The sudden and significant shock to U.S. financial markets from COVID-19 earlier this year pushed most discussions about the transition from LIBOR into the background. Many market participants hoped that, like a bad dream, the LIBOR phase-out deadline would simply go away. However, this does not appear to be the case. With 15 months remaining until the scheduled sunset of this key interest rate benchmark, the transition toward alternative interest rates is returning to focus. Despite the significant milestones achieved by securitization industry participants, there remains significant work to complete in order to minimize potential disruptions to these markets after December 2021--both for new issue and for legacy transactions. S&P Global Ratings has observed significant variability in liability fallback language across major U.S. structured finance (SF) sectors based on its initial review.

Summary of the transition from LIBOR in U.S. securitizations

• S&P Global Ratings rates about 4,800 LIBOR-based securitizations globally, with over 95% in the U.S.
• Due to transaction mitigants, LIBOR exposure doesn't automatically indicate a rating impact.
• Our initial transaction sample review in the U.S. indicates six general categories of liability fallbacks.
• Liability fallbacks are dominated by the transaction party selecting a new rate and fixing the rate at the last posted LIBOR rate.
• Legacy transactions are difficult to amend given required unanimous noteholder approvals.
• Basis risk scenarios are relevant if assets and liabilities don't transition simultaneously.
• The U.S. lags the UK market, where several dozen securitizations have been executed with new risk-free rates (RFRs).
• New RFRs, such as the Secured Overnight Financing Rate (SOFR), are structured differently than LIBOR, so a one-for-one replacement is not feasible.
• Care will be needed to not transfer value between borrower and lender.
• The Alternative Reference Rates Committee (ARRC) recommended that issuance of dollar LIBOR securitizations stop by June 2021 (and issuance of LIBOR-based CLOs by September 2021).
• Transactions with ARRC language could hit triggers before December 2021 if the Financial Conduct Authority (FCA) issues a statement on LIBOR non-representativeness.
• Regulators have announced they are examining financial institution's LIBOR transition plans as part of their routine exams.
• Questions remain, such as "Will transaction cash flows be impacted if there are disputes?", "Will proposed New York legislation providing safe harbor against litigation be enacted?", and "Will 'synthetic LIBOR' be published beyond 2021, and will it apply to dollar LIBOR-lined transactions in the U.S.?"

The Impact Of The Transition From LIBOR Is Not Uniform Across Credit Markets

While LIBOR is the world's most widely used interest rate benchmark, the impact of its discontinuance is not equally spread across credit markets. The largest dollar exposures to LIBOR remain in the derivatives markets, estimated at approximately $200 trillion, dwarfing cash market securitization exposures of approximately $700 billion (S&P Global Ratings rated U.S. securitizations). The derivatives market transition appears to be advancing in light of the International Swaps and Derivatives Association (ISDA) decision-making around swap documentation transition (see below) and derivatives clearing preparation. For example, transitioning price alignment (and discounting) for USD over-the-counter-cleared swaps from the Effective Federal Funds Rate (EFFR) to SOFR is anticipated before year-end 2020.

However, the challenges of ending LIBOR are more acute in cash markets and particularly in SF. This is primarily because of the typical multi-creditor classes of securities in securitizations and related challenges gaining unanimous investor approvals for amending fallback provisions in transaction documents. Bilateral agreements found in sectors outside securitization, including the derivatives market, can be more easily amended with mutual borrower and lender consent. Furthermore, many non-SF securities utilizing LIBOR mature or can refinance before late 2021, affording another opportunity to change benchmark interest rates.

New Rate Selection Requires Care To Minimize Value Transfer

The transition from LIBOR to new interest rates raises the question about a potential value transfer between borrower and lender in securitizations. In fact, one of the guiding principles of official bodies, such as ARRC, in the design of their recommendations and fallbacks for the LIBOR sunset is to minimize any transfer of value between these two parties. This is challenging in practice for outstanding transactions because existing documentation (other than those with ARRC fallbacks) typically does not specify adding a credit spread to new RFRs like SOFR. If a new RFR was substituted for LIBOR (which contains bank credit risk) without a credit spread, then that could represent a transfer of value from lenders to borrowers and be more susceptible to challenge from investors.

The Transition From LIBOR Is Happening Faster In The UK Than In The U.S.

