Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings

Proposed decision under Article 23D BMR for
6 sterling and yen LIBOR settings
Consultation Paper
CP21/19
June 2021

CP21/19 Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings Moving around this document Use your browser’s bookmarks and tools to navigate. To search on a PC use Ctrl+F or Command+F on MACs. How to respond We are asking for comments on this Consultation Paper (CP) by 27 August 2021. You can send them to us using the form on our website at:
www.fca.org.uk/cp21-19-response-form
Email:
cp21-19@fca.org.uk
Sign up for our news and publications alerts
See all our latest press releases, consultations and speeches.
Contents
1 Summary 3
2 Proposed exercise of the Article 23D(2) powers 8 3 Why we are proposing to use Article 23D powers in this way 9
Annex 1
Questions in this paper 26
Annex 2
Abbreviations used in this paper 27

CP21/19
Chapter 1
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings 1 Summary Why we are consulting
1.1 We are seeking views on our proposed decision to use our Article 23D(2) powers under the
Benchmarks Regulation (BMR) as amended by the Financial Services Act 2021 (the FS Act) to require the administrator of LIBOR, ICE Benchmark Administration (IBA), to change the way 1-month, 3-month and 6-month sterling and Japanese yen LIBOR settings (hereafter referred to as ‘the 6 LIBOR settings’) are determined beyond end-2021.
1.2 The FS Act received Royal Assent on 29 April 2021 and the relevant provisions for our
enhanced BMR powers are expected to be shortly brought into force (‘commenced’) by the Treasury regulations. The proposed approach described in this consultation is contingent on those powers being available to us. We will make decisions about the potential exercise of the new powers only when the relevant provisions have been commenced.
1.3 Our Article 23D(2) decision is also contingent on our other, future decisions under
the BMR. That is, it would apply if we decide, as we currently expect, to compel the continued publication of the 6 LIBOR settings by the administrator under Article 21(3) and to designate the 6 settings as Article 23A benchmarks. We are not consulting on our future decisions under Article 21(3) or Article 23A.
1.4 We propose to require IBA to publish these settings on a changed methodology (ie
‘synthetic’) basis in line with our proposed technical specification set out in Chapter 2, taking effect immediately after end-2021 when the relevant panel-bank LIBOR settings will cease and any designation under Article 23A would take effect.
1.5 The proposed approach to using our Article 23D(2) powers is intended to secure
an orderly cessation of LIBOR and to advance our objectives of ensuring consumer protection and the integrity of the financial system.
1.6 In March 2021, we announced the future cessation or loss of representativeness of
the 35 LIBOR settings. We said that we have decided not to require any panel banks to continue to submit to LIBOR beyond the dates from which they have already notified their departure, nor to require IBA to continue to publish LIBOR on the basis of panel bank submissions beyond such dates. We said we would consult on using our Article 23D(2) powers to require the 6 LIBOR settings to be determined under a changed methodology after end-2021, and are now doing so.
1.7 Based on information currently available to us, we expect it will be desirable to use our
Article 21(3) BMR power to compel IBA to continue to publish the 6 LIBOR settings
beyond end-2021. We can only mandate the publication for a period up to 12 months at a time. The BMR requires us to review, by the end of each compulsion period, whether it is necessary to continue the compulsion for an orderly cessation. The maximum period we can compel the administrator is 10 years. For Japanese yen LIBOR settings, we advise market participants now that we intend to compel only for one 12-month period until end-2022, after which point publication of Japanese yen LIBOR will cease. We are not consulting on this.

CP21/19
Chapter 1
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
1.8 We expect to designate the 6 LIBOR settings as Article 23A benchmarks, with this
designation taking effect immediately after end-2021. They would then be subject to a change of methodology immediately after end-2021 by our using the Article 23D(2) powers.
1.9 We are seeking views on our proposal. We will consider any representations to help us
in ensuring that we strike a fair balance between the rights of those affected and the interests of the wider market. How it links to our objectives
1.10 Article 23D(3) BMR provides that we may only exercise our Article 23D(2) powers if we
consider:

• it appropriate to do so having regard to the desirability of securing that the
  cessation of the benchmark takes place in an orderly fashion, and
• it desirable to do so to advance either or both of our statutory objectives to:
  – secure an appropriate degree of protection for consumers – protect and enhance the integrity of the UK financial system
  1.11 Our overview document sets out further the background to the BMR and the
  amendments under the FS Act to give us enhanced powers to help manage the orderly wind down of critical benchmarks and to support our objectives. What we are consulting on
  1.12 Based on information currently available to us, we consider that using our powers
  for the 6 LIBOR settings is desirable to secure an orderly cessation of LIBOR and to advance our consumer protection and integrity objectives.
  1.13 Our current view is that we are satisfied with the rationales supporting our intervention
  as set out in paragraphs 3.2-3.25. Feedback to date has been strongly supportive of such intervention. However, as indicted in paragraphs 1.28-1.29, we are interested in and will take into account any further views, information and data suggesting we should or should not use our Article 23D(2) powers for the 6 LIBOR settings to secure an orderly cessation and to advance our consumer protection and integrity objectives.
  1.14 Subject to their designation as Article 23A benchmarks taking effect, we propose to
  require IBA to change the way the 6 LIBOR settings are determined using a modified methodology based on the 2 following components:
  a. the relevant forward-looking term rates published by IBA in the UK for SONIA, and by QUICK Benchmarks Inc. in Japan for TONA, plus b. the fixed spread adjustment that applies as part of the ISDA IBOR fallbacks for the relevant LIBOR currency and tenor setting and that is published for the purposes of the ISDA IBOR Fallbacks Supplement and Protocol Please see details of our proposed technical specification for each of the 6 LIBOR settings in Chapter 2.

CP21/19
Chapter 1
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
1.15 We are also seeking views on how we propose to use our Article 23D(2) powers. That
is, whether the proposed modified methodology is appropriate, having regard to the following factors (as set out in the Article 23D Statement of Policy):

• fair approximation of the value LIBOR would have had
• least disturbance or disadvantage to affected parties
• market support on a fair way of calculating a replacement value for LIBOR
• availability to the benchmark administrator of robust and transparent inputs
• impact on the administrator
• appropriate length of time over which we require or may require a changed
  methodology
• likely effect outside the United Kingdom of exercising the power
  Outcome we are seeking
  1.16 We are proposing to use our Article 23D powers in a way that is:
  a. appropriate to secure the cessation of an Article 23A benchmark in an orderly fashion, and b. desirable to advance either or both of our consumer protection and integrity objectives. Who this applies to
  1.17 We expect that this consultation will be of interest to:
• regulated and unregulated users of LIBOR
• the administrator of LIBOR, IBA
• providers of component inputs for a potential ‘synthetic’ LIBOR
  The wider context of this consultation
  LIBOR transition
  1.18 We, alongside the Bank of England and other regulators internationally, have been
  encouraging transition away from LIBOR to alternative Risk-Free Rates (RFRs). The Working Group on Sterling Risk-Free Reference Rates (RFRWG) has been working with the Bank of England and us to catalyse a broad-based transition to SONIA, the RFR for sterling LIBOR, by end-2021. There are similar working groups in other LIBOR currency jurisdictions including the Cross-Industry Committee in Japan. We have also been encouraging adoption of robust fallbacks into all new and, where possible, legacy contracts so they continue to operate if and when LIBOR ceases or becomes permanently unrepresentative.

