July 24, 2018 - On Thursday, July 19th, the Alternative Reference Rates Committee (ARRC) went public with its SOFR status and plans. Specifically, the ARRC held a public roundtable where it unveiled an “indicative indicative” term SOFR rate, announced that it was ahead of its transition plan, explained many of the Greek letters in ISDA’s public consultation on LIBOR-SOFR spread adjustments, and updated the “LIBOR fallback” plans for FRNs, business loans and securitizations. We recap the revelations below.
Randal Quarles, the Fed’s Vice Chair for Supervision, kicked off the Roundtable by offering a notable statistic: Based on its data, the Fed sees around 6-7 transactions daily for one- and three-month LIBOR; they see 2-3 per day for 6-month LIBOR and less than one per day for 12-month LIBOR. This underpins ARRC’s view that there may not be enough actual interbank lending to support a robust LIBOR.
Next, Sandie O’Connor, Chair of the U.S. ARRC, discussed SOFR’s strong start and – of particular interest to loan denizens – unveiled an “indicative indicative” forward looking three-month SOFR (see slide 6 of her presentation). To wit, the indicative three-month SOFR ranged between 1.9% and 2% during May and June. At the same time, three-month LIBOR was in the 2.3-2.35% context, suggesting a spread differential of roughly 30-40 bps. As SOFR enthusiasts might know, overnight SOFR (a secured, risk-free rate) has been trending above overnight LIBOR (an unsecured rate), thus raising questions about the relationship between their yields. Loan market participants, who rely more heavily on one-month and three-month rates, were eager to see how term SOFR actually looked. And now we know.
The next items that may particularly interest loan-ers were the discussions around LIBOR fallback language in cash products. In other words, if LIBOR were to go away, to what rate does your loan, floating rate note or securitization “fall back”. This is of critical importance because many cash asset classes had negligible fallback language and had tenors that extend past the potential end-2021 LIBOR cessation date.
At the ARRC Roundtable, the cash working groups – including the business loans group co-chaired by the LSTA and ABA – discussed their progress on creating actual fallback language. Before providing the status update, it is important to note that fallback language has three main components: 1) What is the trigger (e.g., what event “triggers” the conversion from LIBOR to a new rate)? 2) What is the fallback rate (e.g., is it SOFR or some other rate)? 3) What is the “spread adjustment” between LIBOR and the new reference rate?
FRNs: The FRN working group is proposing a prescriptive “hardwired” approach. In other words, today the group is defining the triggers, fallback rates and credit spread adjustments. But because many of these rates do not yet exist, the FRN proposal builds in “waterfalls”. If a certain rate does not exist at the trigger date, then parties go to the next potential step in the waterfall. As delineated in their slides, the FRN group defines the first potential replacement reference rate as X-month SOFR selected/endorsed/recommended as the replacement for X-month LIBOR by the relevant governmental sponsor. If this rate doesn’t exist, the waterfall goes to an interpolated SOFR determined by the calculation agent. If that rate doesn’t exist, the waterfall goes to SOFR. If SOFR doesn’t exist, the waterfall goes to an alternate/substitute/successor rate selected by the relevant governmental sponsor. If even that rate doesn’t exist, the waterfall ends at the Overnight Bank Funding Rate (OBFR), which is a rate that exists today.
But that is not it. In addition to the reference rate, the fallback needs to add a “base rate modifier” to make up the difference between LIBOR and the new reference rate. A similar waterfall ensues. By taking this approach, the FRN group hopes to “hardwire” a solution today, even though not all the necessary components exist today.
Syndicated and Bilateral Loans: This working group – co-chaired by the LSTA and the ABA – explained that there were more nuances in the loans space. Working group members agree that the hardwired approach makes sense at some point. However, because loans are easily amendable – in contrast to FRNs – some working group members recommend an “amendment “approach, similar to what occurs in the loan market today. Others advocate for a “hardwired” approach similar what the FRN group is doing. We lay out the approaches and relative merits below.
As the presenters explained, the loan “hardwired” approach is similar to the FRN approach, but also benefits from the flexibility of loans. Specifically, the loan hardwire starts with term or interpolated SOFR and a base rate modifier to capture the difference between SOFR and LIBOR. If either of these are not available, then the “hardwired” approach would flip to the amendment structure similar to what we see today. The benefit of this approach is that, assuming that the new reference rate and modifiers are in place, the credit agreements can flip seamlessly and do not require a time consuming amendment/vote mechanism at the time of LIBOR cessation (which might be a volatile time).
But other presenters noted that loans, unlike other asset classes, are uniquely repayable and amendable. They ask why we should hardwire today when there will be more information on SOFR and base rate modifiers soon, and loans may go through at least one amendment/refinancing cycle prior to any end of LIBOR. In turn, a group feels that the current “amendment” process works; in this situation, the agent/borrower identifies that LIBOR has ceased to function, identifies a new reference rate (and potentially a spread adjustment to capture the LIBOR-new rate differential) and the required lenders have an objection right. This avoids locking contracts into rates that do not currently exist and whose behavior is not yet known.
Both arguments have merits, and hopefully the broader market will have the opportunity for input soon. Stay tuned.