While there has been significant discussion about the transition from LIBOR in U.S. securitization markets, the transition has been quickest in the UK. Since 2019, new floating-rate UK securitizations and covered bonds have largely used the Sterling Overnight Index Average (SONIA) as the benchmark rate in lieu of Sterling LIBOR. S&P Global Ratings has rated 54 new UK securitizations (including covered bonds) with SONIA-based liabilities through August 2020, totaling approximately £35 billion. This activity does not address legacy contracts, however, which, in the UK, like in the U.S., pose challenges for amending interest rate provisions.

Many Challenges Are Facing Legacy U.S. Securitizations

While U.S. securitizations since 2019 have generally incorporated hardwired ARRC fallback language in transaction documents governing liabilities, this is not the case for most older, legacy transactions where the permanent cessation of LIBOR was not contemplated. The greatest challenge with this large group of transactions seems to be the difficulty in amending liability interest rate language, often requiring unanimous investor approval per the Trust Indenture Act of 1939. Furthermore, given the size and potential disruptive nature of the transition from LIBOR that extends beyond securitizations, ARRC also proposed New York legislation to help address transactions with limited fallbacks and create a safe harbor against potential litigation. However, as of this writing, the timing and ultimate success of that legislation is uncertain. We are just starting to see transaction amendments and field issuer and investor questions around the transition from LIBOR. Key challenges facing legacy U.S. securitizations include:

• The difficulty of amending transaction documents given the unanimous investor approval typically needed;
• The lack of clear fallback provisions and/or differences in fallback language between assets and liabilities;
• The extensive logistics (notifications, fielding questions, adjustments, compliance with consumer regulations where applicable) needed for amending LIBOR-based collateral (consumer and corporate loans) in securitization;
• The introduction of basis risk if LIBOR-based assets and liabilities (and any related hedge agreements) are not switched to the same benchmark at the same time;
• That embedded derivatives in transactions with LIBOR payments, although limited in the U.S. market, may not be fully aligned with both assets and liabilities;
• That our initial transaction sample review suggests over 75% of legacy transactions contain language enabling a transaction party to select a replacement rate or causing interest rates to convert to a fixed rate (the last-quoted LIBOR);
• That there is no term SOFR rate currently available, and none are expected before mid-2021;
• Operational and system challenges using a compounded daily alternative reference rate in lieu of LIBOR and selecting "in arrears" or "in advance" interest calculation methodology (the "in advance" method being favored by the majority of securitization industry participants for liabilities);
• That some underlying loan contracts (collateral) are silent on interest rate changes and would likely follow the regular amendment process; and
• Potential litigation and/or disputes on rate selection.

Six Fallback Language Categories Were Observed In A Sample Of Legacy U.S. Transactions

We continue to assess outstanding transactions containing LIBOR in assets and liabilities for potential rating impact. Given the inherent customization present in SF, it's not surprising that there's variability in fallback language among transactions containing LIBOR-based liabilities. In our initial cross-sector review of U.S. SF transactions, the documentation governing fallbacks for the liabilities typically contemplates a temporary rather than permanent disruption of LIBOR. So far, we've observed that the ultimate fallback language can be generally classified across six broad categories (see table 1). In most cases, the remedy if LIBOR is unavailable is to go through a series of additional steps, usually starting with contacting multiple banks in London for dollar deposit rate quotes followed by polling U.S. banks for rate quotes and ultimately leading to a final fallback state.

Table 1

There Is Significant Variability In Fallbacks Across SF Sectors

The use of LIBOR is uneven across the U.S. SF market. Three sectors--ABS student loan, legacy RMBS (2008 and earlier), and CLOs--account for over 90% of the LIBOR-linked transactions (by count). As we initially examined a sample of fallbacks in each sector, it became apparent that there is significant variation. Chart 1 shows this variation across the six fallback categories previously discussed. This variation speaks to different risk profiles in the SF sectors to be assessed.

Chart 1

Within SF sectors, there is variability as well. For example, the vast majority of LIBOR-linked securities in ABS is in the student loan sector, with very limited auto ABS using LIBOR-linked liabilities. Exposures drop off significantly in other sectors, which generally utilize fixed rates. Table 2 presents some of these differences.

