Term SOFR and Credit-Sensitive Benchmarks

When LIBOR was phased out, the obvious replacement, SOFR, came with two features that made some borrowers uneasy. First, SOFR is set overnight, so a loan that resets daily and compounds in arrears only tells you your final interest cost after the period has already elapsed — awkward for a corporate treasurer who wants to know a coupon amount in advance. Second, SOFR is collateralized by Treasuries, so it barely moves when bank credit stress rises, unlike LIBOR, which was an unsecured bank-to-bank lending rate that would spike exactly when banks' own funding costs spiked. Both gaps spawned alternatives: Term SOFR, a forward-looking rate derived from SOFR futures, and various credit-sensitive benchmarks built to rise with bank funding stress the way LIBOR used to.

Two different fixes for two different complaints

Term SOFR is published for common tenors (1-month, 3-month, 6-month) and is calculated from SOFR futures prices, giving a rate known at the start of the period, just like old LIBOR was. But regulators, worried about a repeat of LIBOR's manipulation problems in a market with thin transaction volume, restricted Term SOFR's use mostly to business loans and a handful of specific cash products, discouraging its use in most derivatives, where compounded-in-arrears SOFR remains standard. Credit-sensitive benchmarks (like AMERIBOR or BSBY-style alternatives, built from actual bank commercial paper and CD issuance) address the second complaint by design: they rise when bank funding costs rise, which is exactly the behavior some regional and community banks wanted, since their own cost of funds also moves with broader bank credit conditions, not with Treasury repo rates.

Worked example

A regional bank makes a 1-year floating-rate commercial loan referencing 3-month Term SOFR + 2.50%. At origination, 3-month Term SOFR is 4.80%, so the borrower's first-quarter rate is a known 7.30%, fixed at the start of the quarter, just as it would have been under LIBOR. Compare this to a comparable loan referencing daily compounded SOFR + 2.65% (a wider spread often used to compensate lenders for giving up the advance-known feature): the borrower doesn't know their exact quarterly rate until the quarter ends and the daily compounding is finalized, even though the average level over time should be similar to Term SOFR plus its spread, absent big rate swings mid-quarter.

Worked example: credit-sensitivity kicking in

During a period of regional banking stress, a bank's own cost of issuing short-term commercial paper jumps 60bp relative to Treasury bill yields as depositors and investors grow wary of bank credit risk generally. A loan referencing a credit-sensitive benchmark built from bank CP rates sees its reference rate rise by roughly that same 60bp, automatically pushing up the interest the bank collects on that loan just as its own funding costs are rising — partially hedging the bank's net interest margin. A comparable loan referencing SOFR would see no such adjustment, since SOFR is collateralized by Treasuries and doesn't move with bank credit stress.

What this means in practice

Choice of benchmark is a real economic decision, not just an administrative label: a lender whose own funding cost tracks bank credit risk has a natural preference for credit-sensitive benchmarks on the asset side, to keep net interest margin more stable, while derivatives markets have overwhelmingly standardized on SOFR (mostly compounded in arrears) because it's the cleanest, least manipulable rate for discounting and hedging.