SOFR transition was meant to eliminate basis risk—the mismatch between what you hedge and what you actually owe. Instead, it created an entirely new category: the unresolved residual basis that sits on treasury books across US, UK, and European banks. Six years into the migration, treasury teams still lack a consensus methodology for measuring, reserving against, and ultimately closing out these microscopic but persistent pricing gaps.
Why SOFR Transition Created Basis Risk Rather Than Eliminated It
The transition from LIBOR to SOFR was pitched as a de-risking exercise. Regulators—the Federal Reserve, the FCA, the ECB—pushed it hard because LIBOR was administered, manipulated-prone, and based on sparse actual lending transactions. SOFR, by contrast, is transaction-based: it reflects actual overnight repo rates in the market.
The problem: SOFR and LIBOR were never economically identical. LIBOR priced in credit risk (a bank’s cost of wholesale funding). SOFR does not—it’s a secured overnight rate. That spread—the LIBOR-OIS (overnight index swap) basis, or the SOFR credit spread adjustment (CSA)—was always there. But regulators and dealers compressed it into a single historical spread adjustment at the moment of transition (June 2023 in US dollar markets, late 2021 in sterling). They called it the “adjustment spread” and treated it as static. It isn’t.
A UK-based treasury manager at a £200bn mid-sized bank describes the result: “We hedged our floating-rate debt issuance using SOFR swaps. The adjustment spread we locked in at transition worked fine for six months. But the actual market basis between SOFR and what our hedging counterparty priced kept moving. We’re now fifty basis points away from our models on some tenor buckets. That’s real P&L slippage, and there’s nowhere in our system to record it because the hedge formally ‘matched’ the underlying on day one.”
This is the core unresolved problem: basis risk didn’t vanish. It migrated from LIBOR discontinuation risk (the known unknown) to residual basis volatility (an unknown unknown that regulators still haven’t formalized how to measure).
The Three Unresolved Basis Risk Vectors
1. Tenor Basis Mismatch
When dealers converted LIBOR derivatives to SOFR, they used term SOFR (SOFR for 3 months, 6 months, 12 months) as the economic equivalent. But term SOFR is less liquid than the overnight SOFR on which it’s built. It’s a constructed rate—a forward-looking quote based on overnight futures. Overnight SOFR, by contrast, is fully hedgeable in the repo market.
The result: a term SOFR basis. A treasury team hedging a 6-month floating-rate coupon with a SOFR swap based on 6-month term SOFR faces basis risk if the market’s implicit 6-month term SOFR (derived from the overnight curve) diverges from the quoted 6-month term rate. At present, this basis is typically 2–5 basis points, but it moves. It’s uncompensated basis risk on positions that are meant to be fully hedged.
A Singapore-based treasury officer at a regional bank observed: “We have USD 2 billion in 6-month SOFR floaters outstanding. Our swap hedge is based on 6-month term SOFR at the time of issuance. But the dealer is now quoting the fair value of 6-month term SOFR differently depending on the horizon and the convexity adjustment methodology. We have no agreed protocol for which quote is ‘right’ for marking our hedge to market. So the basis goes unmonitored.”
2. Cross-Currency Basis
US dollar SOFR and sterling SONIA (Sterling Overnight Index Average) are both transaction-based, both post-transition. But they were transitioned on different schedules and with different liquidity profiles. SONIA is older, more embedded in sterling markets. SOFR is younger, and its liquidity still concentrates in short tenors.
When a UK bank borrows in USD and swaps the proceeds back to GBP, it’s exposed to the cross-currency SOFR-SONIA basis—the spread between what it costs to borrow USD and convert to GBP versus the direct cost of GBP funding. That basis is driven partly by the relative creditworthiness of USD and sterling funding, partly by supply and demand for the swap, and partly by the relative maturities of the underlying rate structures.
The unresolved issue: there’s no standard market convention for how to measure, report, or reserve against this basis in consolidated treasury books. Some banks use the FX forward points as a proxy; others use swap curve spreads; others don’t measure it at all. The result is inconsistent hedge effectiveness reporting across institutions and no regulators explicitly requiring a specific methodology.
3. Counterparty Credit Spread Disconnection
Here’s where the system becomes genuinely opaque. When you convert a LIBOR hedge to SOFR, you’re implicitly moving from a rate that priced in average bank credit risk to a rate that prices in zero credit risk. The adjustment spread is supposed to bridge that gap. But that spread is fixed. Your counterparty’s actual credit spread—the cost of that bank’s unsecured funding—is not fixed.