Table 2

Legacy RMBS Sector Fallback Distribution

Chart 2 displays the ultimate liability fallback distribution of a sample of 450 legacy RMBS transactions across 1999-2008 vintages. While this represents only a portion of approximately 3,000 legacy RMBS transactions containing LIBOR, we believe it has adequate representation across the major fallback categories identified. Bank polling is a larger category in this cohort than in most active SF sectors. In addition, ARRC language doesn't exist in this sector, as it generally began appearing in 2019, while sample vintages for this sector stop in 2008. To the extent some of the original securitization sponsors no longer exist, it remains to be seen how active transaction parties will be in seeking the best possible solution.

Chart 2

Conversion To A Fixed Rate And Transaction Party Discretion Are The Most Common Fallbacks

During an initial review of U.S. transactions, we examined liability fallbacks across U.S. SF sectors. For active market segments (excluding legacy RMBS), two fallback categories dominated. (see chart 1). There are usually several steps indicated prior to ultimate fallback language, which was aggregated in chart 1.

Fixing LIBOR at the last-quoted LIBOR rate (converting to a fixed-rate security)  

The first fallback, if not amended, could introduce interest rate risk and pricing/liquidity issues. But fixing the liability rate in a low-interest-rate environment as currently exists may not necessarily be problematic from a credit perspective, and further analysis may be needed to determine any rating impact. Indeed, the current 12-month forward-dollar LIBOR rate is currently around 0.36%, or the market's expectations of spot LIBOR one year from now. This rate is very low by historical standards and well below LIBOR rates when most legacy transactions were originated.

The transaction party selects a replacement rate/terms  

The second fallback provides discretion to a transaction party, but the party's intentions will be important for market participants to learn well in advance of the scheduled transition date. ARRC's May 2020 best practices publication recommends parties that have discretion over choosing a replacement rate make their intentions known by at least six months ahead of LIBOR replacement, which would imply no later than June 2021.

ARRC Fallback Language Incorporated Since 2019

In anticipation of the transition from LIBOR, since 2019, securitization sponsors in the U.S. have enhanced the interest rate replacement language in their liability documents for new issue transactions. We have observed that these changes have generally followed key elements of the ARRC's recommended hardwired fallback language with the three key steps: trigger events, a waterfall of replacement rates, and a spread adjustment. In most key areas, such as alternative rates and spread adjustments, ARRC recommendations are consistent with those of the ISDA, which is a positive for hedging in securitizations. As of Oct. 1, 2020, the ISDA has indicated it plans to release in January 2021 its amendments to the 2006 Definitions, which include specific recommended fallbacks and triggers.

Transactions With ARRC Precessation Triggers Merit A Closer Look

Most 2019 and 2020 U.S. securitizations with LIBOR liabilities have incorporated ARRC fallback language. In most cases, they include several trigger events that, if breached, would begin the transition process to a new rate. One of those events, a precessation trigger, could be triggered well before the scheduled date for the transition from LIBOR of December 2021 and to that extent bears watching. Precessation would occur if the regulator of the LIBOR administrator were to make a public statement that LIBOR is no longer a representative rate to a specific date in the future. In particular, we believe recent transactions with an early opt-in feature to switch rates would likely exercise those rights prior to the scheduled end of LIBOR as would be specified in a precessation statement.

The ARRC's recommended precessation trigger is "a public statement or publication of information by the regulatory supervisor for the administrator of the benchmark announcing that the benchmark is no longer representative."

Activating such a trigger, in our view, would be a positive event because it would signal a planned and orderly shift away from LIBOR. Two open questions are how long it may take the relevant transaction party (often the servicer or collateral manager) to amend any LIBOR rates on underlying collateral and whether there is material exposure to basis risk in these scenarios due to an extended mismatch. When this trigger is activated, the spread added to the replacement RFR could be calculated and determined for use later on the future date when LIBOR will be deemed to be nonrepresentative.

Asset Fallbacks Are Often Silent But Easier to Amend Than Liabilities

Although there is a general lack of robust hardwired interest rate fallback in many existing loan agreements, these agreements, unlike securitizations, are individually easier to amend than multi-creditor securitization facilities because they are typically bilateral. How asset interest rates change relative to liability interest rates will become important in scenario analysis of different interest rate and basis risk assessments of SF securities. We believe that for U.S. securitizations where contracts currently permit the adjustment of benchmarks, most active servicers, master servicers, and collateral managers aim to time a switch in the rate index on underlying assets and liabilities as closely as possible to minimize any basis risk. Switching interest rates on thousands of individual consumer loans will present significant logistical and operational challenges, including the time needed to address consumer questions, comply with consumer lending laws, send notifications, etc. Similar logistics surrounding contract amendments are expected in the corporate loan market where loan agreements often lack clear transition language if LIBOR phases out. Finally, collateral terms in consumer and corporate loans have generally trailed hardwired liability fallbacks in recent securitizations.

