We’ve written before about our anxiety regarding the fact that SOFR does not really seem fit for purpose to support commercial mortgage lending or indeed any cash product.  (The nonsense about charging interest in arrears should have been a tell, to be honest.)  Of course, the real problem is the absence of a credit-sensitive component to the new index, particularly in this time and place.  That strikes me as almost fatal to the ambitions of the ARRC to remake the market in its image.  SOFR is an open invitation for value transfer from lenders to borrowers at a time when inflation is closer than the horizon and an inexorable climb in the short end of the yield curve is most certainly on offer.

The ARRC made an effort to wrap its mighty bureaucratic mind around preserving a credit component to SOFR but ultimately gave up.  Subsequently, ARRC leadership has shown what appears to be an obdurate insistence that a credit-sensitive rate or an add-on is simply not achievable, is, in fact, not terribly important and is, perhaps, really not a very good idea in the first instance.  Moreover, and a bit surprisingly, they put out that the borrower community would never accept an index with a credit-sensitive component.  Huh?  That’s odd since we’ve been blithely using LIBOR for the past 40 years; and moreover, I didn’t know the borrower community had a veto here.  ARRC has been telling the market that they should get over the fact that SOFR will not be credit sensitive and get on with it.

Not So Fast, as Lee Corso might say (apologies for the football allusion).  Notwithstanding what feels like a bum’s rush to get all financial markets to move to SOFR by January 1 of next year, while LIBOR continues to be published and presumably representative through June of 2023, we are not done.  Ameribor continues to gain traction, particularly amongst the sub-regional banking sector where Ameribor, which indeed includes a credit component, is viewed as more representative of those banks’ cost of funds (and they’re right).  The bank regulators continue to tease us by being very careful not to say that SOFR is the only choice available to the market to replace LIBOR.  Moreover, this past week, LSTA published an option to include credit-sensitive rates in their benchmark fallback waterfall, but neither specified a single rate to use nor where in the fallback waterfall that would occur.  It could be the first priority (And it should be.  Maybe not?).  Note the LSTA nodded toward AFX’s Ameribor Index, the IBA’s Bank Yield Index, HIS Market Credit Rate and Bloomberg’s Short-Term Bank Yield Index (BSBY) as runners, so long as those rates are IOSCO compliant (and indeed, we should expect that they will be).

Nonetheless, the ARRC and some constituencies in the US regulatory establishment have been pressing back hard against any suggestion that it’s not simply SOFR all the time, suggesting that we should stop toying with LIBOR and stop worrying about the pesky absence of a credit component.  All of this of course is happening against a backdrop of a cash market that is strongly signaling that someone will have to rip LIBOR from its cold, dead hands.

I continue to think that this is an important issue and, to the extent we reflectively continue to add ARRC’s SOFR fallbacks into our documents, we may regret the choices being made today when we discover that we’ve created a cohort of financial assets that may trade at a discount when the market embraces a new, better and credit-sensitive underlying index.

This has been bothering me for a while… but I think there are a couple of fixes.  First, and perhaps most concrete, you could just go with one of the new credit-sensitive add-ons in lieu of ARRC’s 11.448 bps adjustment for one-month LIBOR.  But if that’s still taking too much of a bet on future market conditions, we have an alternative.  Why not condition use of any of the SOFR rates (in the SOFR waterfall) on a determination that that choice is consistent with market conventions?

To the extent that many of these financial assets may not finally abandon LIBOR until long into 2023, there’s a lot of time for markets to evolve.

To that end, we have proposed amending both the definition of the “Benchmark Replacement” and “Benchmark Replacement Adjustment” to condition the choice of term SOFR, compound SOFR or any of the alternatives to a determination by lender (or servicer) that such Benchmark Replacement or Benchmark Replacement Adjustment is consistent with…

any evolving or then prevailing market conventions for determining an unadjusted Benchmark Replacement or a Benchmark Replacement Adjustment for US dollar denominated securitized commercial mortgage loans at such time.

The addition of this language will facilitate the ability of the Designated Transaction Representative (DTR) to move from LIBOR to the best then-available market index when the time comes as the fifth bucket in the ARRC waterfall is:

the sum of (a) the alternate rate of interest that has been selected by lender as the replacement for the then current Benchmark, given due considerations to the then prevailing market conventions for determining a rate of interest as a rate of replacement for the then current benchmark where US dollar dominated floating rate commercial mortgage loans included in securitizations at such time and (b) the Benchmark Replacement Adjustment.

            This will allow the DTR to embrace the then-best market convention.  That’s the same language that is included in the Benchmark Replacement Adjustment.  Not only can the DTR select the index, but it can also select the Benchmark Replacement Adjustment (and we’re already hearing quite a bit of support for IBA’s Bank Yield Index for such purpose).

While we’re of a mind to fix what is broke, let’s spend a moment thinking about the Designated Transaction Representative.  The market has moved to pick institutions like Note Administrators to be the DTR and for the DTR to work through benchmark transition and benchmark conforming changes.  We’re doing some of that now and it’s an illuminating exercise.  These institutions, while extraordinarily good at doing what they do, don’t do new things comfortably that are not hardwired or mechanistic, and this, in some significant measure, is one of those things.  There are elements to the process of swapping index rates that are simply not Newtonian.  Only by way of example, selecting the Benchmark Determination Date is not without a significant element of judgment.  But, if we continue to choose these institutions as the DTR, there’s one thing for sure that should be pulled back to a participant with “skin in the game,” and that is any option for early adoption (if one is indeed included in the documents).  The question of when a transaction moves from LIBOR to SOFR or indeed any other alternate rate, is not one susceptible to determination in a legalistic or mechanistic way.  It’s a business-driven choice, so at least in this respect, the DTR should be required to take direction from the collateral manager, the preferred shareholders or other transaction participants vested in the structure.

Listen, we continue to build this airplane while flying it and these choices are complex and the environment in which these choices are made is continuing to evolve.  Moreover, even applicable governmental authorities have seemed to struggle with staying the course as evidenced by the delay in LIBOR’s hard cessation from January 2022 to June 2023.  I’m not comfortable thinking that this little drama of ours is over and that we have found an absolutely suitable pathway to the post-LIBOR world. As a result, baking in some optionality to decisions that might not be made for months, if not more than a year, seems to be an unalloyed good thing.