As alerted in our previous Advisories, LIBOR, the “world’s most important number,” is being phased out. Created almost 50 years ago on August 15, 1969—opening day of the Woodstock music festival—LIBOR began as a floating, market-determined interest rate for syndicated loans, but over time has become the benchmark interest rate for an estimated $350 trillion in outstanding financial arrangements around the world. These contracts include public and private loans and bonds; consumer financial products such as credit cards, mortgages and student loans, and some $200 trillion in interest rate derivatives.
Due in large part to concern that the determination of LIBOR (formally known as ICE LIBOR) is based on fewer and fewer interbank transactions, and therefore is an increasingly unreliable benchmark for the global financial markets, regulators worldwide have been working to develop alternative benchmarks. Over the past few years, the US Federal Reserve, the UK’s Financial Conduct Authority (FCA) and other regulators have convened industry-led working groups to develop risk-free rates (RFRs) as an alternative to LIBOR, with the goal of replacing LIBOR by the end of 2021.
In the US, the Alternative Reference Rates Committee (ARRC)—a private industry group convened by the Federal Reserve Board and the New York Federal Reserve Bank to plan the market’s transition away from US dollar LIBOR—has selected SOFR (the Secured Overnight Financing Rate) as the new interest rate benchmark for US dollar-denominated transactions in bond and loan markets. In the UK, SONIA (the Sterling Overnight Index Average) has been chosen as the new interest rate benchmark for pound sterling transactions. Other financial markets, including for transactions denominated in euro, Swiss franc and Japanese yen, have developed their own risk free rates (EONIA, SARON and TONAR, respectively).
Differences between LIBOR and SOFR
The transition away from LIBOR by 2021 presents a series of significant challenges to the financial markets, for numerous reasons. Many of the challenges stem from the basic differences between LIBOR and the proposed replacement rates, for example SOFR.
First, LIBOR is an inter-bank, unsecured lending rate, whereas SOFR is based on overnight transactions secured by US Treasury securities, a rate considered “risk free.” As a result, LIBOR is generally higher than SOFR, often by 20 basis points or more, which difference often widens at times of stress in the credit markets. Therefore a simple switch from LIBOR to SOFR, without more, would mean a lower interest rate, so in an existing transaction a transition from LIBOR to SOFR would require an upward adjustment—referred to as a “replacement benchmark spread”— to ensure that the pre- and post-amendment rate levels are compatible. The negotiation between creditors and borrowers of the amount of the replacement benchmark spread may present a challenge, because SOFR—which is tied to the securities repurchase (repo) market—is at times subject to significant volatility, particularly at month-, quarter- and year-ends. In the US, the ARRC is expected to recommend a specific methodology for determining the replacement benchmark spread, but when amending existing contracts creditors and borrowers will be under no obligation to accept it.
Second, while LIBOR is available for various tenors (e.g, one-month, three-month, six-month, etc.), SOFR is currently only available as an overnight rate on the website of the SOFR benchmark administrator (the Federal Reserve Bank of New York). For now, the lack of forward-looking term SOFR makes corporate treasurers reluctant to agree to use SOFR in their loan agreements, as they cannot predict how the new benchmark will perform. While private parties are developing forward SOFR curves for different periods (the CME Group, for example, currently publishes one-month and three-month SOFR futures), it will take time for curves to be developed and then gain widespread market adoption.
Finally, given the absence (so far) of a published forward-looking term SOFR, other methods of calculating SOFR are under consideration, each with its own challenges. For example, should SOFR be accrued from the beginning of an interest period on a daily (overnight) basis, with the final interest rate for the period only determined at the end of each interest period? Or should SOFR instead be determined for a given interest period by compounding daily SOFR for the previous one-, three-, or six-month period? While the first option would better reflect market interest rates during the interest period, neither the creditor nor the borrower would have predictability in terms of future interest income/expense. Many corporate treasurers would be informed immediately preceding the payment date how much interest would need to be paid, raising operational back- and middle-office issues both for creditors and borrowers; a problem compounded for those non-US borrowers required to close a foreign exchange transaction in advance to effect US dollar payments.