LIBOR Exposure Does Not Automatically Indicate A Rating Impact

A securitization's exposure to LIBOR doesn't automatically signal a rating impact, as some transactions possess mitigants. Some of these include:

• Limited to no reliance on excess spread (such as senior tranches with higher credit enhancement);
• Interest that may be payable-in-kind (such as junior to mezzanine tranches in CLOs);
• Pass through coupons with net weighted average coupons limiting liability coupons to asset yields less fees (such as in RMBS);
• Clear fallbacks, such as transactions with ARRC-hardwired fallbacks, and those with specified alternative rates (non-ARRC), such as the prime rate;
• Single or highly concentrated investor groups where liability amendments to new replacement rates are feasible; and

Because these factors vary by transaction, additional analysis is required to determine the credit impact of the transition from LIBOR. We have commenced this review to categorize and assess potential ratings impact. To the extent servicers, master servicers, trustees, collateral managers, calculation agents and other transaction parties plan to make amendments or announce specific alternative rates, spread and calculations under the terms of the transaction documents, we will seek that information to reflect any potential credit risk profile changes.

As previously discussed, borrowers and lenders switching from LIBOR aim to make interest rate changes value-neutral such that the dollar LIBOR rate is as close as possible to the SOFR rate plus an appropriate spread. The more closely this is adhered to, all things equal, the less likely there would be a rating impact.

New Risk-Free Rates, Such As SOFR, Are Structured Differently Than LIBOR

Central banks have developed new RFRs, such as the Secured Overnight Financing Rate (SOFR) in the U.S., to replace LIBOR. There is one new rate for each of the five currencies currently quoted by LIBOR. There are fundamental differences with their design vis-à-vis LIBOR that do not permit a simple substitution for LIBOR in transactions without any adjustments, such as a credit spread. Many transaction documents don't specifically mention adding a spread to a replacement rate, yet this is an important step to minimize value transfer. The design of new RFRs attempts to address two main shortcomings of LIBOR: its susceptibility to manipulation on account of the expert judgment involved in determining LIBOR, and its distortion risk because LIBOR no longer represents the wholesale markets it was intended to represent (unsecured interbank lending volumes are too small and therefore unrepresentative of benchmark lending). While monetary authorities appear open to multiple alternative replacement rates for dollar LIBOR, it is clear that SOFR is the leading candidate given traded derivative products and tax guidance rulings. Moreover, ARRC recommendations are more tied to this rate. (See the appendix for comparison of differences between LIBOR and SOFR.)

Anticipating Interest Rate Scenarios Arising From The Transition From LIBOR

While the transition from LIBOR is a dynamic process that continues to unfold with new developments, we can project a number of interest scenarios at this stage. These may be reassessed and examined in more detail as additional information becomes available and the relative merits of each case are considered in each sector and transaction. Some situations are likely to result in running cash flow analyses to assess rating impacts, while others may have adequate mitigants and not need such analysis. Not all information is currently available for outstanding LIBOR-linked securitizations, such as what replacement rate the designated transaction party will select as required under transactions documents. These hypothetical scenarios do not include the potential impact of disputes or adverse tax or other legal challenges to these transactions, not to mention favorable treatment under proposed legislation.

Table 3

The Transition From LIBOR For Consumer SF Sectors

Interest rates on SF liabilities that transition toward other benchmarks including SOFR are likely to transition prior to underlying asset rates linked to LIBOR. While securitization sponsors have indicated they prefer to change asset rates at the same time as liability rates, there may be a practical need to switch them at different times. As a result, there could be some transition period where an incremental amount of basis risk exists (provided transaction parties follow ARRC recommended guidelines on spread adjustments). Additionally, some basis risk may arise where underlying consumer loan assets in a securitization are converted to a new interest rate but the credit spread is phased in over a one-year period (consistent with ARRC guidelines) but where the liability rate changes at once, including the full credit spread adjustment. Depending on how much reliance there is on excess spread and other transaction mitigants, further analysis may be needed to quantify the potential impact of this gap.

