I wrote back in the early days of 2020, or as we call it now, the “Time Before,” that we thought it made sense for key market participants to consider an early move to SOFR pricing, not just as the backstop but as the interest rate of the loans. Frankly, we were thinking about the SASB market, and we thought that would be indeed well received by the investor community.
Of course, at that point we were blissfully on the broad, sunlit uplands of an amiable 2020, and many of us had the bandwidth to deal with this looming, but not immediately threatening, problem. Then COVID-19 was ushered in (cue dastardly villain stage left) and everything largely stopped. Funny how questions about mortality can sort of drive heretofore fascinating controversies (okay, maybe actually not so fascinating) around alternate index rates out of mind.
While the ARRC and a coterie of committed technocrats in and out of government and major institutions, did continue to beaver away on LIBOR conversion, many institutions, and perhaps an even larger percentage of the smaller ones and non-bank lenders, have not had the bandwidth to think hard about such things. Consequently, in the real world of mortgage finance, progress has slowed for all these many months.
Now, it’s not been nothing. The SASB market has largely embraced the SOFR backstop. The GSEs have put a stake in the ground and have begun to offer SOFR priced loans this September, with an expectation that they will no longer buy or sell LIBOR priced assets starting in January 2021. The ARRC and ISDA have reconciled many of the annoying variances (in particular the timing of the LIBOR transition and the amount of the spread adjustment) between the cash and the derivatives market and both have continued to nibble away at areas of uncertainty in the existing guidance, largely dispensing with alternatives to the hardwired approached, while differences between syndicated, securitized and bi-lateral contracts have been mostly reconciled (other than the adoption of Daily Simple SOFR in the business loan market, which we do not expect to see mirrored in other product markets).
Waiting until the tail end of 2021 to begin actively underwriting and pricing mortgage loans sounds like a really bad idea. First, and let’s get this out of the way, if anyone out there still thinks that this transition is not going to happen promptly in early 2022, then I’ve got a bridge I’d like to talk to you about. If you’re waiting for term SOFR to come out so that you won’t have to move twice from LIBOR to compound SOFR in advance to term SOFR, well, you’re not wrong, but that’s a scary, unhedged bet.
Waiting any longer to make a SOFR loan has… negative externalities. Even if you are starting or have already started to build out back-office underwriting, develop new policies and procedures using SOFR and are dealing with a legacy book, rewriting documentation and training one’s staff (an unappreciated heavy lift) and re-educating the borrower community, none of that is really going to gel until you actually make a loan. It’s once you actually endeavor to make a SOFR priced loan that you’ll see what works, what doesn’t work, how it works and it has to be sold to the borrower community.
Waiting is a bad idea. You should fear a chaotic end to LIBOR.
So, get it started now? Test drive the process before it actually needs to work and begin to work towards scaling your processes across all your platforms before you have no choice. This is an airplane you don’t want to fly while you’re building it.
With protective cover of the GSEs already in the space with some increasing borrower awareness due to the vast scale of the GSEs’ footprint, a deal could get done now. A handful of first moving banks are already likely to launch SOFR lending within the next two months. Join the party!! In the SASB market, for both economic and, shall we say… political reasons, large sectors of the investor marketplace are ready and indeed anxious to start to book SOFR-priced loans. No one in the investor community wants to be an antediluvian knuckle-dragger here clinging to his or her guns, religion and LIBOR.
We’re locked, loaded and ready to go. The ARRC language is stabilized, you can pull SOFR compounded in advance (preferably) off a Bloomberg screen and, of course, there’s bragging rights for going first, right?
But one tiny little cavil around my enthusiasm for being a first mover that I feel I need to share; just a little, itsy-bitsy concern for our market that’s characterized by fairly long-tenured paper. I almost hate to mention it, I feel like such a quibbler here; not on side, but here’s the problem. LIBOR has a credit component and SOFR does not. When credit conditions and perhaps more importantly liquidity conditions, deteriorate, LIBOR moves up and SOFR will not budge and indeed might actually go down.