Amending Existing Contracts
Perhaps the greatest challenge to the transition from LIBOR to SOFR will be to amend the contractual terms of existing financings that are due to mature after the LIBOR transition date. An estimated $35 trillion of currently outstanding LIBOR-linked financial transactions expire after 2021 (in comparison, the US national debt is $22 trillion). The problem is particularly acute if, prior to the parties successfully amending the contracts, LIBOR itself is no longer published or is otherwise no longer considered a reliable measure of inter-bank lending rates. Concern has been raised about so-called “Zombie LIBOR,” where LIBOR remains in legacy contracts after the point when it is no longer supported or reported.
Loan documents for LIBOR-based loans generally provide a definition of LIBOR, with that definition providing certain “fallbacks” in case LIBOR is no longer determinable based on the method provided in the document (generally a designated display page on a Reuters or Bloomberg rate screen). However, these fallbacks are—particularly for many older agreements—generally triggered only when LIBOR is unavailable (for example, if for some reason LIBOR is not displayed on the designated rate screen on the interest rate determination date), but do not consider a scenario where LIBOR no longer exists. Loan agreements also typically contain provisions that apply an alternate base rate in the event that either (a) LIBOR cannot be determined, (b) dollar deposits are not being offered in the London interbank market or (c) LIBOR no longer reflects the lender’s cost of funding a loan. Those alternate base rates are often based on the Prime Rate, the Fed Funds rate (plus a margin) or some other agreed upon rate, but these alternate rates were added as a short term solution for a temporary disruption, not as long-term replacements for LIBOR, in particular as those rates are often significantly more expensive than LIBOR.
More recent LIBOR definitions will generally provide a different fallback, already contemplating a time when LIBOR no longer exists, and industry groups have been working to develop a consistent approach. In 2018, the ARRC released market consultations on potential fallback language for syndicated loans, floating rate notes, bilateral loans and securitizations. In April 2019, the ARRC published its recommendations of fallback language for syndicated loans and floating rate notes, based on feedback it received from market participants. Even so, while at least one major bank has adopted the ARRC recommendations in whole, fallback language is still being developed and it will be difficult for the market to develop adequate language until the uncertainties surrounding SOFR are resolved.
Loan modification itself will be a challenge, even after market-standard fallback language has become more fully developed. Loan modification negotiations for bilateral loans between lenders and sophisticated borrowers should be relatively straightforward, though any discussion of a benchmark spread adjustment may be a challenge, especially if negotiated at a time of market stress, when LIBOR and SOFR diverge more significantly. Syndicated loans, the documentation of which often require the approval of lenders holding 100% of the outstanding loan for any proposed modification of the method for calculating interest, will be a greater challenge, particularly if there are numerous members of the lending syndicate with different levels of sophistication regarding the market shift from LIBOR to SOFR. In addition to the obvious LIBOR provisions, amendments may also need to be made to provisions regarding break-funding, make-whole and increased costs, among other clauses. The coordination role of administrative agents will be critical.
But the greatest challenge will likely be to modify floating rate notes (FRNs) that have been widely distributed, as generally the approval of noteholders holding 100% of the outstanding notes is required to amend existing terms and conditions affecting interest rates. To the extent a LIBOR-based FRN is held by a significant number of retail investors, and the terms of the FRN require 100% approval for amendments and have an old-style LIBOR definition, then liability management exercises (such as debt-for-debt exchange offers) should be considered to help mitigate the risk, if at least in part.
