March 31, 2022 - We are now a full quarter into the SOFR regime and, aside from a few frictions, loan originations have adapted remarkably well. Below we discuss what we’ve learned thus far in the post-LIBOR world – and what’s on deck for the coming year.
On March 31st, LSTA’s Meredith Coffey joined a session at the ABA Business Law Section Spring Meeting discussing the phase-out of LIBOR on loans. While the shift to SOFR has been relatively smooth, new loans must address two major differences between LIBOR and SOFR. First, LIBOR is a forward looking “known in advance” rate. This means that borrowers and lenders lock in their rate in advance of the interest period – and this feature has defined loan documentation and operations for decades. Second, LIBOR is a credit sensitive rate, whereas SOFR is a risk-free rate – and this means that the SOFR loans require some adaptation to maintain economic neutrality with LIBOR loans.
As the LSTA deck demonstrates on slides 3-4, the ARRC and the market considered roughly five types of replacement rates. First, there are the daily rates – Daily Simple SOFR and Daily Compounded SOFR. Because these rates are calculated daily during the interest period, the borrower and lenders do not know their rate upfront, there is more operational complexity and documentation must change materially. Ultimately, there was little appetite for the daily rates. Next, there are the “known in advance” rates – SOFR Compounded in Advance, Forward Looking Term SOFR and some credit sensitive rates (“CSRs”) such as BSBY, CRITR and some Ameribors. These rates are operationally and documentationally simpler. However, banks rebuffed the backward-looking “SOFR Compounded in Advance” and regulators largely rejected broad use of the CSRs. This left Forward Looking Term SOFR, which the ARRC recommended for fallbacks and tolerated for new issue business loans and CLOs. Thus, the US loan market has rapidly adopted Term SOFR for new loans (slide 5). (In contrast, regulators in other jurisdictions, such as the UK, have pushed back more aggressively on Term Replacement Rates, limiting their adoption.)
The next US challenge, though, was the economics. As slide 6 notes, SOFR is a risk-free rate, so the Term SOFR curve is lower and flatter than LIBOR. To make SOFR more economically similar to LIBOR, one can use Credit Spread Adjustment (“CSA”) – either in the form of a flat adjustment (like 10 bps) across the curve or a “CSA Curve” (like 10 bps for 1M, 15 bps for 3M and 25 bps for 6M) – or increase the SOFR loan margin to make the all-in rate more similar to a LIBOR loan.
So, what has the market done? This topic was discussed by LevFinInsights this week and will be covered further on April 1st by the LSTA’s Tess Virmani and Ted Basta at the ABA conference. The CSA norms have been shifting rapidly. In January, 69% of institutional loans used a CSA. However, as market conditions increasingly favored borrowers in February, the CSA share declined to 50%. (When a CSA is not observed as a separate field, theoretically the spread adjustment may be included in the margin. However, most loans still are pricing in increments seen in LIBOR, such as 300 bps/350 bps/400 bps. If a CSA were embedded in the margin, theoretically more loans should be pricing in the 312.5 bps/362.5bps/412.5 bps context.) Market conditions shifted away from borrowers in March – and the share of CSAs climbed again to 67%. That said, LFI noted that it was a different “kind” of CSA: Most January deals used a CSA curve (10/15/25 bps), whereas most March deals used a flat 10 bps CSA.
While SOFR originations have gone reasonably smoothly, there still are trillions of dollars of loans that must get off LIBOR by June 30, 2023. As flagged on slide 7, there are several ways to do this. First, parties can simply refinance into a new SOFR loan. Alternatively, parties can use fallback language, either at LIBOR cessation or earlier if their loan has an “early opt-in” feature in its fallback. And here’s where the market comes into play again. Because interest rates were so low, until late February LIBOR and SOFR were similar and shifting to SOFR (with any spread adjustment) would be costlier to borrowers than simply remaining in LIBOR. However, with market volatility and March interest rate hikes, LIBOR has increased much more rapidly than SOFR. Thus, LIBOR is now higher than either SOFR+market CSA (10/15/25 bps) or SOFR+ARRC Hardwired Fallback CSA (11/26/42 bps). In turn, borrowers could benefit by choosing to opt in early to SOFR loans. While lenders might miss some of the economics in the next 15 months, early opt-ins significantly reduce the execution risk of transitioning trillions of dollars of loans in June 2023. (To facilitate remediation and reduce execution risk, the ever-helpful LSTA is developing SOFR Amendment Forms.)