If you hedge a floating-rate debt issuance with a SOFR swap from Counterparty A, you’re implicitly bearing the credit risk that Counterparty A’s funding spread widens relative to SOFR. If your debt investor was previously pricing in an average banking sector LIBOR spread, and Counterparty A is now perceived as riskier than average, you have uncompensated basis risk. The hedge doesn’t perform as expected because the creditworthiness discount embedded in the original LIBOR rate isn’t replicated in the SOFR swap.
This risk is almost entirely unmeasured. Ask a treasury risk manager to explain how they monitor for credit spread basis risk on hedging counterparties, and you’ll get evasion. It’s not in most CVaR (Credit Value-at-Risk) models. It’s not explicitly required in Basel III hedge accounting documentation. But it’s real, and it compounds when you run a multi-dealer hedge book.
What Treasury Teams Are Actually Doing (Badly)
Most treasury teams have fallen into one of three camps:
**The Ignorer.** “We hedged everything at transition. The books match. We move on.” This is technically compliant with hedge accounting rules but leaves uncompensated basis risk on the books. When auditors ask, these teams typically respond: “The hedge is effective to within 80–125% of the underlying,” which is the regulatory test but not the economic reality. Five basis points of unresolved basis times USD 5 billion notional is USD 2.5 million of unreserved slippage. Over a portfolio, this becomes material.
**The Over-Hedger.** Some teams, recognizing the basis risk, have added “buffer” hedges—extra notional above the underlying to over-compensate for expected basis drift. This is economically sensible (you’re paying for insurance against basis widening) but creates a mismatch problem with accounting: now the hedge formally fails the 80–125% effectiveness test, and you must move the variance to P&L volatility, which conflicts with earnings guidance and treasury compensation metrics.
**The Fragmenter.** Larger banks with geographically dispersed treasury operations have left basis risk management to regional teams. A London team manages sterling-USD cross-currency basis. A New York team manages USD tenor basis. A Singapore team manages AUD SOFR exposure. There’s no consolidated view of residual basis across the firm. This created, in at least one case, a situation where two regional teams were inadvertently hedging each other’s basis, creating notional hedges of over 150% of actual exposure—discovered only during an audit.
The right approach—systematic measurement of all three basis vectors, explicit revaluation at quarter-end using market-consensus methodologies, and clear documentation of the economic decision to accept or hedge residual basis—remains the exception rather than the norm.
Why Regulators Haven’t Stepped In
The Federal Reserve, FCA, and ECB have all issued post-SOFR transition guidance. But all of it focuses on the mechanics of the transition (ensuring all contracts are amended, that fallback language is correct, that systems are updated) rather than the economics of residual basis risk management.
The regulatory gap exists because:
1. **Legacy LIBOR discontinuation was the stated problem.** Once LIBOR ceased (USD LIBOR ended on June 30, 2023; GBP LIBOR ended on December 31, 2021), regulators considered the mandate achieved. Residual basis was treated as a normal market risk, not a transition risk.
2. **Basis risk is distributed.** It doesn’t concentrate in one institution or asset class. The 2–5 basis point tenor basis on USD 50 trillion in derivative notional is a system-wide drag, but it’s not a solvency risk for any one bank. Regulators prioritize concentration risk over diffuse but persistent drag.
3. **The accounting standards (ASC 815-20 for US banks, IFRS 9 for others) don’t require explicit measurement of basis effectiveness beyond the 80–125% test.** As long as the hedge doesn’t fail that test, variance is acceptable. Regulators inherit the accounting framework rather than impose a more stringent view.
The result: treasury teams operate in a framework where residual basis risk is technically permissible but economically unmonitored.
The Practical Impact on Liquidity and Capital Planning
For a treasury manager running a USD 10 billion floating-rate debt program, unresolved basis risk translates into three concrete problems:
**Earnings Mismatches.** If your hedge basis drifts 3 basis points per annum on a 5-year position, that’s 15 basis points of cumulative cost not captured in your funding cost forecast. Across a large debt program, this becomes a consistent miss against treasury’s budgeted cost of funds—leading to either lower-than-expected return on equity or unbudgeted treasures of hedging ineffectiveness.
**Refinancing Risk.** When you refinance a matured SOFR hedge, the market’s consensus adjustment spread may have shifted. If you’re hedging at a wide spread relative to current market, you’ve crystallized a loss. But if your original hedge was tight relative to current market, you’ve gained—and that gain gets buried in unexamined basis tracking.
**Cross-Border Complexity.** Why G-SIBs Are Using AI to Solve the 120 Trillion Dollar Cross-Border Liquidity Trap highlights the acute challenge of managing global liquidity. When you add unresolved basis risk across USD, GBP, and EUR SOFR variants, you’re running a liquidity position that you can’t fully measure. A surprise widening of cross-currency SOFR-SONIA basis could force you to post unbudgeted collateral or unwind positions at losses.
