Why LIBOR went away: the benchmark that outlived its market

The benchmark outlived the market behind it

LIBOR disappeared because a benchmark embedded in financial contracts had become difficult to support with the transactions it was supposed to measure. Manipulation damaged its credibility, but removing dishonest submissions would not have restored a deep market for unsecured bank borrowing at multiple maturities. The eventual solution combined stronger benchmark design with a coordinated migration of contracts. That distinction explains why the transition took years rather than ending with the first enforcement actions. [15]

The endpoint is no longer a future deadline. The last three synthetic US dollar settings were published on September 30, 2024. The FCA and Bank of England confirmed the following day that all 35 LIBOR settings had permanently ceased. Synthetic rates had been a temporary bridge for legacy contracts, not a revival of the original bank-submission system. This article reviews that completed transition using the official record available on October 4, 2026. [1]

What LIBOR represented, and why it spread

LIBOR, the London Interbank Offered Rate, was a family of reference rates rather than one universal interest rate. Currency and maturity mattered: three-month dollar LIBOR was a different setting from one-month sterling LIBOR. A floating-rate contract could specify a particular setting plus a negotiated margin. The common reference made pricing and hedging easier across otherwise different loans, securities and derivatives. [10]

The benchmark was intended to represent unsecured wholesale bank funding. Its term rates combined expectations about interest rates over the borrowing period with compensation associated with lending to banks without collateral. A business loan indexed to LIBOR consequently linked its payments to a bank-funding benchmark even when the borrower had no dealings in the London interbank market. This was an economical convention, but it also transmitted the benchmark’s weaknesses into contracts far beyond the market used to construct it. [5, 6]

Analytically, a reference rate is a shared measuring device. Its usefulness comes partly from widespread adoption: lenders, borrowers and hedgers can agree on a starting point without negotiating an entirely new measure of market interest rates. That same coordination advantage makes replacement difficult. Changing the measuring device in one contract may leave a linked hedge, accounting system or security using the old one.

Misconduct and market erosion were different problems

The 2012 Barclays action illustrates the integrity problem. On June 27, the UK Financial Services Authority imposed a £59.5 million penalty for benchmark misconduct. Its findings included submissions influenced by derivatives traders’ positions and reduced LIBOR submissions during the financial crisis because of concerns about adverse media coverage. Those are different incentives: one seeks a trading advantage; the other seeks to influence perceptions of a bank’s financial condition. The regulator also identified deficient controls and failures to respond to escalated concerns. [3]

The structural problem persisted after governance improvements. In his July 27, 2017 speech, FCA chief executive Andrew Bailey explained that the underlying unsecured wholesale term-lending market was no longer sufficiently active. He cited one currency–maturity combination in which about a dozen contributing banks had collectively undertaken only 15 potentially qualifying transactions during all of 2016. A daily benchmark could therefore depend heavily on judgment despite the appearance of a precise published number. [2]

The implication is more fundamental than a need for better policing. Strong controls can govern how an estimate is produced; they cannot create transactions that market participants no longer routinely undertake. A trustworthy submission process and a representative underlying market are separate requirements. LIBOR’s failure involved both.

Reform became a change in financial infrastructure

The Financial Stability Board’s July 22, 2014 reform program pursued stronger existing benchmarks and development of alternative near-risk-free rates. That was a broader international project than simply renaming LIBOR. Different currencies and markets needed reference rates appropriate to their underlying transactions and financial products. [4]

In the United States, the Federal Reserve Board and New York Fed convened the Alternative Reference Rates Committee, or ARRC. Its selection of SOFR in 2017 and the beginning of New York Fed publication in April 2018 provided the basis for the dollar transition. SOFR draws on transactions in the Treasury repurchase-agreement market, where cash borrowing is secured by Treasury collateral. [16]

Building a replacement required an ecosystem as well as a calculation. The ARRC’s retrospective describes work on contract fallbacks, legislation, market conventions and operational preparation. A loan system needs reset dates, day-count treatment and payment calculations; a derivatives market needs usable hedging instruments; contracts need a successor that remains legally workable. The transition’s length reflected those dependencies and the difficulty of coordinating many institutions, rather than uncertainty about whether the old benchmark had weaknesses. [15]

SOFR changed the economics as well as the label

The New York Fed describes SOFR as a broad measure of overnight Treasury financing costs. Its published SOFR averages compound realized overnight rates over 30, 90 and 180 calendar days, while the SOFR Index permits calculation over custom periods. Those backward-looking averages are distinct from a rate predicting borrowing costs over the coming three months. [6]

Term SOFR provides a forward-looking term structure derived from SOFR derivatives markets. The ARRC endorsed its use within a defined scope, including business lending and legacy cash-product fallbacks, while preferring overnight SOFR or averages for many uses. Its recommendations restricted broad derivatives use, with exceptions including hedges of Term SOFR cash products. The distinctions matter because “SOFR” alone does not specify exactly how a contract calculates interest. [7]

Near-risk-free does not mean unchanging or immune to market stress. Treasury repo conditions can move, and interest-rate policy affects financing costs. The crucial economic difference is that secured overnight SOFR does not embed the same unsecured term bank-credit component as LIBOR. Replacing one with the other therefore changes more than the source of the daily number. [5, 6, 7]

