In July 2017, the head of the Financial Conduct Authority (the “FCA”), Andrew Bailey, delivered a speech in which he questioned the future sustainability of the London Interbank Offered Rate (“LIBOR”) and announced that, by the end of 2021, the FCA will no longer use its powers to persuade or compel panel banks to quote LIBOR. According to the FCA, LIBOR is unsustainable predominantly because reforms resulting from the 2012 rate fixing scandal have been unsuccessful in grounding the rate in transactions that are representative of the LIBOR market, as opposed to the subjective judgments of panel banks. The phase-out of LIBOR will affect a wide range of financial contracts which rely on the rate, including some credit agreements and derivatives contracts. With just over two years remaining until the phase-out, parties to financial contracts that reference LIBOR should consider and prepare for the use of alternative benchmark rates.
In this bulletin, we provide a brief overview which (i) examines the definition of LIBOR and the ways in which it is used, (ii) discusses why LIBOR is problematic, (iii) examines some of the key proposed alternative benchmark rates, (v) discusses the readiness of the loan market to transition away from LIBOR, and (v) summarizes the fall-back language recently recommended by the Alternative Reference Rates Committee (the “ARRC”) for inclusion in syndicated credit agreements that reference LIBOR.
1. What is LIBOR and Why is it Important?
LIBOR is the world’s most widely used benchmark rate for short term unsecured funding between banks in the interbank market. While there is no single definition of LIBOR, it is often described as the answer given by a LIBOR panel bank to the following question: “at what rate could you borrow funds, were you to do so by asking for and then accepting inter-bank offers in a reasonable market size just prior to 11am”? A panel of banks (comprised of 11-18 banks per currency) answers this question on each business day in London by submitting quotes for five currencies (Swiss Franc, Euro, Pound Sterling, Japanese Yen and U.S. Dollar) and seven maturity periods (overnight, 1 week, 1 month, 2 months, 3 months, 6 months and 12 months). The ICE Benchmark Administration Limited (“IBA”) determines LIBOR for the day by excluding the highest and lowest 25% of submissions and averaging the remaining submissions.
LIBOR is used for various purposes, including in the calculation of floating or adjustable rates in corporate loans, derivatives, bonds and other financial contracts; in the pricing of retail products, including mortgages, car loans and student loans; as a measure of the strength of the economy; and to predict central bank interest rates. The IBA has estimated that LIBOR is quoted in US$350 trillion of outstanding contracts with maturities ranging from overnight to over 30 years.
2. The Problems Associated with LIBOR
During the global financial crisis of 2007-2008, it was uncovered that a number of banks had been manipulating LIBOR for profit. The ensuing scandal resulted in fines amounting to over US$9 billion against certain banks, criminal charges against individual traders and brokers, and calls for legislated reform to LIBOR. Serious concerns were raised about LIBOR, including that (i) the rate is based on quotes from panel banks which estimate their borrowing costs and is not based on actual transactions, (ii) there is no regulatory framework governing LIBOR, and (iii) some panel banks were falsely submitting low LIBOR quotes that did not reflect their actual cost of funds, so as to inflate their profits and give a more favourable impression of their financial strength.
In response to this crisis, in 2012, the United Kingdom government formed the FCA, which was granted broad powers to investigate and regulate financial markets. The UK government also appointed the head of the FCA to conduct an extensive review of LIBOR. The final report concluded that it would be less disruptive to financial markets to implement a comprehensive reform of LIBOR rather than replace it with a new benchmark rate, and recommended, among other things, that (i) responsibility for LIBOR be transferred from the British Bankers’ Association to a new administrator and (ii) LIBOR submissions rely on transactional data rather than panel bank estimates.
