Libor: It’s not you, it’s SOFR

Since 25 January 2021, all new derivative contracts that reference Ibors will transition to ISDA fallback rates, once/if those Ibors cease to exist. At the same time, the ISDA protocol window remains open, so that derivative trades referencing Ibors set before this date (all legacy contracts) can receive the same treatment. For corporate liability managers looking for an efficient route from pre-Ibors to post-Ibors, this presents a clear solution.

All new derivative contracts that reference Ibors will transition to ISDA fallback rates, once those Ibors cease to exist

There are two key nuances to be aware of though:

The fallback rates themselves are set in arrears, as they are calculated as the risk-free rate plus a fixed spread (the 5yr median of it). The spread by definition is fixed, so the arrears characterisation comes from the risk-free rate.

Some sizeable gaps continue to exist between the risk-free rates and the respective fallback rates. For example, USD 3mth Libor is now at sub-20bp while the 3mth USD fallback rate is at 34bp; in excess of a 15bp differential (currently).

On the first point, this is an area of contention if the swap is an overlay to a loan where the rate is set in advance, as it would be when referencing 3mth USD Libor plus a credit spread. On the second point, any differential between the Ibor rate and the fallback rate is netted out, to the extent that the same spread is applied to the underlying loan.

When we look at 3mth USD Libor specifically, we find that the historical median spread is currently at 26bp (over and above the comparable SOFR). That spread, along with all other applicable risk-free to Ibor spreads, is due to be fixed at any time. There was never a specific date set, but consultation is complete, and all is in place to have the spreads fixed.

The advantage of delaying the fixing of the SOFR to 3mth Libor spread is that it slowly eases lower as we progress over the coming months. We know this as it is calculated as a 5yr median, and so we have most of the applicable data. But also, we know with reasonable certainty where SOFR will trade in the coming months (as the Federal Reserve has the funds rate on hold at least through to 2023, or so they say.)

Delaying the fixing of the SOFR to 3mth Libor spread could see the fixed spread fall to the low 20bp area if we wait long enough. That would reduce the noise around the differential between the market spread and the fixed spread.

We can't wait too long

If the SOFR to Libor spread is fixed today it would be 26bp. Waiting would allow the spread to drift lower. Better to fix early, and get that information set and in the market

At the same time, we can’t wait for too long. This is because all spreads are likely to be fixed at the same time when the end to the respective Libors is officially announced. For the likes of GBP Libor, this needs to be well before the end of 2021, as that is when GBP Libor ceases to exist. Even though USD Libor continues through to mid-2023 for legacy product, the spread is likely to be fixed at the same time.

So for example, should the SOFR to 3mth Libor spread be fixed at 26bp, that would be a sensible spread to also apply to loans, especially those that are overlayed with derivatives that will come under ISDA-style fallback rate treatment.

All things considered, it likely makes sense to bite the bullet and fix the spread(s) early.

Any contention on the size of the spread (currently 26bp) is far lower for longer tenors. So for example, the market spread between SOFR and Libor is just 15bp for a 3yr tenor – quite a difference from a fixed 26bp. But the market spread for a 10yr tenor is 22bp, and 24bp for a 30yr tenor. These are much closer to a potential fix at 26bp.

For players that execute derivatives referencing USD 3mth Libor today (and since 25 January), these are the terms that are being signed up to - there is no choice in the matter. An alternative is to execute new business in SOFR or fed funds referenced swaps.

For players that have legacy positions, there is a choice

For players holding Libor referenced derivative legacy positions who don't sign up to the ISDA protocol, such positions could be marked to market, while simultaneously setting a new trade with a non-Libor reference.

Or, something else. But that “something else” needs to be sorted well before mid-2023 (when the ability to reference legacy product to USD Libor comes to an end).

When we switch our focus to the loans market, we find there is no equivalent fast track route to fallback rates. Existing Libor denominated loans can be left as is through to mid-2023, but from then onwards an alternative to the Libor reference needs to be in place, and conversations on how to make that happen clearly need to have happened well ahead of this. The delay to mid-2023 means that such conversations can start in 2021, but should really be crystalised in 2022. Waiting till 2023 is leaving it too tight, likely making hedging increasingly difficult.

There is an additional twist here - not insignificant rumps of the loans market are uncomfortable with the in-arrears computation. In many cases, systems are not set up this way, and instead require a forward-looking discount calculation, one that would come from an "in-advance" calculation (just as Libor is computed and dealt with today). Where there is no forward rate available, the preference is for a simple averaging calculation of the overnight SOFR rates looking back over the period in question, so "in arrears". This so-called waterfall approach still has forward rates as the ultimate preferred methodology that would be used as soon as they become available.

The 1mth and 3mth futures contracts should help to map out term rate to the benchmark 3mths point. Beyond that, cash volumes are really needed.

We know that the provision of Term (in-advance) SOFR rates are in the works; we await the outcome of a request for proposal process on this. The calculation itself is not the issue. Rather the robustness of the calculation is, and that is down to having suitable volumes, and as a second round, having suitably actionable quotes to help add shadow volumes to actual ones.

And it is not that these are shots in the dark, as there is already a functioning SOFR futures market, which by implication maps out implied term rates. The issue remains one of getting enough volumes in those futures, and ideally enough volumes in benchmark cash markets tenors. The 1mth and 3mth futures contracts should help to map out the term rate to 3 months. Beyond that, cash volumes are really needed.

There is an important sequence of events ahead for USA loans, and it starts with treatment for derivatives. First, the ICE Benchmark Administration (IBA) will announce cessation dates for all Ibor currencies, including USD Libor. At that point, a spread adjustment fixing date would occur, which would fix all spreads from risk-free rates to respective Ibor rates, including SOFR to 3mth Libor (now at 26bp). These spreads would also be the Alternative Reference Rates Committee (ARRC) recommendation to be applied to transform non-consumer cash loans from Ibors to fallback rates (SOFR plus the fixed spread).

No new USD Libor linked loans is the target from mid-year 2021 onwards

Note that this is an ARRC recommendation, not a directive, but a sensible one all the same that will likely be widely applied. In most cases, the new rates would be calculated as the applicable fixed spread (fixed by definition) plus the variable SOFR rate, most likely calculated as an average of the daily rates over the period in question. This would be calculated in arrears. A decent rump of players would see this as sub-optimal, preferring instead the fixed spread plus SOFR set in advance over the period in question. For this, we'd need term rates.

For new loans, the ARRC has laid out a preferred timeline. No new USD Libor linked loans is the target from 30 June 2021 onwards (including CLOs from September). Success on this front requires 2Q 2021 to be a big quarter, where the beginnings of non-Libor linked product should come to the fore. From 31 March onwards, derivative trades should be quoted in SOFR, and not in Libor. Again that points to 2Q 2021 being the key quarter for new product. The main issue here will be reluctance from players holding out for term SOFR, which risks pushing a lot of new SOFR-linked into 3Q.