February 28, 2019 - In the last few years at ABS Vegas, the buzziest topic has been CLOs – their regulatory encounters, their return profile, the value in their capital stack and their perennial arbitrage challenges. This year, at the successor SFIG Vegas, LIBOR took over as most prominent subject. Indeed four of the conference’s panels were devoted in total or in part to LIBOR cessation. Presenters told attendees what they must do to avoid market disruption; as a public service, we offer their streamlined suggestions and steps below.

Step One: Identify your exposure. Several panelists noted that both issuers and investors need to know their total LIBOR exposure. These exposures can show up in assets, liabilities, hedges and more. A full accounting is the first step to managing your risk.

Step Two: Know your potential replacement rates. For US dollar LIBOR, most market participants believe SOFR, a risk-free overnight rate constructed from Treasury repos, is the most likely replacement. But there are different SOFRs, with different attributes and drawbacks.

• Forward Looking Term SOFR. Most loan market participants want a forward looking term SOFR, which would have a term curve like LIBOR. This would allow market participants to know their interest rates in advance and, because it is so similar to LIBOR, would be relatively easy to operationalize. However, because actual SOFR is an overnight rate, a forward looking term SOFR must be derived from SOFR futures trading. As a non-trivial drawback, there is no absolute guarantee that this rate can be developed.
• SOFR Rate Compounded in Advance. This rate would be determined by compounding the equivalent number of days for the previous time period. For instance, if a borrower wanted a 30-day SOFR contract, it would use the rate calculated by compounding overnight SOFR for the previous 30 days. A hypothetical day-count example may be useful. If a borrower wanted a 30-day compounded in advance SOFR contract beginning April 1^st, it could use the overnight SOFR beginning March 2^nd and compound it every day until March 31^st. The compounded rate on March 31^st would be locked in for the 30-day contract starting on April 1^st. The positive is that this rate is known in advance and would be operationalized similarly to term SOFR. The negative is that, for longer contract periods, it could be stale. Importantly, compounded SOFR is publicly available. Clarus publishes compounded SOFR on a daily basis; in the p. 2 COW in a recent LSTA Newsletter, we showed a time series compounded SOFR rate. In addition, as Risk reported, the FOMC also suggested (on p. 10 of its recent minutes) it could publish the rate starting in 2020.
• SOFR Compounded in Arrears. This rate is determined by compounding SOFR during the interest period. Let’s continue with the hypothetical example above, with a borrower wanting a 30-day compounded in arrears SOFR beginning April 1^st. In this case, overnight SOFR would be compounded every day from April 2^nd to April 30^th. The positives are that this is the true, exact interest rate on the loan and it is neatly hedgeable with swaps. The main negatives are that the final compounded rate is not precisely known at the beginning of the interest period, and many borrower and lender systems cannot easily operationalize a daily compounding rate yet.
• Simple Daily SOFR in Arrears. This rate is pulled daily, just like compounded in arrears, but it is not compounded.  Using the example above, the overnight SOFR rate would be applied to the loan every day from April 1^st to April 30^th. The positives are that this rate is basically the exact rate of interest, it is close to being perfectly hedgeable and loan systems already can do this with daily LIBOR and Prime. The negative is that the exact rate will not be known until the end of the interest period and it creates far more data flow through systems.

Step Three: Create workable LIBOR fallbacks (which answer the question, “If LIBOR went away tomorrow, to what rate (and how) would my loan fall back?”). The ARRC issued consultations on FRNs and syndicated loans last September. After receiving a number of responses, the body is preparing to release recommended language soon. Ideally, recommended fallback language for bilateral loans and securitizations, which had their own consultations in December, will follow soon thereafter.

Step Four: Begin to think about operationalizing replacement rates in loan systems. The ARRC and the LSTA have been working with market participants to begin the operationalization discussion. In addition, the LSTA is joining a Finastra Webcast next week to discuss this specific issue. (LSTA members can also reach out directly to mcoffey@lsta.org or ehefferan@lsta.org.)

Step Five: Start issuing SOFR loans. Sounds crazy, right? After all, is SOFR really ready for prime time? Still, the best way to reduce the “LIBOR cessation” problem is to stop making LIBOR loans. Moreover, there have been a number of new SOFR securities issued. According to SFIG presenters, there have been nearly $60 billion of SOFR issuances since mid-last year. To be fair, there are some caveats. First, much of the issuance has been by the public sector. (Fannie Mae alone has issued roughly $13 billion in SOFR notes.) Second, much of the issuance has been on the shorter end of the curve (see last week’s COW for a tenor distribution). Third, most notes have been purchased by money market investors, who are quite familiar with using an overnight benchmark. Fourth, even with money market buyers, SOFR issuances have generally used Daily Simple SOFR because many buyers’ systems cannot yet operationalize daily compounding.  But “can’t today” and “can’t ever” are not the same thing. According to a recent FCA speech, so far in 2019, there have been 15 issuances totaling nearly £16 billion that have referenced compounded SONIA, the Sterling replacement rate. One might posit that, if we can put a man on the moon, we can put a loan on SOFR.