New Loan Agreements, Prior to LIBOR Cessation
For new loan agreements being entered into now and using LIBOR as the interest rate benchmark, the ARRC has proposed two different approaches for making future amendments when LIBOR ceases: the “amendment” approach and the “hardwired” approach. Each of these approaches has slight variations when applied to syndicated loans as opposed to bilateral loans. Generally, the “amendment” approach provides that, upon the occurrence of a defined LIBOR replacement trigger (certain cessation or pre-cessation events), the lender (in the case of a bilateral loan) or the borrower and the administrative agent (in the case of a syndicated loan) will have agreed to amend the loan agreement to replace LIBOR with an alternate benchmark rate (which may include term SOFR) unless the other party or parties to the loan agreement (for example, a certain percentage of “required lenders”) object in writing within a specified timeframe. Under the “hardwired” approach, upon the occurrence of a defined LIBOR replacement trigger, LIBOR is automatically replaced with term SOFR, compounded SOFR or another alternate benchmark rate determined in accordance with the terms of the loan agreement using an agreed “waterfall.” Each of these approaches has advantages and disadvantages.
Generally speaking, the “amendment” approach provides the parties with greater flexibility in establishing a rate to replace LIBOR upon the occurrence of a LIBOR replacement trigger. However, a disadvantage to the “amendment” approach is that the parties may not be able to agree on a replacement rate when LIBOR replacement is triggered and, in that case, the existing (and inadequate) fallbacks will remain in the loan agreement. Depending on the specific wording of these fallbacks and the then-current market, the result will either be inadequate, unduly expensive or unworkable (particularly for loans having longer tenors), and will inevitably in some cases lead to litigation.
At the same time, while the “hardwired” approach has the advantage of not depending upon the parties reaching agreement to a replacement rate in the future at the time a LIBOR replacement trigger occurs, the parties do risk agreeing in advance to a replacement rate (e.g., term SOFR or compounded SOFR) that does not currently exist and may never fully develop. The ARRC’s “hardwired” approach includes a required benchmark spread adjustment based on spread adjustments (or adjustment methodologies) published by relevant governmental bodies or ISDA. While the “amendment” approach contemplates that the parties will select a benchmark spread adjustment at the time of the amendment, the parties would still need to agree on the amount (or the methodology for determining the amount) of the adjustment, though giving “due consideration” to certain defined factors. Given that LIBOR is generally higher than SOFR, these spread adjustment provisions are important. Without such provisions, borrowers and the lenders will have different incentives in determining whether a LIBOR replacement was actually triggered, with borrowers likely preferring an early switch to SOFR and lenders preferring a later switch (all other things being equal). For this reason, the ARRC spent considerable effort to develop objective and knowable triggers as part of its consultations.
Our recent experience suggests that parties are more comfortable with the amendment approach than with the hardwired approach until more information regarding replacement rates becomes available.
The migration away from LIBOR presents other risks to both creditors and borrowers.
Hedging. ISDA is undertaking separate consultations for the derivatives markets, and consulted with the market in July 2018, announcing final recommendations at the end of 2018. While ISDA undertook its consultations before the ARRC took its own, there is concern that the ISDA fallbacks and ARRC fallbacks may not align, generating worries of fallback misalignment between loans / notes and their respective interest rate hedges. For example, unlike the ARRC proposal which includes a pre-cessation trigger (e.g., a public statement by the regulator for the administrator of LIBOR that LIBOR is no longer representative), the current ISDA proposal does not include pre-cessation triggers. Accordingly, loan agreements that include the ARRC’s pre-cessation trigger may result in the replacement of LIBOR before it has ceased to be published, while any associated hedges using ISDA’s fallback language would continue to be based on LIBOR until it is officially discontinued. Another example of potential divergence between the LIBOR replacement proposals applicable to the cash (loan) and derivatives markets relates to the method of determining replacement benchmark itself. ISDA has announced that it will utilize a compounded replacement rate calculated in arrears as its fallback for derivatives. To the extent the loan market adopts a forward-looking term SOFR (or some other methodology for determining the replacement benchmark) instead of a compounded SOFR in arrears, there will be a mismatch between loans and their associated hedges.