Emerging Industry Responses
Some large banks have begun to address this. How Global Transaction Banks are Using AI to Solve the 100 Trillion Dollar Liquidity Puzzle describes how leading institutions are applying machine learning to uncover hidden patterns in liquidity flows. A few treasury operations have adapted this to SOFR basis tracking—building daily recalibration models that recalculate the fair value of term SOFR across tenors and identify divergences from market consensus pricing.
The FCA and ECB have also hinted at future guidance. The December 2024 ECB monetary policy statement referenced ongoing monitoring of “remaining SOFR-related basis dynamics,” suggesting that regulatory interest is growing. But no explicit rule change has been announced.
The Algoy Perspective
The SOFR transition is being treated as a solved problem. It isn’t. Regulators and dealers closed the LIBOR discontinuation hole, but they inadvertently opened a new one: a system-wide, distributed, but economically real basis risk that sits on treasury books with no agreed-upon methodology for measurement or reservation.
What separates leading treasury operations from the laggards is not technology—it’s rigor. The best-performing teams have built daily basis monitoring dashboards that explicitly track tenor basis, cross-currency basis, and counterparty credit spread disconnection. They revalue hedges not just at month-end but continuously, and they document the economic decision to accept or hedge residual basis. This requires investment in data infrastructure and the willingness to challenge the assumption that “if the hedge passes the 80–125% test, it’s effective.”
For a treasury team at a mid-sized bank or a regional operation, this is especially critical. You lack the scale to absorb persistent basis drift as a cost of doing business. You also lack the technology investment budgets of global systemically important banks (G-SIBs). The solution isn’t to adopt an elaborate AI-driven model; it’s to establish a repeatable quarterly process: identify all material hedging relationships, calculate the residual basis using a documented market-consensus methodology, and reserve or hedge that basis explicitly.
What Should Happen Next
Three regulatory and industry actions would resolve this:
1. **Explicit SOFR basis measurement standard.** The International Swaps and Derivatives Association (ISDA) or a central bank consortium should publish a standardized methodology for measuring tenor basis, cross-currency basis, and counterparty credit spread basis. Until that exists, each institution invents its own, and comparability across financial statements breaks down.
2. **Hedge accounting update.** Accounting standard-setters should clarify that basis risk on SOFR-hedged positions must be separately disclosed—either as a reserve, a separate P&L line, or explicit narrative. The current 80–125% test is too blunt an instrument to capture economic reality.
3. **Regulatory guidance on cross-currency basis.** Given the maturity of sterling and dollar SOFR markets, the FCA and Federal Reserve should jointly issue guidance on how to measure and manage SOFR-SONIA basis, particularly for multinational treasuries.
Until those happen, treasury teams should assume that their residual SOFR basis risk is unmonitored and uncompensated. That’s not a regulatory failing; it’s an opportunity. The teams that build systematic tracking now will have a clearer picture of their true cost of hedging—and a competitive advantage when rates finally stabilize and basis risk becomes tradable.
How Should Your Treasury Team Address SOFR Basis Risk?
Establish a quarterly process: identify all hedged floating-rate positions, calculate residual tenor basis using Bloomberg fair value pricing, cross-currency basis using FX swap spreads, and counterparty credit spread disconnection using historical credit spread indices. Document the methodology. Decide whether to hedge or absorb each basis vector. Reserve explicitly if you absorb. Review with your CFO and external auditors quarterly to ensure alignment on the economic approach.
Frequently Asked Questions
Is SOFR basis risk the same as LIBOR basis risk?
No. LIBOR basis risk was discontinuation risk—the possibility that LIBOR would cease and you’d be forced into a disadvantageous fallback. SOFR basis risk is volatility in the spread between SOFR and other rates (term SOFR, SONIA, counterparty credit spreads). It’s residual, persistent, and largely unmonitored by regulators or auditors.
Do I need to reserve for SOFR basis risk under IFRS 9 or ASC 815-20?
Not explicitly. As long as your hedge passes the 80–125% effectiveness test, accounting standards don’t require a separate reserve. But good risk management practice—and increasing auditor scrutiny—suggest that you should quantify it and document the economic decision to accept or hedge it.
Which SOFR basis component is most material?
For most treasuries managing USD funding, tenor basis (the spread between different maturity term SOFR rates) is the largest unmonitored component. Cross-currency basis matters primarily for multinational firms. Counterparty credit spread disconnection is smallest in absolute terms but most opaque.
