First, the ARRC, playing Charlton Heston, playing Moses, brings down from on high the ten commandments of SOFR and lo, we were sore afraid and with veneration, professed we had no God but SOFR. A solution of sorts to a somewhat self-inflicted problem. As we have observed before, we continue to think the solution to the problem of bankers diddling LIBOR is to punish bankers and shore up the system to make it more robust and not to blow up trillions of dollars of transactions and 40 years of precedent. But that train has left the station.
However, the ARRC, having thought about the current state of affairs some, concluded an 18-month delay in the extirpation of US LIBOR was okay, and the cash market was relieved to the extent they were thinking about it at all.
Nonetheless, the finger-waggers in the prudential banking regulatory establishment have made it clear that the banks better start using SOFR right away and stop this untoward dalliance with LIBOR. But LIBOR use remains a stubborn reality. Hmmm. I wonder why?
We have observed in this commentary before that delay in the end of LIBOR would certainly be appreciated and the December announcement was darn near essential to avoid a chaotic exit from LIBOR on its originally scheduled expiry date of December 2021. But this is really an intermezzo in the sad progress to SOFR, and the powers continue to insist it will replace LIBOR in the middle of 2023, come hell or Red Sea high water (don’t worry, we have now exhausted my biblical metaphors).
The existential problem, of course, is that SOFR is and will continue to be a flawed index because it lacks a credit component. Delay won’t fix that. This means that there is a possibility, no actually, a likelihood, that significant value transfer from lender to borrower (and, I guess, theoretically from borrower to lender) will occur when economic stress or dislocation occurs. In that alternate universe where LIBOR continued to exist, LIBOR would have spiked on risk and dislocation, compensating lenders for enhanced risk, whereas our new friend, SOFR, may actually decline as the risk-off safety trade balloons the demand for Treasuries. We’ve asked why this is not as big a problem as we continue think it is, but increasingly we do so sotto voce so as not to offend our betters.
Giving credit where credit is due, the ARRC tried to figure out how to add a dynamic credit component to SOFR, but eventually threw up their short little regulatory arms and said there was nothing to be done and we should all learn to live with it.
Not so fast. Bloomberg and Markit are racing to develop a SOFR credit sensitive add-on and of course, the American Financial Exchange, the folks behind Ameribor, continues to beaver away on a competing alternate rate, which does have a robust credit component (you know, and I’m just saying, Ameribor is quite popular in the regional banking market where they actually know something about credit. They continue to embarrassingly insist that SOFR is not a good match for their liabilities. Pesky little banks’ bankers.) Note the political nicety of both Bloomberg and Markit calling their alternate index rates “add-ons” even though they really could be recast as completely freestanding alternate rates, but no one wants to offend the ARRC. These “add-ons,” are being developed across all the principal tenors of traditional LIBOR based on a wide basket of actual market quotations, including both interbank trading (admittedly thin) and a variety of other actual trades in actual markets. Hmm, doesn’t sound like a bad idea, does it? And, of course, you will be able to grab these quotes in real time off your screen just like…LIBOR.
It seems like this process of coming up with a credit-sensitive index is gaining momentum and now indeed it is conceivable, if not likely, that one of these new indexes or add-ons will compete robustly for market share soon.
However, to quote those two mid-20th century intellectuals, Laurel and Hardy, “It’s another fine mess you’ve gotten us into.” If a credit-sensitive alternate to SOFR gains material traction, we’ve got a new and difficult legacy problem on our hands. After the better part of two years of being told to embrace SOFR, SOFR has begun to colonize our documentation. At this point it is hardwired as the alternate rate in billions of dollars of SASBs, the GSEs are now directly lending on SOFR, and a significant percentage of all other largish commercial mortgage debt now has hardwired SOFR into notes and loan agreements, awaiting the inevitable demise of LIBOR.
The SOFR benchmark replacement mechanic is the sort of thing you would expect a horse built by committee to be. It contains a mechanic to “fix” the problem that SOFR has, over many years, been lower than LIBOR, but it does this by simply adding bps to base SOFR.
But here’s the rub, the way the ARRC language works, that adjustment is most certainly to be the difference between LIBOR and SOFR over a five-year look-back period. This will be set upon the occurrence of a “Benchmark Transition Event” which is going to happen very shortly when the ARRC confirms that the LIBOR administrator will cease to provide the LIBOR benchmark in June 2023.
When that happens, it will freeze the calculation of the difference between LIBOR and SOFR permanently at roughly 11 or 12 bps for one-month tenors. That is, like, forever.
The SOFR benchmark replacement mechanic in our loan documents now will direct the so-called Designed Transaction Representative (the poor bastard in charge of figuring out the new interest rate of the loan) to select SOFR plus the aforementioned fixed forever Benchmark Replacement Adjustment.
So, what happens when the market, or significant portions of it, sensing the error of its ways, decides to move away from this version of SOFR to embrace one of the new credit-sensitive SOFR add-ons? We’ll then be in a situation where new add-on loans will be protected with extra spread for extra risk while legacy loans, and here we’re talking about loans which now default to or will default to the current version of SOFR, do not.
To be clear, the current version of the LIBOR replacement mechanics does not permit a lender or a Designated Transaction Representative to add or substitute one of the new dynamic add-ons to base SOFR.
What this means is that we’ll be looking at a situation where loans converting to the current version of SOFR will trade at a discount to new paper with a more dynamic economic spread. That’s not something that might happen, that’s something that will happen.
So, we should try and fix this now, right? Representatives of the ARRC have said pretty clearly that it can’t be done and more recently have said that the resistance of the borrower community will prevent any re-introduction of a credit risk component into our loan documents. Huh? Done deal? They’re in charge? Everyone lived with the credit-sensitive LIBOR for 40 years and it seemed to work pretty well. What’s wrong with actually replicating something that really did work? What’s the compelling evidence of the unwillingness of borrower-side market participants to agree to actually replicate the economic consequences of LIBOR by adding a credit component into the new SOFR structure? How bad can it be? And, by the way, it’s conceivable that come the summer of 2023, these credit risk add-ons might actually be inside the 11 or 12 bps of the static SOFR adjustment and might that be an attraction to borrowers. In any event, we won’t know if these nebulous borrower constituencies will win the day, unless we try. The thought of adding another alternative to our already mind numbingly long and complex alternate rate mechanic is a tad daunting, but what else can we do? The prudential regulators could give us some help here by embracing the inclusion of dynamic spread adjustment mechanic (which should, if you think about it, make a great deal more sense to the prudential regulators than a static adjustment). Of course, it would be delightful if ARRC would step away from its dogged refusal to consider alternatives and bake more flexibility into the currently generally approved ARRC mechanic. Will those things happen? Maybe, but not without the industry pressing its case. Our industry should re-engage with the ARRC and seriously discuss the applicability, reasonableness and robustness of some of these new alternate index add-ons. We should consider adjusting our current ARRC transition language to make it at least possible to embrace a new dynamic rate add-on which might gain currency in our markets between now and the summer of 2023. We should engage with the New York legislature and our gloriously elected representatives in Washington to make sure that whatever safe harbor for transitioning out of LIBOR gets passed, it does not mechanistically hamstring us by tying indemnification only to the use of the current version of SOFR.
All this needs to get discussed. The good news is that there is some light in the tunnel beyond the straightjacket of an un-credit responsive index rate. While certainly everyone is tired of dealing with this issue and wrestling with its complexity, we need to get back into the big muddy and figure this all out.
So, when you’re tired of worrying about the pandemic and need something else to think about, I urge renewed consideration of the appropriate index for floating rate paper starting in a theatre (of the absurd) near you soon.