The Secured Overnight Financing Rate, or SOFR, was introduced last year by the Alternative Rate Reference Committee (ARRC) in an attempt to move away from the scandal-plagued LIBOR. It measures the cost of secured overnight borrowing by calculating an average of overnight repo transactions.
SOFR has clear benefits over LIBOR: it is based on real market transactions (not estimated) that are far greater in number (transactions used to calculate 3-month LIBOR hover around $500mm, while daily SOFR volumes range between $700-800B). However, SOFR’s daily calculation methodology leaves financiers scratching their heads when determining how to apply SOFR to a longer period, such as a 3-month bond.
For loans with a tenor greater than one day, the Fed recommends taking an average of SOFR rates across the loan period. This approach is fine for market participants that don’t need to know the cost of capital until the end of the loan – but this is not exactly orthodox. Another alternative is looking backward and pricing the debt based on the historical daily average of the market (the borrower pays what the market used to charge), but this approach may be less accurate over the long run. Participants are hopeful that a healthy, forward-looking derivatives market will soon flourish and provide sensible longer-term rate references and, indeed, the ARRC expects to release term reference rates based on that market by the end of 2021. Importantly, that goal relies heavily on robust volumes in the SOFR market.
Thus far, bond issuance volume has come from government sponsored enterprises – Fannie Mae, Freddie Mac and Federal Home Loan Bank are the largest participants – followed by banks and insurance firms. Corporate SOFR linked issuance is limited, although more corporate entities are including fallback language in their debt offerings, which states how interest will be calculated via SOFR if LIBOR ceases to function during the duration of the bond. Corporations are not likely to flood the SOFR-linked issuance market unless reliable measures of term rates exist.
SOFR has attracted respectable issuance to date, but we expect greater corporate issuance will be required to get the market comfortable with the pivot away from LIBOR. The introduction of a SOFR-linked Treasury security is one reasonable way to help facilitate the transition and the U.S. Treasury has publicly mulled the introduction of such an instrument. The bell doesn’t toll for LIBOR until 30 months from now, however we think the owners and issuers of the ~$200T of outstanding LIBOR-referenced debt will need more than just encouraging statements to get comfortable: they will need a market.
The Alternative Reference Rates Committee, A User’s Guide to SOFR, April 2019
CME Group, Adoption of SOFR and SONIA as Floating Rate Note Benchmarks, May 8, 2019
Federal Reserve Bank of New York, Secured Overnight Financing Rate Data, NY Fed Website
J.P. Morgan, LIBOR Primer: Setting the Stage for SOFR, January 15, 2019