The transition from the London Interbank Offered Rate (LIBOR) to the Secured Overnight Financing Rate (SOFR) successfully established a robust, transaction-based, nearly risk-free foundation for U.S. dollar markets. But it left an important market-structure question unresolved: how should corporate credit instruments reflect changes in systematic credit conditions?
As discussed in a previous blog for BWC, this question is central to the next stage of benchmark reform.
Most floating-rate corporate loans answer that question implicitly. They are priced at SOFR plus a fixed spread. That spread compensates lenders for borrower-specific credit risk, but it also reflects the systematic credit conditions prevailing when the loan is originated. Once the spread is fixed, however, those components remain bundled together even as market-wide credit conditions change.
Amazon provides a useful example. In June 2026, the company established a $17.5 billion senior unsecured delayed-draw term loan facility. Borrowings under the facility carry a Term SOFR margin of 62.5 to 87.5 basis points, depending on Amazon’s credit rating. Under the conventional structure, that margin is fixed when the loan is priced, even if broader credit conditions change over the life of the facility.
A more transparent structure would separate them: SOFR + systematic credit spread + borrower-specific spread.
Transaction-based credit spread measures such as the USD Across-the-Curve Credit Spread Index (AXI) and USD Financial Conditions Credit Spread Index (FXI) provide one way to implement such a structure.
Separating Two Different Types of Credit Risk
SOFR was deliberately designed to be nearly risk-free. That is a strength, not a flaw. But SOFR was never intended to measure credit conditions.
Research by Antje Berndt, Darrell Duffie and Yichao Zhu developed AXI as a measure of the recent cost of wholesale unsecured debt funding for publicly listed U.S. bank holding companies and commercial banks. The index uses credit spreads on unsecured debt instruments across maturities, weighted by transaction and issuance volumes. The researchers also developed FXI by broadening the underlying universe to corporate debt issuers.
The distinction is important. SOFR measures the nearly risk-free rate; AXI or FXI can measure systematic credit conditions; and a remaining fixed spread can reflect borrower-specific credit risk.
This framework does not recreate LIBOR. Instead, it separates risks that are combined under conventional SOFR-plus-spread pricing.
What Would This Look Like in Practice?
In March 2026, Amazon issued floating-rate notes with coupons of compounded SOFR plus 44 and 59 basis points. In June, the company established a $17.5 billion senior unsecured delayed-draw term loan facility under which Term SOFR borrowings carry a margin ranging from 62.5 to 87.5 basis points, depending on Amazon’s credit ratings.
Suppose, purely for illustration, that AXI were 50 basis points when an Amazon loan priced at SOFR plus 87.5 basis points. The same initial coupon could instead be expressed as:
SOFR + AXI + 37.5 basis points.
Nothing changes economically on day one. The all-in rate is identical. The difference emerges afterward. The borrower-specific component remains fixed, while the systematic credit component moves with market conditions.
That distinction becomes clearer when looking at another large investment-grade issuer. Oracle issued a $500 million floating-rate note in January 2025 at compounded SOFR plus 76 basis points. A year later, in February 2026, it issued another $500 million floating-rate note at SOFR plus 111 basis points, a 35-basis-point increase.
Oracle’s expected ratings on the two transactions remained Baa2/BBB/BBB, although Moody’s and S&P had changed their outlooks from stable to negative by the second transaction. The widening therefore cannot simply be equated with a ratings downgrade. It potentially reflects some combination of changing Oracle-specific risk, market technical factors, and broader credit conditions.
A structure that separately identifies a systematic credit component would make that decomposition more explicit.
Why Dynamic Credit Pricing Could Matter in Stress
The distinction becomes particularly important during periods of financial stress.
In a recession or financial shock, the Federal Reserve may ease monetary policy, causing risk-free short-term rates to decline. At the same time, credit spreads can widen as risk appetite deteriorates, and funding conditions tighten.
Under a conventional SOFR-plus-fixed-spread loan, the borrower’s coupon may therefore decline even as the economic cost of providing credit increases. SOFR reflects monetary policy easing, but the fixed margin does not respond to contemporaneous changes in systematic credit conditions.
This issue is particularly relevant for committed bank credit facilities. Corporate borrowers may draw revolving credit lines during periods of market stress, precisely when liquidity becomes more valuable and bank funding conditions may deteriorate. Research underlying AXI specifically identifies this potential mismatch: SOFR can decline during stress even as bank funding spreads increase, potentially encouraging greater use of SOFR-linked commitments at a time when providing those commitments becomes more costly.
A credit-sensitive supplement changes that dynamic. As SOFR declines, a widening AXI could offset part of the decline, allowing the loan’s coupon to respond both to monetary policy and to changing bank funding conditions.
There is, however, an important counterargument: credit-sensitive pricing can be procyclical. If AXI rises during a recession, borrowers could face higher credit spreads precisely when revenues and liquidity may already be under pressure.
That trade-off should not be minimized. But fixed spreads do not eliminate systematic credit risk; they allocate it differently. Lenders providing committed financing must anticipate the possibility that their own funding conditions will deteriorate during stress and incorporate that risk into pricing and balance-sheet decisions.
The relevant question is therefore not whether systematic credit risk exists, but how it should be measured, priced, and shared between borrowers and lenders.
From Concept to Implementation
Incorporating a systematic credit component would require changes to loan documentation, operational systems, and hedging practices. Market participants would also need confidence in index governance and sufficient underlying transaction volumes. Broader adoption could eventually support derivatives and other hedging instruments, although liquidity would need to develop alongside the cash market.
These are meaningful considerations, but they are considerably more manageable than the wholesale benchmark transition required by LIBOR’s cessation. The objective is not to prescribe a single contractual structure, but to demonstrate that systematic credit risk can be identified and priced separately from borrower-specific risk.
FXI could extend the same principle beyond bank funding conditions. Because its underlying universe includes both financial and nonfinancial corporate debt, FXI provides a broader measure of corporate credit conditions. Its potential applications need not be identical to those of AXI: AXI has a particularly direct economic connection to bank lending, while FXI may be more relevant where broader corporate credit conditions are the desired reference.
The Next Step in Benchmark Reform
The LIBOR transition answered one question decisively: the foundation of U.S. dollar floating-rate markets should be a robust, transaction-based, nearly risk-free benchmark. SOFR provides that foundation.
The next question is how markets should handle the credit risk that SOFR intentionally leaves out.
Today, much of that risk is embedded in a fixed spread established when a loan or bond is originated. Separating that spread into a market-wide credit component and a borrower-specific component could make the economics of corporate borrowing more transparent and allow pricing to adjust as financial conditions change.
AXI and FXI offer a transaction-based approach. They need not replace SOFR or recreate LIBOR. The more important principle is that systematic credit risk and borrower-specific credit risk are different risks, and benchmark architecture should be capable of pricing them separately.
With SOFR now firmly established, that distinction deserves greater attention from borrowers, lenders, investors and policymakers.
Featured Author:
Alex Roever is Senior Advisor to SOFR Academy, an American financial market infrastructure company. He previously spent more than 25 years at J.P. Morgan Securities, where he served for nearly a decade as Managing Director and Head of U.S. Interest Rate Strategy. Mr. Roever represented J.P. Morgan on the Financial Stability Board’s Market Participants Group, which was charged with recommending enhancements to major interest rate benchmarks, including IBORs, following the Global Financial Crisis. He is a CFA Charterholder and has also served as a Senior Director at CFA Institute.
All views expressed by members are their own and not reflective of the views of the Bretton Woods Committee.

