation the UK Government plans to bring forward, will not have control over the economic terms of that action. The FCA will consult on what that action may look like and what the limitations of any solution may be for the 'tough legacy', but it is important to note:
- Any solution will only be designed for legacy contracts, it will not be available for new business, so this not an alternative to moving new activity away from LIBOR as soon as is prudent;
- (as the FCA have said), use of the powers will not provide a solution in all circumstances, in particular if the relevant alternative inputs are not available to the benchmark administrator. Given the challenges I have already covered in creating a robust dynamic credit spread, the market would need to be prepared for a fixed proxy based on historical averages; and
- neither authorities, nor market participants in ISDA's consultation, or in bond markets as revealed by practice, want forward looking term risk-free rates to be the basis of derivative markets. It is very important, if these tools are to be used without unwanted side effects, to see derivative markets move away first, either organically, or through the ISDA protocol.
The message remains clear: those who can transition away from LIBOR should do so on terms that they themselves agree with their counterparties.
## Conclusion
Libor transition is challenging. It has been embedded in the financial system for a number of decades across a broad range of products.
But we are moving away from Libor and the problems of Libor. This means facing up to transition now, taking action to ensure new issuance moves to robust alternative rates and concrete plans are in place to deal with
7 https://www.parliament.uk/business/publications/written-questions-answers-statements/written-statement/Commons/2020-0623/HCWS307/
8
legacy exposures. The UK Government announcement on its intention to legislate and the finalisation of the ISDA protocol are important steps in the 'endgame' becoming clearer.
There is more to do and work is underway to remove the remaining barriers to transition. After John has spoken we'll be joined by the chairs of the UK and US industry groups who continue to lead this work, building consensus on key areas such as product conventions and cash fallback mechanisms. Hopefully, the message is now clear to those who have remained on the side-line: this is a necessity not a choice and there are eighteen months left to transition.
And, for the avoidance of doubt, I'm taking it that eighteen months in Libor is past the middle of the seventh, so you don't need to get up now, stretch and sing 'Take Me Out to the Ball Game' before John speaks.
9