In 2014, the IBA was appointed to oversee LIBOR and implemented various control measures to prevent future rate manipulation; one of its goals was to base the rate on transactional data guided by standardized parameters rather than subjective estimates by panel banks. Despite efforts, this proved challenging to implement due to the shrinking interbank lending market and the consequent small number of transactions on which LIBOR submissions could be based. For example, in 2017, The Loan Syndications and Trading Association (the “LSTA”) estimated that over 70% of three-month U.S. dollar submissions were derived from the “expert judgement” of panel banks and not actual or interpolated transactions. Furthermore, panel banks became increasingly reluctant to provide quotes based on judgments that were not supported by transactional data given their increased exposure to liability for inaccurate or misleading submissions and the high regulatory costs. Consequently, the FCA was granted temporary legal power to compel banks to quote LIBOR. Given this fragile regulatory environment, the FCA shifted away from its original view that LIBOR should be maintained but reformed, to the decision that LIBOR needs to be phased-out in favour of alternative risk-free benchmark rates that are firmly based on transactions.
3. Alternative Risk-Free Benchmark Rates
The UK Working Group on Sterling Risk-Free Rates (the “UK Working Group”) has recommended the Sterling Overnight Index Average (“SONIA”) as their preferred alternative to LIBOR for rates quoted in sterling. SONIA is an overnight rate published by the Bank of England that is calculated as the weighted average rate of unsecured transactions in the sterling market brokered by the Wholesale Markets Brokers’ Association in London each business day. In April 2018, the Bank of England implemented reforms to SONIA, including methodological changes to broaden the number of transactions included in the rate’s calculation, thereby increasing the average transaction volume by three times its prior volume, and trading volumes have continued to grow. As of July 2019, the notional of outstanding cleared SONIA-linked swaps exceeded £10 trillion, and in the first half of 2019, SONIA accounted for over 45% of notional swaps trading in sterling. The UK Working Group’s objective is to reduce existing reliance on LIBOR and pave the way for a transition to SONIA in the sterling bond, loan and derivatives market, with the result that SONIA will be the “primary sterling interest rate benchmark” by end of 2021.
In the United States, the ARRC has recommended the Secured Overnight Funding Rate (“SOFR”) as an alternative to LIBOR for U.S. dollar contracts. SOFR is an overnight collateralized rate calculated based on transactions in the Treasury repurchase market, which has a well established and growing market. Since the Federal Reserve Bank of New York began the daily publication of SOFR in April 2018, average daily trading volume has exceeded US$800 billion in 2018 and reached as high as US$1 trillion in July 2019. Similarly to the UK Working Group, the ARRC is tasked with preparing a transition plan for SOFR to replace U.S. dollar LIBOR.
The FCA has endorsed SONIA and SOFR as alternative benchmark rates to LIBOR. The key benefit of both SONIA and SOFR is that these rates are comparatively risk-free relative to LIBOR; they are quoted using transactional data from active markets that are collected by the relevant central bank and do not rely on submissions from panel banks, and thus are less likely to be manipulated. There are a number of other material differences between the alternative rates and LIBOR: (i) SONIA and SOFR are only available in sterling and U.S. dollars, respectively, rather than the multiple-currency offering of LIBOR; (ii) LIBOR is a forward-looking term rate available in several different maturities, while SONIA and SOFR are currently only overnight rates that are backward-looking; and (iii) LIBOR is based on an unsecured interbank lending market and incorporates bank credit risk, while alternative risk-free rates do not capture bank credit risk. In the case of SOFR, since it is secured by U.S. Treasuries, it is expected to be lower than LIBOR and therefore will require an adjustment in pricing to be more comparable to LIBOR.
In Canada, the Canadian Overnight Repo Rate Average (“CORRA”) is a comparable risk-free benchmark rate which measures the average cost of overnight collateralized funding. The CORRA is based on actual transactions using calculations from on-screen trades through interdealer brokers. It is largely used for overnight index swaps and related futures. In March 2018, the Bank of Canada announced the creation of the Canadian Alternative Reference Rate Working Group (the “CARR”), which has a mandate to identify and develop a new term risk-free Canadian dollar interest rate benchmark, anticipating that if other currencies move to the primary use of term risk-free rates, market participants will likely want to have a similar option in Canadian dollars. The CARR will also examine potential enhancements to CORRA, including broadening the volume of trades used to calculate the rate. More recently, in July 2019, the Bank of Canada announced that it intends to become the administrator of CORRA in 2020, with Lynn Patterson, the Deputy Governor of the Bank, stating that “[t]he Bank expects that, over time, CORRA will be further adopted across a wide range of financial products and could potentially become the dominant Canadian interest rate benchmark, particularly in derivatives markets”.