CP21/19
Chapter 1
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
1.19 However, we expect that there will be a pool of outstanding legacy contracts that have
no realistic prospect of being amended to transition away from the 6 LIBOR settings by end-2021. For these settings, subject to the results of our consultation and our future decisions under Article 21(3) and Article 23A, we expect to require IBA to change the way the benchmark is determined immediately after end-2021. Our overview document sets out more detailed background on LIBOR transition and our powers under the BMR as amended by the FS Act. Our BMR powers
1.20 Article 21(3A) of the amended BMR will require us to assess the capability of the
6 LIBOR settings to measure the underlying market or economic reality they are intended to measure (their ‘representativeness’) beyond end-2021 if we decide to compel their continued publication after that date.
1.21 We already announced on 5 March 2021 that these 6 LIBOR settings will no longer
be representative after 31 December 2021 due to the withdrawal of panel bank contributions immediately after that date. We are not aware of any information or developments that change that assessment or are likely to do so. Accordingly, when we make this assessment also under Article 21(3A) of the amended BMR, we expect to reach the same conclusion that the representativeness of the 6 LIBOR settings is manifestly at risk immediately after end 2021 as relevant panel banks cease contribution. Following the outcome of our assessment and applying our Statement of Policy on our Article 23A BMR powers, we expect to designate the 6 LIBOR settings under Article 23A taking effect immediately after end-2021. This means that the ‘representativeness’ of the 6 LIBOR settings will not be restored by us or IBA, and they will become permanently non-representative after end-2021, in line with our announcement in March.
1.22 In the event that we do designate the 6 LIBOR settings, we intend to use our Article
23D(2) powers to require IBA to determine the 6 LIBOR settings on a changed methodology basis.
1.23 Where we designate a benchmark under Article 23A, there will be a general prohibition
on the use of the Article 23A benchmark by UK supervised entities, except where we permit use for certain legacy contracts. We have consulted separately on our proposed policy for how we would consider whether and how to exercise our ‘legacy use’ power under Article 23C BMR. That consultation closed recently and we are currently considering responses before publishing our final Statement of Policy. We expect to consult further in Q3 on a proposed decision on how we will use our Article 23C power in respect of LIBOR.
1.24 Any ‘synthetic’ LIBOR will be time limited. We encourage market participants to continue
active transition away from LIBOR, and not to rely on a potential ‘synthetic’ LIBOR. Further Treasury legislation
1.25 In May 2021, the Treasury announced its intention to bring forward supplementary
legislation that will seek to reduce yet further any risks to contractual certainty and disputes in respect of ‘tough legacy’ contracts referencing LIBOR, where we have exercised the powers (including the Article 23D(2) powers) given to us in the FS Act.

CP21/19
Chapter 1
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings Equality and diversity considerations
1.26 We have considered the equality and diversity issues that may arise from the proposals
in this Consultation Paper.
1.27 Overall, we do not consider that the proposals materially impact any of the groups with
protected characteristics under the Equality Act 2010. But we will continue to consider the equality and diversity implications of the proposals during the consultation period, and will revisit them when making the final decision. Information we use to inform our decision-making
1.28 Chapters 2 and 3 set out our proposed decision to use our Article 23D(2) powers
based on data and information to the extent they are available to us. This includes information presented to us by market participants and their representatives, LIBOR users, national working groups and overseas authorities. We apply assumptions and estimates based on available information. We will take into account further information as it becomes available to us, including through this consultation, in making any final decision to use Article 23D(2) powers.
1.29 We welcome further views, information and data from market participants and all
relevant stakeholders, including LIBOR users, both within and outside the UK in response to this consultation. Next steps
1.30 We are seeking responses to this consultation by 27 August 2021.
1.31 You can respond using one of the forms described on page 2 ‘how to respond’.
1.32 Following this consultation, we will consider any feedback in finalising our decision. We
will communicate to the LIBOR administrator and markets in a clear and transparent way any final decision to use our powers to require a methodology change of the 6 sterling and Japanese yen LIBOR settings.
1.33 We will publish a Notice in line with the requirements at BMR Article 23D(7) where
we decide to use our Article 23D(2) powers. We will also clearly communicate any subsequent use of our powers.

CP21/19
Chapter 2
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings 2 Proposed exercise of the Article 23D(2) powers
2.1 We propose to require IBA, the administrator of LIBOR, to change the way the following
6 LIBOR settings are determined in line with the technical specifications set out below. This would be required immediately upon any designation of these benchmarks as
Article 23A benchmarks taking effect.

• 1-month sterling LIBOR is to be calculated as the sum of the 1-month Term SONIA
  Reference Rate provided by IBA and the fixed spread adjustment that applies as
  part of the ISDA IBOR fallback for 1-month sterling LIBOR and that is published
  for the purposes of the ISDA IBOR Fallbacks Supplement and Protocol (hereafter referred to as the ‘the ISDA spread adjustment’ for the relevant LIBOR setting).
• 3-month sterling LIBOR is to be calculated as the sum of the 3-month Term SONIA
  Reference Rate provided by IBA and the ISDA spread adjustment for 3-month sterling LIBOR.
• 6-month sterling LIBOR is to be calculated as the sum of the 6-month Term SONIA
  Reference Rate provided by IBA and the ISDA spread adjustment for 6-month sterling LIBOR.
• 1-month Japanese yen LIBOR is to be calculated as the sum of the 1-month Tokyo
  Term Risk Free Rate provided by QUICK Benchmarks Inc. and the ISDA spread adjustment for 1-month Japanese yen LIBOR.
• 3-month Japanese yen LIBOR is to be calculated as the sum of the 3-month Tokyo
  Term Risk Free Rate provided by QUICK Benchmarks Inc. and the ISDA spread adjustment for 3-month Japanese yen LIBOR.
• 6-month Japanese yen LIBOR is to be calculated as the sum of the 6-month Tokyo
  Term Risk Free Rate provided by QUICK Benchmarks Inc. and the ISDA spread adjustment for 6-month Japanese yen LIBOR.
  2.2 Subject to the FS Act entering into force and the designation of the 6 LIBOR settings
  as Article 23A benchmarks taking effect, we propose to exercise our Article 23D(2) powers as above because we consider doing so would be:
• appropriate having regard to the desirability of securing the cessation of these 6
  Article 23A benchmarks in an orderly fashion, and
• desirable to advance both our consumer protection objective and integrity
  objective
  2.3 Please see Chapter 3 for how we have had regard to our Article 23D Statement of Policy.
  2.4 We are seeking views on our proposal.
  2.5 We recognise that in requiring a change to how the LIBOR settings are determined
  it may be necessary to ensure that IBA has in place appropriate measures or policies to address certain operational issues associated with the process of publishing a benchmark, such as matters relating to data sufficiency and availability. We do not consider those to be part of the substantive proposal and we are not inviting views on such matters.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings 3 Why we are proposing to use Article 23D powers in this way
Part A: Decision on whether to use the Article 23D powers
3.1 We consider it appropriate to use our Article 23D(2) powers in respect of the 6 sterling
and Japanese yen LIBOR settings, assuming we make Article 21(3) and Article 23A decisions as we set out in Chapter 1, in order to secure an orderly cessation of these benchmarks beyond end-2021, and to advance our consumer protection and integrity objectives. Below we explain how we have reached this view in line with our Article 23D Statement of Policy. Appropriateness of using the powers to secure an orderly cessation of the benchmark
3.2 We have assessed data and information currently available to us regarding the likely
existence and scale of outstanding legacy contracts referencing the 6 LIBOR settings that will not be able to move away from using the settings before or upon the end of the relevant LIBOR panels at end-2021.
3.3 We have taken into account data and information available to us, including:

• data from our regular joint FCA and PRA data requests to regulated firms
• regulatory data reporting sets available to us, eg under EMIR and MiFIR, and product
  sales data reporting for mortgages
• data and information provided to us by firms, trade associations and their
  representatives in the context of our regular engagement with them
• publicly available data on securities referencing LIBOR
• views, data and other publicly available information (eg surveys) from overseas
  authorities, national working groups (incl. their sub-working groups such as the Sterling RFRWG Tough Legacy Task Force), market participants and LIBOR users both within and outside the UK
  3.4 We applied the factors set out at paragraph 2.4 of our Article 23D Statement of Policy
  to this pool of outstanding legacy contracts. That is, we considered information available to us on the scale, duration and nature of these contracts, sophistication of parties to these contracts, as well as the practicability and likelihood of amending these contracts in a fair way by mutual agreement by end-2021 without our intervention. Based on these factors, we consider with a high-level of confidence that there will be a material amount of legacy contracts, both within and outside the UK, referencing each of the 6 LIBOR settings with maturities beyond end-2021 that contain no fallbacks or inappropriate fallbacks that cannot practicably be amended by the time the relevant LIBOR panels cease. These are ‘tough legacy’ contracts for our purpose of considering whether to use our Article 23D(2) powers. We consider that, without our intervention, these contracts may not function as intended or could be at risk of frustration beyond end-2021, which would potentially lead to a disorderly cessation.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
3.5 We assess that most of these contracts are in cash markets (ie bonds and securitisations,
loans including mortgages and commercial lending) referencing the 6 sterling and Japanese yen LIBOR settings.
3.6 For derivatives, information available to us suggests that they face fewer problems
than cash markets. We expect the vast majority of derivatives contracts to be able to transition away when the relevant LIBOR panels cease. This is because parties can adopt an established industry standard documentation/mechanism, such as the ISDA IBOR Fallback Protocol or the mechanism CCPs are adopting for cleared derivatives, to facilitate amending multiple contracts at once. ‘Tough legacy’ contracts referencing the 3 sterling LIBOR settings
3.7 In bond markets (including securitisation), we understand that there are around 60
transactions that have been converted through consent solicitation since 2019 to transition away from LIBOR. Based on our analysis, there are currently estimated to be around 480 outstanding transactions in the UK at end-2021 referencing 1 of the 3 sterling LIBOR settings, with a total notional value around £90 billion. There is likely to be a material amount of these outstanding legacy contracts referencing each of these 3 sterling settings at end-2021, by number and value of contracts, which either have no fallbacks or inappropriate fallbacks to deal with the cessation of LIBOR. While we expect it is possible to amend more of these legacy contracts, we anticipate that there may be difficulties in amending some contracts within the timeframe available, particularly where they require high numbers of bondholders to consent or it is difficult to engage with some parts of the bond or note holder community.
3.8 We estimate that there are around 200,000 outstanding mortgages in the UK
(including unregulated Buy-to-Let mortgages) referencing 1 of the 3 sterling LIBOR settings, with a total mortgage value around £30 billion. We understand that terms of mortgage contracts may vary widely. It is likely that many contracts do not contain appropriate fallbacks to address the cessation of LIBOR. Among these contracts, we expect that some mortgage lenders may not be able to amend contractual provisions without explicit borrower consent. Obtaining consent for multiple contracts may be difficult within the timeframe available. Many retail mortgage borrowers are unlikely to be familiar with LIBOR transition and may decline to engage with lenders’ efforts to amend their contracts.
3.9 Sterling LIBOR is significantly used in the commercial lending markets in the UK.
We estimate that there are several hundred billion pounds of lending to businesses, across a very large number of individual contracts referencing 1 of the 3 sterling LIBOR settings at end-2021. We judge that there is likely to be a significant amount of these legacy contracts with no or inappropriate cessation fallbacks. While we expect it possible to amend many of these legacy contracts, it may not be realistic to do so in all cases before end-2021. Some may be more difficult to amend within the timeframe available. This could be because of the diverse nature of borrowers, for example, some borrowers may be unable or unwilling to engage with lenders to amend their contracts. ‘Tough legacy’ contracts referencing the 3 Japanese yen LIBOR settings
3.10 A survey by the Japan Financial Services Agency (JFSA) and Bank of Japan (BOJ) as
of end-2020 on the use of LIBOR by financial institutions found that the majority of outstanding legacy contracts in cash markets at end-2021 referencing Japanese yen LIBOR do not contain appropriate fallbacks to deal with the cessation of Japanese yen LIBOR.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
3.11 In bond markets (including securitisation), based on information available to us, we
estimate that there are at least 320 outstanding bonds/securitisation transactions in the UK and Japan at end-2021 referencing 1 of the 3 Japanese yen LIBOR settings, with a total notional value of around 14 trillion Japanese yen (around £90 billion). Based on the JFSA-BOJ survey results, we judge that the majority of these outstanding legacy contracts, by number and value of contracts, either have no fallbacks or inappropriate fallbacks to deal with the cessation of LIBOR. While we expect more of these contracts to be amended or transitioned away, we estimate that there is likely to be a material proportion of these contracts, referencing each of the 3 Japanese yen LIBOR settings, that might be difficult to amend by end-2021.
3.12 Based on the results of the JFSA-BOJ survey, we estimate that there is a significant
amount of outstanding loans, by number and value of contracts, at end-2021 referencing 1 of the 3 Japanese yen LIBOR settings. The majority of these loans do not contain appropriate fallbacks to deal with the cessation of LIBOR. While we expect more of these contracts to be amended and transitioned away, we estimate that there is likely to be a fraction of these contracts, referencing each of the 3 Japanese yen LIBOR settings, that might be difficult to amend by end-2021. Conclusion
3.13 To support an orderly cessation of these ‘tough legacy’ contracts at end-2021, our
view is that we would need to sustain LIBOR for a limited period once the relevant panel-bank contributions cease for the 6 sterling and Japanese yen LIBOR settings.
3.14 If we decide to compel publication of the 6 LIBOR settings under Article 21(3) beyond
their planned cessation at end-2021, and subsequently designate them under Article 23A, we consider that unless we use our Article 23D(2) powers, there is no other suitable way for the administrator to continue the publication of these permanently non-representative LIBOR settings for a reasonable period to support an orderly wind‑down. Desirability of using the powers to advance our consumer protection and integrity objectives
3.15 We consider that the scale of outstanding legacy sterling and yen LIBOR contracts
likely to exist at end-2021 will pose a threat to consumer protection and a risk of market disruption unless we use our Article 23D(2) powers to intervene. Using our Article 23D(2) powers to advance both our consumer protection and integrity objectives is therefore desirable.
3.16 In the absence of using our Article 23D powers, the 6 LIBOR settings will cease. As
a result, there is likely to be negative impact on consumers who have exposure to outstanding legacy contracts (eg retail mortgages, or, through their pensions or other investments, bonds and securitisations) that contain no or inappropriate fallbacks and cannot practicably be transitioned away by the time the relevant LIBOR panels cease at end-2021. These contracts may no longer function as intended, and consumers may suffer financial loss or unexpected change in the cost of their contracts, posing a material risk to ensuring an appropriate degree of consumer protection.
3.17 Our proposed decision to use Article 23D(2) powers to require continued publication
of the 6 LIBOR settings under a robust, changed methodology as specified in Chapter 2 would provide contract continuity for consumers and others whose contracts would continue to reference LIBOR (albeit under a changed methodology) after end-2021.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
3.18 Our proposed use of Article 23D(2) powers would also help achieve fair outcomes
for consumers, where they are unlikely to be able to manage the consequences of the cessation of LIBOR without our intervention. We discuss in further detail below how our proposed methodology aligns with the approach adopted by most market participants who are transitioning away from LIBOR, therefore ensuring least disturbance to consumers who would continue to rely on LIBOR beyond end-2021. Please see paragraphs 3.54-3.55.
3.19 We have also taken into account the likely financial effect on consumers. We are
proposing to use our powers in a way that would achieve a fair and reasonable approximation of the value panel-bank LIBOR would have had, while avoiding some of the volatility that has characterised LIBOR itself during periods of market stress. Please see detail in paragraphs 3.27-3.49.
3.20 We consider that our intervention would also advance our ‘integrity’ objective in the
following ways. If we do not use our Article 23D(2) powers to sustain the 6 LIBOR settings under a changed methodology beyond end-2021, there could be disruption to the integrity of certain markets and the wider financial system.
3.21 Our proposed methodology seeks to produce a ‘synthetic’ rate that is robust enough
to support an orderly wind-down. During the wind-down period, ‘synthetic’ LIBOR remains LIBOR and should flow through to allow the continued operation and valuation of outstanding legacy contracts. This means that relevant market transactions (eg mortgage interest payments) can continue, albeit the underlying methodology of LIBOR will be on a changed basis. This will minimise market disruption and help maintain the ‘orderliness’ of the financial system.
3.22 Our proposed approach to using our Article 23D(2) powers will also help maintain
transparency in the market, as it allows outstanding legacy contracts that reference LIBOR to continue to function in line with the already defined rights and obligations in the contracts. It provides clarity on how the contracts will be determined in the future through the wind-down period.
3.23 Requiring LIBOR to continue to be published on a ‘synthetic’ basis will also help
maintain the resilience of the financial system. As ‘synthetic’ LIBOR remains LIBOR and should flow through to contracts, firms party to legacy contracts referencing LIBOR will be able to continue to fulfil and benefit from these contracts. They will be able to serve their customers and meet obligations to counterparties for those outstanding legacy contracts. Without our intervention using Article 23D(2) powers, there is a risk of market disruption affecting the resilience of the financial system.
3.24 We have considered and chosen robust and transparent component inputs for our
proposed methodology for a ‘synthetic’ LIBOR as we explain in paragraphs 3.57-3.61. This means that ‘synthetic’ LIBOR will be robust against market abuse, maintaining the ‘cleanliness’ of the financial system.
3.25 We consider it feasible to produce the 6 LIBOR settings through our proposed,
changed methodology. We set out details in paragraphs 3.57-3.63.
Q1: Do you have any views, information or data which suggest that we should or should not use our Article 23D(2) powers for the 6 LIBOR settings to secure an orderly cessation and to advance our objectives of consumer protection and integrity?

