Transitioning Away From LIBOR: Understanding SOFR’s Strengths and Considering the Path Forward

September 18, 2020
Nathaniel Wuerffel, Senior Vice President
Remarks at the Bank Policy Institute’s Credit-Sensitive Benchmark Symposium (delivered via videoconference) As prepared for delivery

Thanks for the kind introduction, Bill, and good morning to everyone joining virtually.

Events like this play an important role in fostering productive discourse about the transition away from LIBOR, and I am honored to participate in this symposium. As always, the views I express today are my own, and not necessarily those of the Federal Reserve Bank of New York or the Federal Reserve System.1

Right now, we are 470 days from the end of 2021, after which the existence of LIBOR is no longer guaranteed. That timeline might feel tight, but it has actually taken years of preparation, thoughtful consideration, and countless conversations to reach this point.

The transition away from LIBOR is arguably one of the most significant and complex challenges that financial markets will ever confront. In the face of this daunting task, it is important to remember why we are transitioning in the first place, and to ensure that we never have to do it again.

With that in mind, I will start by explaining why the world needs to move from LIBOR to robust alternative reference rates. I will also explain some of the reasons the Alternative Reference Rates Committee chose the Secured Overnight Financing Rate (SOFR) as its preferred alternative to U.S. dollar (USD) LIBOR. I will conclude by highlighting efforts to understand transition challenges for commercial loan products, in line with the focus of today’s symposium.

Why the World Needs to Transition Away From LIBOR

Concerns about LIBOR first became known well over a decade ago. LIBOR panel bank submissions were egregiously manipulated, which highlighted the secular decline in its underlying market.

Starting in 2012,2 LIBOR’s financial stability risks triggered reform efforts worldwide. Global regulators worked to coordinate these efforts,3 and groups like the ARRC were born, tasked with addressing the unique needs of financial markets across countries and currencies.

The ARRC is a diverse group of private market participants convened by the Federal Reserve and responsible for developing the plan to transition away from USD LIBOR.4 Its first task was to recommend an alternative to USD LIBOR. To do so, the ARRC developed key criteria to assess a range of options.

The ARRC’s criteria5 guided its selection process to make sure it chose a “robust” benchmark. Among other criteria, this included evaluating whether the rate was based on a deep, liquid market, and whether it was designed and administered in a fashion consistent with the IOSCO Principles for Financial Benchmarks.6

There are many attributes that can help make a reference rate robust, and I encourage everyone here to do some bedtime reading with the IOSCO Principles. In my own view, a robust reference rate has at least three key attributes:

  • First, it should have a reliable administrator with strong and resilient production and oversight processes.
  • Second, it should be clear what market the rate represents and how it measures that market.
  • And third, it should be based on a market that is deep and broad enough that it does not dry up in times of stress, is resilient even as markets evolve over time, and cannot easily be manipulated. By “deep and broad,” I mean that the market has enough volume and diversity of transactions to serve as the bedrock for the trillions of dollars of financial contracts that will reference it.

When assessed against this last attribute, it is quite clear why LIBOR is inadequate. In particular, over the four decades since LIBOR was formally developed, the wholesale funding market that it seeks to measure has withered. The Global Financial Crisis accelerated the decline of LIBOR’s underlying market, as banks found more stable ways to fund themselves. With so much economic value riding on a thin market, the incentive to manipulate LIBOR increased and—as we all are painfully aware—such exploitation ultimately became a reality.

Even though extensive reforms have been undertaken to make LIBOR more robust, its production primarily relies on expert judgement rather than eligible funding transactions. The U.K.’s Financial Conduct Authority (FCA), which regulates LIBOR, has noted that panel banks, “feel understandable discomfort about providing submissions.”7 So despite the sprawling use of LIBOR today,8 the FCA has said that LIBOR’s survival is not guaranteed after 2021 and everyone needs to prepare to meet that target date.9

No one could have anticipated that LIBOR would grow to be so important to financial markets and then abruptly end in this way. But we can avoid repeating past mistakes by choosing reference rates that are robust.

And it is this history lesson that makes the choice of selecting alternative reference rates so consequential. Although each market participant must choose which alternative it will use, the ARRC was convened to help in this process by recommending an alternative to USD LIBOR that could help many in the global transition.10 As a result, its decisions needed to be informed by intense study, thorough analysis, and healthy debate, conducted with the input of a broad and diverse set of stakeholders.11 Efforts to select an alternative to USD LIBOR took over three years, and ultimately resulted in the ARRC choosing SOFR as the best option.

Why the ARRC Chose SOFR as the Foundation for the Transition Away From USD LIBOR

Why did the ARRC choose SOFR?

SOFR is a broad measure of the cost of borrowing cash overnight in the U.S. Treasury securities repurchase agreement (repo) market. If you look at how SOFR is produced, what it represents, and the robustness of its underlying market, it is clear why the ARRC selected it.