4. The Transition Context
A 2018 report by the International Financial Law Review on their poll of market participants found that the market is generally unprepared for the challenge of transitioning from LIBOR to alternative benchmark rates. Notably, survey respondents were found to be slow in making changes to the way in which benchmarks are treated in their documentation with only 12% having made any documentary changes. Moreover, a majority of the respondents indicated that the trigger for changing their documentation would be externally driven; 80% said the trigger would be the establishment of generally accepted market terms, and 78% said the trigger would require regulatory action.
The uncertainty over the continued publication of LIBOR post-2021 and what will be the resulting widely accepted alternative benchmark rates is undoubtedly making it more challenging for the market to be overly proactive, however, there are some minimal steps that can and should be taken during the transition phase. As the LSTA noted, the transition from LIBOR to an alternative reference rate is a “process, not an insurmountable problem” and so market participants should begin to prepare for the phase-out now as preparing early will minimize both costs and risks.
Credit agreements which reference LIBOR have traditionally contained general fallback language to address the potential unavailability of LIBOR. One common fallback states that if LIBOR is not available, an alternate rate is to be used (such as an interpolated rate or quote from a designated reference bank). Another common fallback states that upon a “market disruption event”, LIBOR loans are to be converted into base rate or prime rate loans. These fallbacks are generally intended to address a situation of the temporary unavailability of LIBOR and therefore do not provide a long-term solution to the permanent discontinuance of LIBOR.
Given the limitation of the traditional fall-back language, parties to credit agreements which reference LIBOR should (i) review their current credit agreements for the sufficiency of the fallback provisions, and (ii) consider what provisions addressing the discontinuance of LIBOR should be inserted into amendments to existing credit agreements and new credit agreements with maturity dates post-2021, giving due consideration to the AARC’s recommended fallback language, which is summarized below.
5. ARRC Recommended Fallback Language
In 2018, the ARRC published guiding principles for fallback language and launched a consultation process on its proposed fallback language to solicit input from market participants. On April 25, 2019, the ARRC released its final recommended fallback language for new originations of syndicated loans and floating rate notes that reference U.S. dollar LIBOR, incorporating the feedback from the consultation process. The release of recommended fallback language for bilateral business loans and securitizations followed on May 31, 2019, which recommendations are substantively similar to those for syndicated loans. In announcing the recommendations, Tom Wipf, the Chair of the ARRC, stated that:
It’s no longer a question of if – but when – LIBOR will become unusable, yet most contracts referencing it don’t adequately account for this eventuality. With LIBOR’s possible 2021 expiration date looming, that obviously poses a massive risk to financial stability and market participants … The fallback language issued today is a critical step in addressing that concern. We encourage market participants to incorporate this language into new contracts and when possible, to begin writing contracts using SOFR instead of U.S. dollar LIBOR.
The recommendations are meant to create consistency throughout loan documentation, as well as increase the transparency and resiliency of fallback language. However, the ARRC notes that the adoption of the fallback language is voluntary
The ARRC has recommended two distinct approaches that market participants may choose from in adopting fallback language in syndicated credit agreements, namely, (a) the “hardwired approach” and (b) the “amendment approach”:
The hardwired approach builds specific fallback language into credit agreements by providing both (i) a waterfall to determine the replacement rates to be used in place of LIBOR and (ii) a waterfall to determine the applicable credit adjustment spread. The inclusion of a spread adjustment addresses the fact that LIBOR and SOFR are different rates (as discussed in Part 3 above) and therefore provides a mechanism to make the rates more comparable through an adjustment that may be a positive or negative value or zero. The waterfall of replacement rates, in order of priority, is as follows: (A) the sum of term SOFR (which is a to be determined, forward-looking term rate discussed in Part 6 below) and the spread adjustment, and if that rate is not available, (B) the sum of compounded SOFR and the spread adjustment, and if that rate is not available, (C) the sum of the alternate rate selected by the administrative agent and the borrower giving due consideration to the then prevailing market convention for determining a replacement rate and the spread adjustment. If term SOFR is the replacement rate, the adjustment waterfall, in order of priority, is as follows: (A) the spread adjustment selected by the ARRC or relevant authority, and if that is not available, (B) the spread adjustment referenced in the ISDA fallback definitions (which ISDA is in the process of developing). If a rate other than term SOFR is the replacement rate, the spread adjustment is to be agreed between the administrative agent and the borrower giving due consideration to the then prevailing market convention for determining such spread adjustment.