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
Part B: Decision on how we would use the powers
3.26 Our proposed decision to require the 6 LIBOR settings to be calculated in line with the
technical specification in Chapter 2 has taken into account the relevant factors on ‘how to’ use our powers set out in the Article 23D Statement of Policy. Fair approximation of the value the benchmark would have had
3.27 Our aim is to achieve a reasonable and fair approximation of the LIBOR benchmark’s
expected value, having regard to the underlying market or economic reality that LIBOR intended to measure before it was designated as an Article 23A benchmark.
3.28 Currently, LIBOR is defined as ‘a widely used benchmark for short term bank borrowing
rates, produced each London business day by IBA for 5 currencies with 7 maturities ranging from overnight to 12 months’. The panel-bank LIBOR methodology is ‘designed to produce an average rate that is representative of the rates at which large, leading internationally active banks with access to the wholesale, unsecured funding market could fund themselves in such market in particular currencies for certain tenors.’
3.29 So, in economic terms, LIBOR is composed of:
a. a measure of the expectation of RFRs over a fixed period; plus b. an adjustment reflecting bank credit risk and liquidity conditions in funding markets over the corresponding fixed period, meaning that, in sterling for example, LIBOR has usually been higher than expectations of RFRs, especially at longer tenors/settings
3.30 While the 6 LIBOR settings cannot be published on a representative basis once the
relevant panels cease, we think that it is possible to measure each of these component parts in a different manner independently and then combine them to produce a reasonable and fair approximation of the relevant LIBOR setting. We propose to modify the way in which the 6 LIBOR settings are calculated, so that IBA is required to include the 2 following components:
a. a forward-looking term RFR based on the relevant overnight RFR (the Sterling Overnight Index Average (SONIA) for sterling and the Tokyo Overnight Average Rate (TONA) for yen); plus b. a 5-year historical median of the spread between the corresponding LIBOR setting and relevant RFR Below, we provide details on the 2 proposed components. a) the first proposed component – a forward-looking term RFR
3.31 The first proposed component is intended to measure, on an ongoing basis, the
expectation of interest rates over the relevant LIBOR setting (eg the sterling 3-month LIBOR setting on 1 July would include the expected interest rate in sterling markets over the 3-month period July to September). A forward-looking term rate can be generated from the fixed rates offered in overnight RFR-referencing derivatives markets, e.g. overnight indexed swaps (OIS), which provide information on market expectations of the varying overnight RFRs over a given, future period. This is the main mechanism by which IBA and QUICK Benchmarks Inc. (QBS) calculate their forward￾looking SONIA- and TONA-based rates. Figures 1A-1B show the high degree of correlation between the term rates and relevant OIS.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
3.32 For our analysis in this consultation, we use OIS data as a proxy for the relevant term
rates where appropriate, because OIS data are available for a longer period than the term rates. This allows our analysis to cover a longer historical period in demonstrating the potential effect of our proposal.
3.33 We think the best way of approximating the first economic component of LIBOR, ie
the expectation of sterling and yen RFRs over the relevant period or term, is to use a forward-looking SONIA- and TONA-based term rate. We have conducted analysis and think that this is a more accurate way of capturing this component of LIBOR than other alternatives, for example, RFRs ‘in arrears’ or backward-looking measures of the relevant RFRs (ie RFRs ‘in advance’).
3.34 An RFR that is compounded ‘in-arrears’ involves taking the RFR for each business
day over the interest period to calculate the rate that is applicable for that interest period. As a result, the applicable interest rate won’t be known until the end of the relevant interest period. This contrasts with LIBOR where the rate is identified at the beginning of the interest period. Because of this, we consider that RFRs ‘in-arrears’ are operationally unsuitable as a component for a ‘synthetic’ LIBOR which legacy contracts would need to rely upon. Please see more detail in paragraph 3.54.
3.35 An RFR ‘in-advance’ is a backward-looking measure of the relevant RFR – so the
interest rate reflects that of the previous interest period (eg if you pay interest on a monthly basis, last month’s interest rate will be what you pay this month). Using backward-looking measures of the relevant RFRs would be less effective in measuring the first component of LIBOR as they reflect the interest over the previous interest period. The discrepancies between backward-looking RFRs and the market expectation of the interest rate could be larger particularly when under market stress, as we saw in March/April 2020. Due to these considerations, we consider that a forward-looking term RFR is an appropriate component to measure the market expectation of interest rates over a fixed term that is reflected in LIBOR itself. Q2: Do you agree that a forward-looking term RFR is an appropriate component for producing a ‘synthetic’ LIBOR to measure on an ongoing basis the expectation of interest rates over a fixed term that is reflected in LIBOR itself?
3.36 For sterling LIBOR, the industry has chosen SONIA as the RFR to which to transition.
The Sterling RFR Working Group also invited interested benchmark administrators to produce a Term SONIA Reference Rate (TSRR) for some limited use cases. The 2 available TSRRs are derived from SONIA OIS tradable quotes and are very highly correlated to SONIA OIS end of day rates as shown in Figure 1A. We have identified IBA as the provider of a TSRR as a component input for the purpose of producing 1-month, 3-month and 6-month sterling LIBOR settings on a ‘synthetic’ basis, subject to the FS Act entering into force and our decision to use our Article 23D(2) powers. Please see further detail in paragraphs 3.57-3.59.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
Figure 1A: Comparing SONIA OIS with TSRR for 1-month, 3-month and 6-month sterling
settings
Notes: FCA analysis based on data from IBA, Refinitiv and Bloomberg between June 2020 and March 2021.
3.37 For Japanese yen LIBOR, the industry has chosen TONA as the preferred RFR. The
Japanese Cross-Industry Committee recommended the Tokyo Term Risk Free Rate (TORF) provided by QBS as the alternative forward-looking term rate for cash markets following a public consultation. TORF is the only forward-looking term RFR for Japanese yen. TORF is derived from, and highly correlated with, TONA OIS end of day rates as shown in Figure 1B. We consider TORF an appropriate component for the purpose of producing a potential synthetic rate for 1-month, 3-month and 6-month Japanese yen LIBOR settings.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
Figure 1B: Comparing TONA OIS with TORF for 1-month, 3-month and 6-month
Japanese yen settings
Notes: FCA analysis based on data from QUICK Benchmarks Inc. (QBS) and Bloomberg between January 2020 and March 2021. b) the second proposed component –the ISDA spread adjustment, based on a 5-year historical median spread between LIBOR and the corresponding RFR ‘in-arrears’, that is published for the purposes of the ISDA IBOR Fallbacks Supplement and Protocol
3.38 Our second proposed component for a modified methodology for the 6 LIBOR
settings is an adjustment spread to be added to the RFRs reflecting assessments of bank credit risk and liquidity conditions in specific funding markets, given these factors are not reflected in a forward-looking term RFR, but are a part of LIBOR. Unlike term LIBOR rates, the relevant RFRs do not incorporate material bank credit risk so, in sterling for example, RFRs are typically lower than LIBOR, particularly at longer tenors or settings. Figures 2A-2B show the basis, or the spread, between the 6 LIBOR settings and the corresponding RFRs over a historical 5-year period. The ‘live’ spreads between LIBOR and RFRs change over time and may change in response to market conditions, such as changing assessments of bank credit risk and funding market liquidity conditions. The spread will need to be calculated and added to the relevant RFR to provide for a reasonable and fair approximation of LIBOR and reduce the value transfer that would occur if only the forward-looking RFR were used.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
Figure 2A: Spread between sterling LIBOR and SONIA OIS for 1-month, 3-month and
6-month settings over a 5-year historical period Notes: FCA analysis based on data from IBA and Bloomberg between September 2017 and March 2021. We use SONIA OIS timeseries, as a proxy for TSRR which has only been published for a short time.
Figure 2B: Spread between Japanese yen LIBOR and TONA OIS for 1-month, 3-month
and 6-month settings over a 5-year historical period Notes: FCA analysis based on data from IBA and Bloomberg between September 2017 and March 2021. We use TONA OIS timeseries, as a proxy for TORF which has only been published for a short time.
3.39 We have not identified an appropriate and robust way of measuring in a dynamic way
unsecured inter-bank credit risk and funding market liquidity conditions in relation to the 6 LIBOR settings beyond end-2021, because there are few transactions in these markets. Prevailing liquidity conditions could affect the availability and robustness of data in calculating the spread, and therefore the robustness, reliability and volatility of a ‘synthetic’ LIBOR. This is why the current panel bank version of LIBOR relies so heavily on judgment-based submissions at longer tenors. Gaps in data, and volatility related to reliance on a very small number of transactions could not only cause deviation from rates that might be available to other participants in the markets if they chose to transact, but could also cause harm to consumers whose contracts continue to reference LIBOR after end-2021. That would be inconsistent with our objective of ensuring an appropriate degree of consumer protection. We do not consider this a suitable option.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