SOFR is fully transaction-based and is produced by the New York Fed in accordance with the IOSCO Principles.12 It is an accurate and reliable representation of conditions prevailing across three key market segments of the U.S. Treasury repo market.13

The term “U.S. Treasury repo market”—which refers to the market that underlies SOFR—might sound like jargon. But you do not need to be a repo market participant to understand this rate or use it in contracts.

The U.S. Treasury repo market undergirds the U.S. and global financial system, and a repo trade is arguably not any more complex than a mortgage. In a mortgage, you borrow cash from a lender collateralized by a house. In a repo, you borrow cash from a lender collateralized by a U.S. Treasury security.

The market underlying SOFR is incredibly deep and has broad participation, with about $1 trillion in transactions every day and a diverse set of market participants including asset managers, banks, broker dealers, insurance companies, pension funds, and corporate treasurers. And because SOFR captures trades across multiple segments of the repo market, it is well placed to have a deep and liquid transaction base even as financial markets evolve.

The U.S. Treasury repo market also is broadly accessible and competitive. There are roughly 2,000 entities eligible to transact in the Fixed Income Clearing Corporation-cleared segment alone, and that is just one part of the U.S. Treasury repo market that SOFR measures. In the tri-party segment of the market, there typically are 40 to 50 different firms borrowing cash each day, and roughly 120 firms lending. The top 10 lenders in the market underlying SOFR account for just over half of all lending activity. The Herfindahl-Hirschman Index (HHI) is a widely used measure of market concentration. For the markets underlying SOFR, the HHI measure is below 600, indicating a competitive market.

Given all of these attributes, it is no wonder the ARRC chose SOFR as its preferred alternative, since the rate is a strong foundation for the transition away from USD LIBOR.

All of which raises the question: if SOFR is so great, why are we here?

Efforts to Understand Transition Challenges for Commercial Loan Products

Given its widespread impacts, it is important for the Federal Reserve and our official sector partners to be fully aware of the challenges to a successful transition. So, as we do with all major policy initiatives, we are holding discussions to understand the perspectives of everyone directly impacted by the transition.

The ARRC has been working hard to develop tools to facilitate the use of SOFR in all products, coordinating with a wide array of market participants and vendors to develop conventions, resource guides,14 fallback language,15 and more. But this transition is complex, and some banks have highlighted their concerns around using SOFR in commercial loans in the absence of a credit-sensitive spread.

Given the importance of this market and the banks’ concerns, the New York Fed hosted a series of four working sessions with banks and borrowers to understand their challenges in transitioning commercial lending away from LIBOR in the absence of a credit-sensitive supplement to SOFR.16 The business loan market is the largest of the cash markets transitioning away from LIBOR at around $3.4 trillion.17

The workshops were intended to explore and understand the nature of the concerns from the perspective of banks and borrowers, and to consider the potential role of a credit-sensitive supplement—but not to recommend a specific credit-sensitive supplement. I’ll highlight a few key takeaways from these sessions.18

The first workshop brought banks together specifically to understand the nature of the challenge. Participants explained that there could be a mismatch between banks’ unhedged cost of funds and SOFR-based commercial loans during an economic downturn, and that a credit-sensitive supplement could provide a natural hedge. A number of the banks thought such a supplement would be particularly important for revolving lines of credit, commercial real estate loans, and commercial and industrial loans.

The second and third workshops brought banks together to discuss data and conceptual design considerations for a credit-sensitive supplement. Participants discussed the IOSCO Principles, as well as data sources that could be relevant to constructing a credit-sensitive supplement, including overnight wholesale borrowings, various short-term money market transactions, and longer-term funding transactions.

Many of the banks expressed views that a supplement to SOFR should be credit-sensitive, dynamic, based on unsecured funding, and reflective of marginal bank funding costs. The banks also discussed a range of views on the precise transactions and issuers to include, and an appropriate frequency and observation period for calculations. There was some discussion that funding costs may vary by the size of the bank and the regulatory oversight to which they are subject.

To increase the robustness of a potential spread, workshop participants discussed broadening the set of eligible transactions, but also noted that this could potentially make a spread less representative of the funding cost for a specific market segment. Participants also discussed the risk of a credit-sensitive supplement being referenced in a broader set of products for which it was not designed or is not sufficiently robust.

The fourth—and most recent—workshop focused on understanding corporate borrowers’ perspectives on these issues. Some borrowers explained that they carefully and competitively bid out their borrowing activity to banks, taking into consideration the specific credit profile of each bank. Some expressed that their loan pricing should reflect the funding profile of the banks in their banking group and that a broader credit spread might introduce a bank credit risk component that is not actually relevant to their own chosen lenders.