I’m sure I’m teaching Grandma to suck eggs here, but I’ve looked at the LIBOR series over the past 30 years, when it’s been as high as 8% and as low as the close to nil levels where it currently sits, and it’s moved periodically relatively quickly. Between 1992 and 1995 it moved up 300 bps. Between 2003 and 2007 it moved up over 400 bps. And, from the end of the Great Recession until 2019 it was up 250 bps and now it’s nil again.
The ARRC has been aware of this issue and has essentially punted on a solution, beyond a semi-official suggestion that lenders should just get over it and price in a cushion ab initio to take this variability into account. Really?
The SOFR mechanics already contain an index adjustment to reflect the difference between LIBOR and SOFR which, if LIBOR was determined to be non-representative today, would be about 11 bps. So already in this bright new world, a borrower’s cost of funds is composed of three components: (i) the SOFR index, (ii) the index adjustment amount plus (iii) the economic spread. I struggle to see how the possibility, not the certainty, but just the possibility, the potential increase in the price of floating rate financial assets down the road can be priced into a deal at inception in a way that satisfies both the borrower and the lender. I’ll be completely gob smacked to find out that the borrower community would be fine paying this additional increment. As the borrower community will struggle to warm up to the notion that LIBOR plus 175 is now SOFR plus 175 plus 11, it’ll be a very heavy lift to ask them for plus something else. Plus, something else? In any active market that I’m aware and indeed if you were able to bake in such an incremental spread, I suspect it would be cold comfort when you subsequently find yourself with a mispriced asset in future.
While we don’t know where interest rates are likely to go after such long period of interest stability, dimes to dollars somewhere during the tenor of your average floating rate commercial loan made in the next few years (with extensions that could push its ultimate maturity out as long as seven years or more), credit conditions (or liquidity conditions) will deteriorate and in the alternate universe where LIBOR was still available as the index, loans would provide a compensating higher yield to the lenders because of that.
How will you be feeling about that when it happens (and it will)? I guarantee you that even if you added 10, 20 or even 30 bps to your coupon ab initio, based on a model that demonstrates in the pretty world of pure mathematics that even if short term interest rates explode later, you would be adequately compensated, it won’t feel good when it happens. And you’ll notice when interest conditions tighten in the out years, even if we don’t have LIBOR to kick around any longer (please, Lord, tell me we’re not going to have to deal with a zombie LIBOR). The spread of newly originated paper will be materially higher to take this into account and of course we will be able to watch things like Prime, ICE Bank Yield Index, Ameribor, etc., to remind us how bad we feel about the difference between the rate on the pre-existing paper and the rate at which you can put money to work today.
There is, of course, a solution and that’s to shorten the tenor of the floating rate market. The other possibility, of course, is to include some sort of repricing mechanic in connection with extensions. Generally, extensions are tied to a simple set of mechanical check-the-box deliverables such as continued debt service coverage, debt yield and of course a payment of a modest fee. Maybe in the future, these should be repricing events. How that repricing or mark-to-market would occur and what its impact would be on the floating rate market is hard to tell, but it’s probably not, shall we say, good. But if we don’t do this, there’s a realistic possibility that holders of floating rate debt may see a material transfer of value from borrowers to lenders over the next decade and I don’t know how we think about that.
Okay, let me get a hold of myself here. SOFR is going to be here soon, it’s how we’re going to price floating rate debt, and this absence of a credit component in the index is an insolvable problem (just ask the apparatchik, who convened a group that might develop a credit sensitive SOFR spread adjustment, only to give up a few days ago). In the end, that’s just the world we’re going to live in. It’s unlikely that something is going to magically occur in the next 14 months to make it better. So, I guess I have convinced myself that it’s time to suck it up and get on with things and we might as well get our first SOFR loans done now, right?