The use of the replacement rate is triggered by the occurrence of one of the following benchmark transition events: (i) an announcement by the benchmark administrator (IBA, in the case of LIBOR), that it has ceased or will cease to provide the benchmark rate, permanently or indefinitely, and there will be no successor administrator, (ii) an announcement by a regulator for the benchmark administrator, a central bank or other relevant authority, that the benchmark administrator has ceased or will cease to provide the benchmark rate, and there will be no successor administrator, or (iii) a public statement by the regulatory supervisor of the benchmark administrator that the benchmark rate is no longer representative. It can also be triggered by an “early opt-in election” if the following occurs: (i) a notification by the administrative agent (or a notification at the request of the borrower) that at least five (this number can be adjusted depending on the comfort level of the parties) outstanding U.S. dollar syndicated credit facilities at that time contain term SOFR plus the applicable spread adjustment, as the interest rate instead of LIBOR, and (ii) the joint election by the administrative agent, the borrower and the required lenders to declare that such early opt-in election has occurred, and the provision by the administrative agent of written notice of such election to each of the other parties.
Once a benchmark transition event or early opt-in election, and the related effective date, has occurred, all references to LIBOR (or any replacement rate) throughout the credit documentation are automatically converted to the new replacement rate and the applicable spread adjustment, as determined by the respective waterfalls as of such time, without the need for any amendment to, or further action or agreement by the parties to, the credit documents, subject to the right of the required lenders to object within five business days, if the proposed replacement rate is selected by the administrative agent and the borrower in the event that the other replacement rates given higher priority in the waterfall are not available. The hardwire approach fallback provisions are also future-proofed; the term “Benchmark” is defined broadly to include LIBOR, initially, and any benchmark replacement of LIBOR, therefore, in the unlikely event that the replacement rate to LIBOR is discontinued, any subsequent replacement rate would be selected through the same mechanism.
The amendment approach does not specifically define the replacement rate and spread adjustment that will be used upon the discontinuation of LIBOR. Rather, it provides a streamlined process for negotiating an amendment to the credit agreement to adopt a replacement rate at some future time. This approach is similar to the variants of LIBOR-replacement language that have become increasingly more common in syndicated credit agreements since the FCA’s announced phase-out of LIBOR given the limitations to the traditional fallback language.
The benchmark transition events under the amendment approach are the same as provided under the hardwired approach. An early opt-in trigger is also available if the following occurs: (i) a determination by the administrative agent, or a notification by the required lenders to the administrative agent (with copy to the borrower), that the required lenders have determined, that U.S. dollar-denominated syndicated credit facilities being executed at such time are being executed or amended to incorporate a new benchmark interest rate to replace LIBOR, and (ii) the election by the administrative agent or the required lenders to declare that an early opt-in election has occurred, and the provision by the administrative agent of written notice of such election to the borrower and lenders, or by the required lenders to the administrative agent. Once a trigger event or early opt-in election has occurred, the borrower and the administrative agent may amend the credit agreement to replace LIBOR with a replacement rate (which may be a SOFR term rate or any other replacement rate) and the spread adjustment, giving due consideration to (A) any recommendation for a replacement rate and spread adjustment by the relevant governmental authority (including the ARRC) or (B) any prevailing market convention for determining a replacement rate, provided that if the replacement rate so determined would be less than zero, such rate will be deemed to be zero.