Some participants expressed concerns about how a potential credit sensitive spread would intersect with existing features that banks may use to manage risk, such as commitment fees and interest rate floors. A number of the borrowers viewed the ability to hedge their loans in derivative markets as an important consideration and wondered whether a credit sensitive supplement could impair hedges and hedge effectiveness.

Both banks and borrowers highlighted the urgency of the timeline for transitioning away from LIBOR, and the complexity of system updates, product adjustments, and other changes.

Together, the working sessions this summer fostered constructive discussions around the challenges associated with transitioning commercial loans away from LIBOR. It would be premature for me to comment on any next steps at this point, but I am glad to participate in today’s symposium to hear further perspectives on these important matters.


Stepping back, I realize that everyone joining us today is exceedingly aware of the transition’s complexity and wide-reaching impacts. We welcome conversations around how to overcome the hurdles related to moving away from LIBOR, and we are ready to support the industry as it works toward solutions. In this effort, no one should lose sight of how we arrived here, and how to avoid that outcome in the future—it’s in all of our best interests to move off of LIBOR and onto robust reference rates.

Remember: we are just 470 days away from the end of 2021. Let’s make this transition stick, so that we never have to go through it again.

1 Betsy Bourassa, Eric LeSueur, and Will Riordan contributed to the preparation of these remarks.

2 The Wheatley Review of LIBOR: Final Report, HM Treasury, September 28, 2012.

3 The International Organization of Securities Commissions’ (IOSCO) developed the Principles for Financial Benchmarks in 2013 to lay out best practices for the administration and production of benchmarks. The Principles were endorsed by the Financial Stability Board, which also established the Official Sector Steering Group (OSSG) in 2013 to coordinate international benchmark reform efforts. National Working Groups in jurisdictions worldwide, like the ARRC, were born out of the OSSG to coordinate where possible while also meeting the unique needs of specific financial markets.

4 Today, the ARRC consists of 37 private sector member firms, 13 public sector ex-officio members, and 5 observers. Today, there are multiple working groups with both member and non-member participants, representing in total more than 300 member and nonmember institutions, comprised of over 2,000 individual participants, working to support a stable transition, with representation across banks, asset managers, insurers and industry trade organizations.

5 Table 2: ARRC Criteria for Potential Alternative Reference Rates, ARRC Second Report, ARRC, March 5, 2018.

6 Principles for Financial Benchmarks, IOSCO, July 17, 2013.

7 In 2017, the ARRC’s efforts—and those of stakeholders worldwide—were accelerated by the FCA, which announced that “the survival of LIBOR… could not and would not be guaranteed” after the end of 2021.

8 As noted in the ARRC’s Second Report, as of 2016, gross exposure to USD LIBOR was close to $200 trillion, roughly equivalent to 10 times the U.S. Gross Domestic Product.

9 The FCA has continued to cite end-2021 as the target date, noting in March 2020 that, “The central assumption that firms cannot rely on LIBOR being published after the end of 2021 has not changed and should remain the target date for all firms to meet.”

10 As Chair Powell has said: “The Federal Reserve convened and supports the work of the Alternative Reference Rates Committee (ARRC) and views SOFR as a robust alternative that will help many market participants in the transition away from LIBOR. However, we have been clear that the ARRC’s recommendations and the use of SOFR are voluntary and that market participants should seek to transition away from LIBOR in the manner that is most appropriate given their specific circumstances.” Also as the ARRC has noted, “The ARRC supports a vibrant and innovative market with reference rates that are robust, IOSCO compliant, and available for use before the end of 2021. The ARRC’s recommendations have always been voluntary and it recognizes that market participants may choose other rates, but any solutions must be robust.”

11 Background on USD LIBOR, ARRC, August 7, 2020.

12 The New York Fed produces all of its reference rates in accordance with the IOSCO Principles. To produce reference rates, including SOFR, the New York Fed uses a sophisticated platform that ensures the production is accurate and timely from the point of collecting the data through publication. To ensure the integrity of the rates, the New York Fed maintains a framework with regular reviews, bright-line conflict-of-interest policies, and backup protocols like the geographic dispersion of the staff and multiple sources of transaction data.

13 SOFR is based on transactions in three key market segments. It incorporates trades on Bank of New York Mellon’s tri-party platform and those for which the Fixed Income Clearing Corporation (FICC) is a central counterparty.

14 Publications, ARRC.

15 Fallback Contract Language, ARRC.

16 Transition from LIBOR: Credit Sensitivity Group Workshops, Federal Reserve Bank of New York, last modified September 17, 2020.

17 Second Report, ARRC. Business loan estimates reflect gross notional exposures as of year-end 2016 and do not include undrawn lines.

18 Minutes and presentations from those workshops are publicly available.

By continuing to use our site, you agree to our Terms of Use and Privacy Statement. You can learn more about how we use cookies by reviewing our Privacy Statement.   Close