3.40 In our view, the fairest and most robust way to calculate the credit risk and funding
market liquidity conditions is to take the median of historical values over a lookback period that reflects a range of economic and market conditions. This approach, which is based on readily available information, is robust against manipulation compared with other alternatives that the industry also considered and discarded (eg an approach based on forward transactions where the underlying liquidity could be volatile) when looking at ways of calculating fair fallback values for LIBOR.
3.41 We consider that a historical median approach, which is less affected by outliers in the
dataset, is more stable and less sensitive to the effects of extreme market conditions when compared with a historical mean approach (which takes an average of all datapoints in the relevant historical period). As a result, a historical median approach is better in calculating a representative credit spread adjustment and minimising potential value transfer. Industry consultations in many jurisdictions also came to this conclusion in settling upon use of a median rather than a mean. See detail in paragraphs 3.50-3.51.
3.42 An alternative solution to using the median is a historical trimmed mean. However, a
historical trimmed mean approach would require discretion in examining and trimming the underlying dataset by currency and tenor settings to exclude outliers, whereas a historical median approach can be consistently applied to all LIBOR settings without such discretion.
3.43 We consider that a 5-year lookback period is better at capturing a range of economic
and market conditions that could occur in the future than a shorter lookback period.
3.44 The vast majority of industry responses to a consultation by the International Swaps
and Derivatives Association (ISDA) on the final parameters of adjustments that will apply to derivatives fallbacks for certain IBORs (including LIBOR) favoured the historical 5-year lookback median approach. This approach has been taken forwards by ISDA and is included in the ISDA Protocol, which has been widely accepted and adopted by industry, both by payers and receivers of floating rate interest payments, including in cash as well as derivative markets in major jurisdictions (see paragraph 3.51). We agree that 5 years is the most appropriate option.
3.45 We also considered how best to estimate the historical credit spread between panel￾bank LIBOR and the corresponding forward-looking term RFR (which is our first
proposed component for ‘synthetic’ LIBOR). The forward-looking term RFRs are relatively new rates and so only limited data are available (from June 2020 for both TSRRs and from January 2020 for TORF). We consider it appropriate for the spread to be calculated based on the difference between LIBOR and relevant RFR ‘in-arrears’, because the RFR ‘in-arrears’ rates have been available for longer than the forward looking term RFRs and so more data are available. This is also the approach widely supported by the industry through various ISDA consultations. We include our analysis on this, and on different lookback periods in Table 1. Overall, the differences between the approaches shown are small.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
Table 1: Comparison between 2-year, 5-year and 10-year historical median spread
between LIBOR and the relevant RFRs ‘in-arrears’ AND 5-year historical median spread between LIBOR and the relevant OIS LIBOR settings 2-year LIBOR – RFR ‘in arrears’ 5-year LIBOR – RFR ‘in arrears’ (ie the ISDA spread adjustment) 5-year LIBOR – OIS 10-year LIBOR – RFR ‘in arrears’ 1-month sterling 0.0045% Source: FCA 0.0326% Source: Bloomberg 0.0303% Source: FCA 0.0525% Source: FCA 3-month sterling 0.0917% Source: FCA 0.1193% Source: Bloomberg 0.1140% Source: FCA 0.1216% Source: FCA 6-month sterling 0.2656% Source: FCA 0.2766% Source: Bloomberg 0.2319% Source: FCA 0.2762% Source: FCA 1-month Japanese yen -0.05697% Source: FCA -0.02923% Source: Bloomberg -0.02167% Source: FCA 0.00857% Source: FCA 3-month Japanese yen -0.02490% Source: FCA 0.00835% Source: Bloomberg 0.01411% Source: FCA 0.03470% Source: FCA 6-month Japanese yen 0.04657% Source: FCA 0.05809% Source: Bloomberg 0.07085% Source: FCA 0.08054% Source: FCA Notes: FCA analysis based on data series from IBA and Bloomberg up until March 2021. The ISDA spread adjustments that apply to ISDA IBOR Fallback Supplement and Protocol were fixed on 5th March 2021. Data included in this table is only for illustration as part of this consultation and should not be used for any other purposes.
3.46 Based on these considerations, we consider the fairest and most robust way to
calculate a replacement value for LIBOR is based on a historical 5-year median difference between LIBOR and the corresponding RFR compounded in-arrears. This is the same as the ISDA approach which was subject to significant market consultation and support (see paragraph 3.51).
3.47 Further, to ensure ‘least disturbance’ to affected parties (see details in paragraphs
3.53-3.56), we propose that IBA should use the ISDA spread adjustments, that apply as part of ISDA IBOR Fallbacks for LIBOR and is published for the purposes of the ISDA IBOR Fallbacks Supplement and Protocol. Following our announcement on the end of LIBOR, these spreads were fixed. The spread was fixed at this point because knowing that LIBOR will cease or become unrepresentative on a particular date could potentially affect the underlying data used in the median calculation. Fixing the spread numbers at the point of this announcement ensured that the data used to calculate the historical median were not affected, or perceived to have been affected, by knowledge of this event. We propose that the ISDA spread adjustments should be a component for our proposed methodology for the 6 LIBOR settings to calculate a fair replacement value for the corresponding panel-bank LIBOR settings.
3.48 We attach importance to ensuring that consumers do not face unfair outcomes as a
result of our intervention using Article 23D(2) powers. We include an example below in Box 1 to illustrate how our proposed methodology for LIBOR, especially our choice of the fixed ISDA spread adjustments, could affect interest rate payments by a retail consumer with a LIBOR referencing mortgage. This is intended as an example to illustrate how interest payments are affected by the ‘spread’ component in LIBOR and ‘synthetic’ LIBOR. It does not show overall interest costs, which will also be affected by the level of risk-free rates. Actual contractual terms required by mortgage lenders vary.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings Box 1: How our proposed methodology for the 6 LIBOR settings could affect retail consumers with a LIBOR-linked mortgage The current panel-bank LIBOR comprises, in economic terms, a forward-looking expectation of risk-free rates plus a ‘live’ or varying spread over risk-free rates. We are proposing to change LIBOR’s methodology after end-2021 so that it would comprise the forward-looking expectation of risk-free rates (eg forward looking SONIA for sterling) plus a fixed spread, specifically the ISDA spread adjustment, intended to be a fair approximation of what the ‘live’ spread might be in a range of future economic conditions. This methodology would be used to calculate the 1-month, 3-month and 6-month sterling and Japanese yen LIBOR settings. 3-month sterling LIBOR is used in most sterling LIBOR retail mortgages. For a mortgage contract referencing 3-month sterling LIBOR, the interest payment a consumer pays would include the costs under the forward-looking risk-free term rate plus the costs under the spread. The interest costs related to a ‘live’ spread change over time as the ‘live’ spread fluctuates in response to market conditions. As an example, for 3-month sterling LIBOR under our proposed methodology, the fixed spread is 11.93 basis points, or 0.1193%. If a consumer has a mortgage of £150,000 referencing the 3-month sterling LIBOR and pays interests monthly, they would pay £15 a month under the proposed fixed spread component as part of the overall interest costs. For the same mortgage, the monthly interest costs the consumer would pay under the ‘live’ spread of a panel-bank LIBOR would have varied from £2 less than a SONIA rate in December 2020, to £76 above a SONIA rate, in April 2020. This is because the ‘live’ spread of 3-month sterling LIBOR over term SONIA varied between minus 1.4 basis points (or -0.014%) and plus 61 basis points (or 0.61%) between December 2019 and May 2021. This fixed spread is calculated using a 5-year historical median between LIBOR and the relevant risk-free rate (eg SONIA). Exactly the same spread is set to be used by mutual agreement in a wide variety of LIBOR contracts between large financial institutions. The 5-year lookback period covers a range of market and economic conditions, including both the brief spike in the ‘live’ spread between March and April 2020, and also a longer more recent period of an unusually low ‘live’ spread, largely due to central banks’ monetary policy response to Covid-19. At end-May 2021, the ‘live’ spread between 3-month sterling LIBOR and a 3-month forward￾looking SONIA rate was 3.3 basis points, or 0.033%. This is 8.7 basis points, or 0.087%, below the fixed 5-year historical median. Current market prices show an expectation that it will increase modestly, moving closer to the 5-year median, by the end of 2021. Whether or not a consumer paying a LIBOR-linked mortgage will end up paying more interest
under synthetic LIBOR than they would have paid under panel￾bank LIBOR depends on what panel-bank LIBOR would have been in 2022 and beyond. But we do not know in advance what values panel-bank LIBOR will take, nor will we know in retrospect because the panel will no longer exist. So, in line with the market consensus, we think that taking a 5-year historical median is a fair approximation of what this spread might be in future. Under this approach, consumers with LIBOR-linked mortgages would also benefit from no longer being exposed to the risk of the spread increasing at times of economic stress (eg in March/April 2020).