Regulation. For financial institutions in particular, there is increased regulatory focus on ensuring that banks are prepared. The FSB (Financial Stability Board) and IOSCO (the International Organisation of Securities Commissions) have been coordinating international efforts for interest rate benchmark reform. In recent months, the US’s Federal Deposit Insurance Corporation (FDIC) focused its Winter 2018 issue of Supervisory Insights to the end of LIBOR, while the US Securities and Exchange Commission (SEC) has identified it as a disclosure and operational concern for both reporting companies and securities industry participants.
Taxation. In the US, a number of issues have been raised about the tax impacts of converting existing loans and other financial instruments from LIBOR to a replacement rate. For example, there is concern that the conversion could result in a determination that there was a material modification of the indebtedness, potentially resulting in a taxable exchange. A similar concern is raised under FATCA, where a material modification to an existing financial instrument can cause the issue to be deemed a new issuance, jeopardizing the “grandfathering” exemption from FATCA withholding for instruments issued before July 1, 2014. This would be a particular concern in the context of an older, existing securitization, where the documentation establishing the issue likely contains no provisions contemplating FATCA withholding. On April 8, 2019, the ARRC sent a letter to the US Treasury and the US Internal Revenue Service requesting formal guidance on these and similar US tax issues.
Potential for Disputes. For many of the reasons discussed above, there may be instances where it will be a challenge to incorporate fallback provisions into an existing financial instrument prior to the cessation of LIBOR because of the inability to obtain requisite consent from the relevant constituents. Absent a statutory or other “fix” that applies across the different market segments, the potential for disputes in these cases is a real concern that should be considered as firms analyze their needs and objectives.
Both creditors and borrowers should already be preparing for the transition from LIBOR to SOFR. We would recommend the following:
First, parties should take stock of their LIBOR exposure under existing loans, notes and derivatives, focusing on transactions maturing after 2021. The definition of LIBOR should be reviewed, as well as the provisions for amending the terms and conditions of the regarding modification. Discussions with counterparties should begin as soon as possible to ensure that counterparties are also aware of the conversion from LIBOR to SOFR.
Second, parties should review standard documents that are likely to be used for future transactions, such as under medium term note (MTN), commercial paper (CP) or certificate of deposit (CD) programs, to check whether amendments should be made in contemplation of future issues. For example, consideration should be given to changing existing program documentation to permit less than 100% approval for amending LIBOR-related interest rate provisions, reducing the ability of small groups of holdout creditors to block necessary amendments. Parties should monitor developments in standard LIBOR replacement language and the developments involving potential term SOFR and “replacement benchmark spread.” In addition, given the concern that both SOFR and LIBOR fallbacks may develop in different directions between standard lending/securities documentation and standard ISDA documentation, companies should review credit and hedging documents carefully to avoid potentially costly gaps.
Third, parties, particularly lenders and agents, should review their internal systems to understand what adjustments may be required for loan accrual in SOFR, whichever SOFR calculation method is ultimately used by the market. Back- and middle-office systems and procedures, such as client invoicing, will also need to be adapted.
Finally, parties, again, particularly lenders and agents, should already begin advising borrowers and issuers that LIBOR is coming to an end, preparing them for the changes to come. Less sophisticated counterparties may need additional time to educate themselves on the upcoming changes to LIBOR and the adoption of SOFR.
While 2021 may still seem well in the future, the adjustments that market participants will need to make are significant, and these adjustments will be undertaken while replacement rates and fallback provisions remain unresolved. The time to take stock of your company’s exposures, and to map a path forward, is now. One of the last bands to perform at Woodstock was Blood, Sweat and Tears. If the issues raised by the transition to risk-free rates remain unresolved by the end of 2021, LIBOR may well end on a similar note.
This post comes to us from Arnold & Porter Kaye Scholer LLP. It is based on the firm’s memorandum, “The End of LIBOR,” dated June 11, 2019, and available here.