So near, SOFR: The race to find a Libor alternative

It’s not that long ago that the Alternative Reference Rates Committee (ARRC) announced, to the surprise of many, that it could not promise a Secured Overnight Financing Rate (SOFR) term rate by the end of 2021. Consequently, some market participants, especially in trade and working capital finance that work off term rates for forward discounting, had their heads turned by the likes of the Bloomberg Short-Term Bank Yield Index (BSBY), which is a term rate. The subsequent re-focus on the delivery of a SOFR term rate was triggered by the unintended spark of life given to BSBY (among others). And now we have one. The Chicago Mercantile Exchange has been designated by the ARRC as the provider of a SOFR term rate, and that SOFR term rate is being jostled into place.

We've gone from anything but Libor to SOFR first

We may always have had a multiple alternative rates' environment post the demise of Libor, but these events have accelerated the concept into market consciousness. In an effort to get ahead of this, the official sector has morphed its stance from a call on market participants to use “anything but Libor” to one of “SOFR first”. The official sector continues to push SOFR in arrears as the preferred version, as it offers the purest compilation (no guesswork). But in an effort to ensure that all participants are catered for, the provision of term SOFR counters any excuse that some market participants might have had to delay transition away from the use of Libor.

But big issues remain. The most pertinent one centres on the treatment of the (bank) credit adjustment spread. Essentially this is the basis point cost that banks need to pay to get access to market liquidity, over and above the risk-free rate. This is incorporated in Libor (and BSBY), as these are in part compiled from where banks can manage to print commercial paper. But, it is not contained in SOFR, as it is by design close to risk free. So one big question is what market participants should do about this if anything. There are a number of options, as follows.

• Use the ISDA fixed Credit Adjustment Spreads on top of SOFR. As you can see in the chart above, we employ the 5yr median spread between SOFR and Libor as calculated by the International Swaps and Derivatives Association (ISDA) to be employed in the calculation of fallback rates. So for example, that spread is 26.2bp for the 3mth rate. Simply adding this to the underlying SOFR rate then takes care of the bank funding cost. Effectively here the bank gets paid a fixed credit adjustment spread, but then takes its chances on the funding markets. Today it's' happy days' for banks, as bank funding rates in the market imply a single-digit credit adjustment spread. But there will be times when the funded credit adjustment spread could be much higher (as is likely given that the 26.2bp is a median calculation). This is the solution being employed for legacy products being transitioned from USD Libor, set for mid-2023.

• Use the market Credit Adjustment Spreads. In the example above, we employ the basis that we observe in the market between SOFR and Libor, which is observable for every tenor right out the curve. And even if or when Libor really ends, it could be replaced by a basis between SOFR and BSBY. This again would be a fixed spread applied on day one or at transition. It will vary over time as it is a traded number that will be responsive to market circumstances, widening during times of funding stress and narrowing when funding spreads are tighter. It should still be a no-arbitrage solution, one that breaks even as opposed to a medium-term average for the spread (which would be where the longer tenor spread converges towards). This is an alternative that players can also use to adjust from legacy Libor for SOFR plus the market basis. I

• Use SOFR with no explicit bank spread adjustment. As you can see in this example, we use SOFR (term or compounding) as our reference rate, and on top of that we add the individual borrowers' unique credit spread, without any explicit adjustment for the cost of bank funding. The keyword here is “explicit”, as in reality there would be an implicit spread applicable for all borrowers to account for the banks' funding costs. If the borrower's credit spread is fixed and the only variable is the movement in the SOFR rate, then we can assume that there is a fixed spread within the corporate credit spread to account for the cost of bank funding. If nothing is made explicit, then assuming the market spread (e.g. 20bp in the 5yr currently) as the size of that spread is appropriate. This is likely the simplest solution. Basically, it accepts a higher unique corporate credit spread for all borrowers, to leave all-in economics unchanged.

Let's pause here to assess what all this means so far. What is happening in all of these three alternatives is an implied fixed versus float relationship between the borrower and the lender with respect to the credit adjustment spread. In all the cases above, the borrower is paying a fixed credit adjustment spread while the bank faces a floating credit funding cost. Right now that is a positive carry trade for banks as funding costs are ultra-low, in fact, there's a single-digit credit adjustment spread in the market. While this should average out over time, it also implies a certain vulnerability for banks that have to take the chance they don’t run into a period of deep negative carry should funding rates spike or creep up, reflecting credit pressures.

There is always the alternative of fully embracing the incorporation of a floating back spread adjustment. That alternative is to transform to a float-float relationship between borrower and lender. Here we come to the next set of options.

• Use a Bank Credit Spread as provided by a third party. Some third-party providers are explicitly providing this, but it can also implicitly be derived as the difference between any of the bank rates and the SOFR rate. For simplicity, for example, IHS among others provides this as a stand-alone spread that can be employed as an add-on to any risk-free rate. Crucially this is a floating spread, which is also theoretically more appealing. So when there are funding pressures on banks for whatever reason, it gets reflected in a wider spread, and in consequence also into the all-in funding cost for the borrower during that period of stress. The converse also applies, with borrowers benefiting if bank credit spreads narrow.

• Forget SOFR and just use BSBY. Using BSBY, as you can see in the chart above, is broadly equivalent to the use of SOFR plus an IHS-style credit adjustment spread. Both of these elements are variable, in the sense that SOFR will move as the Fed adjusts policy over time, and the credit element will move in relation to funding pressures for banks. So the likes of the BSBY, ICE, Ameribor, among others, provide all-in rates that are term in nature, they include the bank credit element and in fact, look and feel a lot like Libor. A positive here is continuing with something that feels quite familiar, but a negative is the risk there could be significant volatility in the rate, or price discovery is made difficult during periods of extreme volatility.

These bank rates are in part anchored by where banks print commercial paper (just as Libor is today). The warning from the official sector is that these bank funding markets could dry up during periods of crisis, leaving the compilation of the reference rates somewhat compromised. While commercial paper is not the only anchor for bank rates, the official sector would clearly prefer for participants not to get too enamoured by these alternative bank rates, even if they are backstopped by compilations that include where banks take in deposits.