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
3.49 Taking the above 2 proposed components together, we show in Figures 3A-3B how
our proposed methodology would have tracked the value of the 1-month, 3-month and 6-month sterling and Japanese yen panel-bank LIBOR settings. This is based on available historical data between January 2011 and March 2021. Q3: Do you agree that the fixed spread adjustment that applies as part of the ISDA IBOR fallbacks for the relevant LIBOR currency and tenor setting and that is published for the purposes of the ISDA IBOR Fallbacks Supplement and Protocol is the fairest and most robust way to calculate the replacement value for LIBOR?
Figure 3A: Comparison between sterling LIBOR and ‘synthetic’ sterling LIBOR under
proposed methodology for 1-month, 3-month and 6-month settings Notes: FCA analysis based on data from Bloomberg/ISDA and IBA between January 2011 and March 2021. We use SONIA OIS timeseries, as a proxy for TSRR which has only been published for a short time, to show how a ‘synthetic’ LIBOR would have tracked LIBOR over a longer historical period.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
Figure 3B: Comparison between Japanese yen LIBOR and ‘synthetic’ Japanese yen LIBOR
under proposed methodology for 1-month, 3-month and 6-month settings Notes: FCA analysis based on data from Bloomberg/ISDA and IBA between January 2011 and March 2021. We use TONA OIS timeseries, as a proxy for TORF which has only been published for a short time, to show how a ‘synthetic’ LIBOR would have tracked LIBOR over a longer historical period. Market support
3.50 Our proposed methodology has taken into account market support that has already
been established on a fair way of calculating a replacement value for LIBOR.
3.51 There has been wide support for and adoption of the historical median approach
over a 5-year lookback period, as a way of calculating the spread between LIBOR and corresponding RFRs ‘in-arrears’. This market consensus has been established through a series of consultations by ISDA, launched at the request of international authorities, and inviting and receiving feedback from market participants in major jurisdictions, across a range of major currencies including but not limited to the 5 LIBOR currencies. Cross-market working groups in the UK, the US, Switzerland and Japan have also endorsed this approach to be incorporated in fallback arrangements in cash contracts such as bonds and loans. The Financial Stability Board’s Official Sector Steering Group has also supported the use of ISDA spread adjustments for cash contracts.
3.52 As discussed in paragraphs 3.36-3.37, national working groups in both UK and Japan
recommended certain limited use cases in cash markets for forward-looking term rates based on the relevant RFR (SONIA for sterling and TONA for yen) as part of the transition away from LIBOR.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings Least disturbance or disadvantage to affected parties
3.53 Our Article 23D Statement of Policy sets out that we will seek to use our powers in a
way that causes the least disturbance or disadvantage to affected parties. In the case of the 6 LIBOR settings, there are parties to outstanding legacy contracts referencing these settings that cannot realistically renegotiate, amend or transition away from LIBOR by end-2021. These contracts would continue to reference LIBOR (albeit under a proposed, changed methodology).
3.54 Our proposed methodology will allow those parties to maintain the ability to continue
to use LIBOR (eg to calculate and make their interest payments). Using a forward￾looking term RFR as a component of a ‘synthetic’ LIBOR, plus the ISDA spread adjustment, provides knowledge and certainty on the interest payment due to be made or received very similar to the way the legacy contract would have operated under panel-bank LIBOR. While using an RFR ‘in-advance’ (ie using backward-looking observations of the RFR) could also achieve payment certainty, it would be less effective in approximating LIBOR (as we discuss in paragraph 3.35). Using an RFR ‘in￾arrears’ as a component for a ‘synthetic’ LIBOR would require parties to these legacy contracts to wait until the relevant interest period has ended before calculating and making interest payments. For example, if a contract references a 3-month LIBOR (1 July to 30 September), using an RFR ‘in-arrears’ means consumers would only find out and make their interest payment at the end of September. This could potentially cause disturbance to affected parties. Many would need to make consequential changes to ensure their contracts could continue to operate.
3.55 Our proposed methodology would also allow affected parties to continue to hedge
their products. Using a term RFR plus the ISDA spread adjustment means that outstanding legacy contracts that continue to reference LIBOR under our proposed, changed methodology would have the same expected value of interest payments as those that are being amended to use RFRs ‘in-arrears’ over the same calendar period plus those same ISDA spread adjustments from the point that the relevant panel￾bank LIBOR settings cease. While realised overnight RFRs, compounded in arrears at the end of a period, are likely to differ from expectations at the beginning of that same period, the basis between the two can also be easily hedged.
3.56 The component parts for our proposed methodology are visible and available to
market participants. This would help minimise disruption should market participants want to extend their use of the relevant components following the end of our requiring IBA to determine LIBOR based on our proposed methodology. Availability to the benchmark administrator of robust and transparent inputs
3.57 The robust and transparent inputs required to produce the ‘synthetic’ LIBOR we
propose for the 6 LIBOR settings already exist and we expect them to be available to IBA.
3.58 As a component for a potential sterling ‘synthetic’ LIBOR, both TSRRs are robust
and transparent in line with BMR requirements. We have assessed them relative to each other on an objective, fair, reasonable, non-discriminatory, proportionate and transparent basis in line with our Article 23D Statement of Policy. Please see paragraph 2.79 of our Article 23D Feedback Statement on how we have considered our competition duty in developing the criteria in assessing potential term rate providers.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
3.59 We have decided to choose IBA as the provider of a TSRR for the specific purpose
of producing a ‘synthetic’ LIBOR for the 3 sterling settings, conditional on the FS Act entering into force and our deciding to use our Article 23D(2) powers. This will support an orderly wind-down of outstanding legacy contracts that cannot transition away by end-2021 and continue to reference LIBOR (under our proposed, changed methodology). We are not consulting on our choice of the TSRR provider.
3.60 We have decided that TORF should be used as a component for a potential synthetic
LIBOR for the 3 Japanese yen LIBOR settings, conditional on the FS Act entering into force and our deciding to use our Article 23D(2) powers. TORF is provided by QBS which is a benchmark administrator supervised by the JFSA. QBS is committed to complying with the IOSCO Principles for Financial Benchmarks. In line with our Statement of Policy, we consider TORF as a robust and transparent input. We expect TORF to be made available to IBA to be used in the production of a potential ‘synthetic’ Japanese yen LIBOR. We are not consulting on our choice of TORF.
3.61 The second component of our proposed methodology is the ISDA spread adjustments
that is published for the purposes of the ISDA IBOR fallbacks for the 6 relevant LIBOR settings. We expect this to be available to IBA to be used as an input in the production of a potential ‘synthetic’ LIBOR. Impact on the benchmark administrator
3.62 If the 6 LIBOR settings are designated under Article 23A BMR, they could then be
subject to a change in methodology under our Article 23D(2) power. The use of the 6 LIBOR settings by supervised entities in the UK, will also be restricted under Article 23C of the BMR to certain legacy use only, and the benchmark will therefore be of limited remaining commercial value to its administrator.
3.63 We have taken into account the operational, as well as the financial and commercial
impact our proposed methodology may have on IBA. As part of this, we have taken into account any potential knock-on financial effect on consumers. We seek to ensure that consumers do not bear any reasonably avoidable additional costs by proposing to use our powers in a way that would achieve a fair and reasonable approximation of the value panel-bank LIBOR would have had (see paragraphs 3.27-3.49). We expect the necessary component inputs of our proposed methodology for the 6 LIBOR settings to be available to IBA. Length of publication on the basis of a changed methodology
3.64 We said that our policy intention is to intervene by using our Article 23D(2) powers
for as short a time as is appropriate to ensure an orderly wind down in line with our consumer protection and integrity objectives.
3.65 We are required by Article 23E to review every 2 years whether the use of our Article
23D powers has advanced our objectives of consumer protection and integrity.
3.66 We also explain in Chapter 1 that we can only compel IBA using our Article 21(3) power
for continued publication of LIBOR for a maximum period of 12 months at a time and we will need to review our decision by the end of that period. We anticipate needing to review, and potentially renew, our requirement to publish synthetic sterling LIBOR. For Japanese yen LIBOR, we intend to compel publication only until end-2022.