Potential Basis Risk With Hedging And SOFR

In cases for which hedges support SF securities (very limited in the U.S. market), changes to benchmark rates, such as a shift from LIBOR to SOFR, could introduce basis risk if the terms of the liability and floating swap do not change at the same time or if replacement rates and/or spreads differ. For example, an SF transaction with fixed-rate assets, floating-rate LIBOR-based liabilities, and an interest rate hedge could see basis risk to the extent the floating rate payable to the securitization trust switches to SOFR prior to the floating rate payable on the notes. This could occur to the extent that the hedge and the notes have different trigger events causing a switch from LIBOR to another benchmark rate. This change could also expose the special purpose vehicle issuer to making an upfront payment or series of payments senior in the priority of payment, the impact of which on the ratings would need to be assessed. Finally, a lack of agreement on a new post-LIBOR rate could lead to hedge termination where the issuer could owe termination payments to the counterparty, but we currently view this possibility as remote.

Acknowledging this potential basis exposure, ARRC has aligned its recommended spread adjustment to SOFR with ISDA by selecting a five-year median SOFR-to-dollar LIBOR difference as the preferred method. By bringing cash markets, as represented by ARRC, in line with derivative markets, as represented by ISDA, this should help market participants better adjust to alternate rates and minimize disruptions. Market participants are awaiting ISDA's scheduled January 2021 release of its amended 2006 definitions containing final adjustments to fallbacks.

Questions Remain On The Transition From LIBOR

Given how ubiquitous LIBOR is among legacy securitization products, there remain many questions and issues we will be following to determine whether and where there may be potential rating impacts. These include:

• To what extent will key transaction parties to existing securitizations adhere to ARRC's recommended fallback language, spread adjustments, asset agreement fallbacks and spreads, and overall transition timetables?
• How far in advance of the scheduled December 2021 LIBOR cessation date will transaction parties that have the contractual right to select a new benchmark rate on the liabilities make their intentions known to the market?
• To what extent will investors agree, particularly where unanimous consent is required, on amendments to transactions documents that govern interest rates?
• Will the proposed New York state legislative solution be successful and, if so, when would that occur?
• To what extent and when might central banks in the U.S. or the UK or the FCA issue a public statement that LIBOR is not a representative rate before December 2021: as early as the end of 2020 or early 2021?
• If the LIBOR methodology is changed by FCA to assist "tough legacy" contracts under UK law, will U.S. market participants be able to use a new "synthetic LIBOR" rate or will it lead to disputes among transaction parties and investors for legacy U.S. securitizations?
• Will replacement rates besides SOFR emerge, particularly those with a dynamic credit component, and be proposed for inclusion in U.S. securitizations?

Being Prepared Is The Primary Challenge Surrounding The Transition From LIBOR

Given the inherent uncertainty surrounding the transition away from LIBOR, it is incumbent on market participants with direct or indirect LIBOR exposure to prepare for a variety of contingencies and outcomes. This may include updating transaction and/or underlying asset documentation, initiating customer outreach on new rates, enhancing interest rate modeling, preparing for potential legal challenges with new RFRs, developing alternate plans for rates in case term SOFR is unavailable in 2021, verifying vendor compliance, managing global exposures to the extent LIBOR is used globally, monitoring all triggers in recent transactions containing ARRC fallbacks, contemplating new issuance in light of ARRC timetable guidelines for 2021, and anticipating possible paths arising from some of the questions above, among many other considerations.

For a large portion of outstanding securitizations, liability documents call for a transaction party, such as the servicer or trustee, to determine a replacement rate for LIBOR. For these parties with discretion over new rates, announcing to investors, rating agencies, and other market participants their intentions on replacement rates and/or adjustments as early as possible will help contribute toward an orderly transition.

Appendix

Related Research

• Will the Expiration of LIBOR Affect The Credit Quality of U.S. Public Finance Issuers?, Feb. 26, 2020
• Credit FAQ: How To Say Goodbye To LIBOR Without Creating Market Chaos, Sept. 23, 2019
• With A LIBOR Phase-Out Likely After 2021, How Will Structured Finance Ratings Be Affected?, Oct. 19, 2017