The effectiveness of the amendment is subject to the consent rights of the required lenders. If an amendment is triggered by a benchmark transition event, it will become effective on the fifth business day after the administrative agent has posted the proposed amendment so long as the administrative agent has not received a written notice from the required lenders objecting to the amendment, in which case, interest will accrue at the alternate base rate (or other equivalent rate specified in the credit agreement) until such time as a replacement rate is selected by the administrative agent and the borrower, and not objected to by the required lenders. If an amendment is triggered by an early opt-in election, it will become effective on the date that the administrative agent receives written notice that the required lenders accept such amendment. Unlike the hardwired approach, the amendment approach fallback provisions are not future-proofed since it only contemplates the replacement of LIBOR, therefore, the appropriate modifications would need to be made to the provisions to address any subsequent discontinuance of the replacement rate to LIBOR.
Additionally, both the amendment approach and hardwired approach contain provisions which give the administrative agent the right to make technical or operational conforming amendments to the credit agreement and other loan documents as it determines may be appropriate, from time to time, to implement and administer the replacement rate, without any further consent of the other parties. The administrative agent is also required to promptly notify the borrower and the lenders of certain events, including (i) the occurrence of a benchmark transition event or early opt-in, (ii) the implementation of a replacement rate, (iii) the effectiveness of any conforming amendments, (iv) the commencement or conclusion of any benchmark unavailability period when interest would accrue at a default base rate in the absence of implementation of a replacement rate, and (v) in the case of the hardwired approach, the removal or reinstatement of any tenor of term SOFR if such tenor is not available for display on an applicable screen or other information service.
The hardwired approach provides certainty and clarity as to what will happen when LIBOR is discontinued, with minimal need for any documentary changes or consents (if any) from the parties once a trigger for the transition to a replacement rate occurs, which should make this fallback approach easier to implement and execute across a bulk of credit documents. It also allows for some flexibility in the replacement rate and spread adjustment that are adopted given the use of the waterfall mechanisms. However, a major limitation to the hardwire approach is that it requires lenders and borrowers to agree to future rates and spread adjustment methodologies that do not yet exist, and thereby forego the ability to capture any economic value from the market environment when the transition occurs. In contrast, the amendment approach provides a mechanism to facilitate future negotiation of the replacement rate and spread adjustment once they are established. However, the amendment approach is limited by the reality that it may not be practical for lenders to amend thousands of credit documents simultaneously if LIBOR ceases to be available unexpectedly without the appropriate pre-planning in place to implement the amendments. It may also encourage gamesmanship among the parties in negotiating a replacement rate depending on whether the economic market is lender-friendly or borrower-friendly at the time of transition. While the amendment approach is the preferred choice by most lenders today, as we move closer to the 2021 deadline and alternative rates become more widely accepted in the loan market, it is likely that market participants will view some variant of the hardwired approach as the more appropriate and efficient choice.
Additionally, the hardwired approach is expected to be somewhat consistent with ISDA’s anticipated fallback language for derivatives contracts. However, the language may differ in certain respects, including (i) the availability of a trigger prior to the complete cessation of LIBOR (i.e. if there is a public statement by a regulator that LIBOR is no longer representative), which could result in differences in the recommended fallback language, and (ii) the choice of term SOFR as the primary fallback rate, given the broad consensus in the derivatives market that a compounded average of SOFR in arrears should be the primary fallback rate. Therefore, the amendment approach offers market participants the most flexibility if parties are concerned with aligning their credit documents with derivatives contracts.
Over the past year, ISDA has conducted a number of consultations with market participants on technical issues related to the development of fallbacks for derivative contracts that reference certain key interbank offered rates, including LIBOR. ISDA is working to publish, in the fourth quarter of 2019, amendments to the 2006 ISDA definitions to provide fallbacks for new contracts and protocols for use of fallbacks in legacy contracts, with expected implementation by early 2020.