CP21/19
Chapter 3
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings Likely effect outside the United Kingdom of exercising the power
3.67 LIBOR is used both within and outside the United Kingdom. In proposing the decision
to use our Article 23D(2) powers for the 6 sterling and Japanese Yen LIBOR settings, we have taken into account data and information available to us on the use of these settings outside the UK. We discuss these in paragraphs 3.3-3.12.
3.68 We welcome further views and data from market participants and all LIBOR users both
within and outside the UK to this consultation.
Q4: Do you have any representations or points about how our proposed 23D(2) decision would impact you?

CP21/19
Annex 1
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
Annex 1
Questions in this paper
Q1: Do you have any views, information or data which suggest that we should or should not use our Article 23D(2) powers for the 6 LIBOR settings to secure an orderly cessation and to advance our objectives of consumer protection and integrity? Q2: Do you agree that a forward-looking term RFR is an appropriate component for producing a ‘synthetic’ LIBOR to measure on an ongoing basis the expectation of interest rates over a fixed term that is reflected in LIBOR itself? Q3: Do you agree that the fixed spread adjustment that applies as part of the ISDA IBOR fallbacks for the relevant LIBOR currency and tenor setting and that is published for the purpose of the ISDA IBOR Fallbacks Supplement and Protocol is the fairest and most robust way to calculate the replacement value for LIBOR? Q4: Do you have any representations or points about how our proposed 23D(2) decision would impact you?

CP21/19
Annex 2
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings
Annex 2
Abbreviations used in this paper
Abbreviation Description
BMR Benchmarks Regulation
BOJ Bank of Japan
CCPs Central Counterparty clearing houses
EMIR European Market Infrastructure Regulation FS Act Financial Services Act 2021 IBA ICE Benchmark Administration IBOR Interbank Offered Rate IOSCO International Organization of Securities Commissions ISDA International Swaps and Derivatives Association JFSA Japan Financial Services Agency MiFIR Markets in Financial Instruments Regulation OIS Overnight Indexed Swap PRA The UK Prudential Regulation Authority QBS QUICK Benchmarks Inc. RFR Risk-Free Rate SONIA Sterling Overnight Index Average TONA Tokyo Overnight Average Rate TORF Tokyo Term Risk Free Rate TSRR Term SONIA Reference Rate RFRWG Working Group on Sterling Risk-Free Reference Rates

CP21/19
Annex 2
Financial Conduct Authority
Proposed decision under Article 23D BMR for 6 sterling and yen LIBOR settings We make all responses to formal consultation available for public inspection unless the respondent requests otherwise. We will not regard a standard confidentiality statement in an email message as a request for non-disclosure. Despite this, we may be asked to disclose a confidential response under the Freedom of Information Act 2000. We may consult you if we receive such a request. Any decision we make not to disclose the response is reviewable by the Information Commissioner and the Information Rights Tribunal. All our publications are available to download from www.fca.org.uk. If you would like to receive this paper in an alternative format, please call 020 7066 7948 or email: publications_graphics@fca.org.uk or write to:
Editorial and Digital team, Financial Conduct Authority, 12 Endeavour Square, London E20 1JN Sign up for our news and publications alerts

© Financial Conduct Authority 2021
12 Endeavour Square London E20 1JN
Telephone: +44 (0)20 7066 1000
Website: www.fca.org.uk
All rights reserved
Pub ref: 007648