6. Future Outlook and Next Steps
As the deadline for the phase-out of LIBOR looms, progress is being made in the development of markets in alternative risk-free rates to LIBOR with substantial growth in trades in SONIA and SOFR in the derivatives market. In the cash market, there has also been substantial progress in the transition to SONIA for new issuances of sterling floating rate notes. However, the transition from LIBOR to the alternative rates in the loan market has been comparatively slower. One reason for this is the current absence of a term structure in the alternative rates. SONIA and SOFR are backward-looking rates that are settled overnight with the final interest payment being known only when the interest period expires. Lenders and borrowers prefer forward-looking term rates like LIBOR as the total interest payable is known at the beginning of the interest period. If a forward-looking term SOFR is established, borrowers would be able to continue to predict funding costs and plan their cash flows, and it is anticipated that lenders’ back-office operations and loan conventions (such as interest calculation methodology and treatment of weekends) would require minimal changes. If a term SOFR is not established, the likely alternative of compounded SOFR in arrears would be more difficult to adopt and necessitate operational changes for both borrowers and lenders, as well as changes to conventions.
The ARRC has indicated a preference for a forwarding looking term SOFR as the fallback rate for cash products, however, the development of this rate is subject to the ARRC achieving consensus among its members that term SOFR is a “robust transaction-based, IOSCO-compliant benchmark,” and it is unclear that such rate will be produced before the end of 2021. The UK Working Group has similarly expressed their support of the development of a robust and compliant term SONIA for use in limited cash products, but its timing is also uncertain. At the same time, the establishment of term rates is dependent on the development of the use of alternative rates in the derivatives market, the timing of which is not guaranteed, despite the progress being made. These factors have likely caused many market participants to delay their transition away from LIBOR as they take a wait-and-see approach to the establishment of term rates. However, in a recent speech at the Securities Industry and Financial Markets Association’s LIBOR transition briefing, Andrew Bailey cautioned against this approach. He emphasized that the “base case assumption” should be that LIBOR will not be published after the end of 2021, and that even if LIBOR continues to be published, it is unlikely to pass the regulatory tests of representativeness given the anticipated decline in the number of panel banks that will remain to quote LIBOR. He concluded that, to avoid complications, the best way forward is not to enter into new LIBOR contracts, but rather to transition to an alternative benchmark rate well before the end of 2021.
We will continue to monitor the phase-out of LIBOR and will provide further updates as notable developments occur.
by Kathy A. Martin and Maria Sagan
 The ARRC has recognized that an overnight benchmark rate may not be viable for some market participants and products and therefore has included in its transition plan a goal of creating a forward-looking term rate for use in cash products: see ARRC Report, supra note 22 at 20-23.
 Ibid at 7. The ARRC notes that an administrative agent may also choose to include as an additional trigger, even though it is highly unlikely, a public statement by any government authority having jurisdiction over the administrative agent that LIBOR is no longer representative or not permitted to be used by the administrative agent. The ARRC did not include this in the recommended fallback language given their focus on market-wide events and aligning the language as closely as possible with derivatives: see Ibid. at 16.
 See, for example: The Brattle Group, Report prepared for the International Swaps and Derivatives Association, “Anonymized Narrative Summary of Responses to the ISDA Consultation on Term Fixings and Spread Adjustment Methodologies” (20 December 2018); the International Swaps and Derivatives Association, “Preliminary Results of ISDA Supplemental Consultation on Spread and Term Adjustments for Fallbacks in Derivatives Referencing USD LIBOR, CDOR and HIBOR and Certain Aspects of Fallbacks for Derivatives Referencing SOR” (30 July 2019); and the International Swaps and Derivatives Association, “Preliminary Results of ISDA Consultation on Pre-Cessation Issues for LIBOR and Certain Other Interbank Offered Rates (IBORs)” (9 August 2019).
 ARRC Syndicated Recommendations, supra note 37 at 20 and 35. The ARRC is working to develop forward-looking term rates based on SOFR derivatives markets. The Federal Reserve Bank of New York also plans to begin publishing a series of backward-looking SOFR averages in the first half of 2020 along with the daily SOFR: see Federal Open Market Committee of the Federal Reserve System, "Minutes of the Federal Open Market Committee" (29-30 January 2019) at 10; and Michael Held, "SOFR and the Transition from LIBOR" (Speech delivered at the SIFMA C&L Society February Luncheon, New York City, 26 February 2019).