The global transition did not make SOFR the replacement for every currency. Sterling moved toward SONIA, whose reformed methodology covers overnight unsecured sterling transactions. The Bank of England began the reformed calculation for transactions dated April 23, 2018. Its unsecured design also demonstrates why the phrase “risk-free rate” does not universally mean a secured repo benchmark: maturity, market depth and methodology all matter. [8]

Why the exit had several dates

The end of new business and the end of legacy publication served different purposes. US banking agencies’ November 30, 2020 statement encouraged banks to stop entering new dollar-LIBOR contracts as soon as practicable and by December 31, 2021. Allowing existing contracts additional time did not mean encouraging a fresh stock of LIBOR exposure. [9]

At the end of 2021, 24 of the 35 LIBOR settings ceased. Five dollar settings continued on a panel-bank basis until mid-2023, while selected sterling and yen rates continued temporarily on a synthetic basis. This staged approach allowed more contracts to mature or be amended without an abrupt, simultaneous interruption across every market. [10]

The final dollar panel publication was June 30, 2023. Overnight and 12-month dollar LIBOR then ceased permanently. The one-, three- and six-month settings continued synthetically, using the relevant CME Term SOFR rate plus a fixed ISDA spread adjustment, until their final publication on September 30, 2024. A screen could still display a LIBOR-labelled number during that interval even though its economic construction had already changed. [11, 1]

Scroll horizontally to see all columns.

The contract problem: replacing a rate without replacing the bargain

A fallback provision specifies what happens when the named benchmark cannot be used. Some legacy clauses were designed for a short interruption, not permanent cessation. Others did not supply a practical successor. A temporary last-known-rate mechanism can behave very differently when the rate never returns. The Federal Reserve described this group as tough legacy contracts when implementing the federal LIBOR Act. [12]

The December 16, 2022 final rule identified SOFR-based replacements for specified contracts subject to that Act after June 30, 2023. It also addressed statutory safe harbors and continuity. It was not a blanket declaration that every contract mentioning LIBOR had identical terms or would convert in the same way. Applicability, existing fallback language, contractual choices and product category remained important. [12]

The statute specifies tenor adjustments, including 0.11448 percentage points for one-month LIBOR, 0.26161 for three months and 0.42826 for six months. These equal 11.448, 26.161 and 42.826 basis points, respectively. The statutory consumer-loan treatment included a one-year transition in the adjustment. A benchmark spread is separate from the borrower-specific lending margin. [13]

The Federal Reserve’s May 2021 Financial Stability Report explains that the ISDA adjustments used a historical five-year median difference between each LIBOR setting and its associated fallback rate. The March 5, 2021 announcements fixed those spreads for relevant fallbacks. A fixed historical adjustment reduces a particular discontinuity at conversion; it does not recreate a time-varying bank-credit premium forever. Two benchmarks with different underlying economics will not move together perfectly in every future environment. [14]

The lasting significance

An illustrative contract makes the distinction clearer. A legacy loan might replace three-month LIBOR plus a two-percentage-point borrower margin with the specified SOFR-based successor, the applicable transition spread and that same borrower margin. The successor rate’s observation and compounding conventions determine the resulting cash flows. This is a conceptual example, not a claim that all converted loans used that formula or that newly negotiated SOFR loans must carry a statutory legacy spread.

The transition removed reliance on a discontinued benchmark, but it did not abolish basis risk, funding risk or disputes about contract interpretation. Those are different economic and legal questions. Its more durable achievement was to connect widely used reference rates more closely to active markets while developing a mechanism for changing the contractual infrastructure around them.

LIBOR’s history is therefore a story about the limits of financial standardization. A convenient convention can remain dominant after the market behind it has changed. Replacing it requires both a better measure and a workable path for the obligations already written against the old one. The scandal accelerated attention; the thin underlying market made the long-term case for departure. [2, 3, 4]

Sources

1. FCA, Bank of England and sterling working group: The end of LIBOR, October 1, 2024SourceBack to text: ↑1↑2↑3
2. FCA: Andrew Bailey, The future of LIBOR, July 27, 2017SourceBack to text: ↑1↑2
3. FSA: Barclays benchmark misconduct enforcement, June 27, 2012SourceBack to text: ↑1↑2
4. FSB: Reforming Major Interest Rate Benchmarks, July 22, 2014SourceBack to text: ↑1↑2
5. New York Fed: Five things about LIBOR, ARRC and SOFR, November 26, 2018Official sourceBack to text: ↑1↑2
6. New York Fed: Reference rates and SOFR averages methodologyOfficial sourceBack to text: ↑1↑2↑3↑4
7. ARRC: Summary and update of Term SOFR scope-of-use recommendations, April 21, 2023Official source · PDFBack to text: ↑1↑2
8. Bank of England: SONIA reform implementation announcement, October 16, 2017SourceBack to text: ↑
9. Federal Reserve, FDIC and OCC: LIBOR transition statement, November 30, 2020Official releaseBack to text: ↑
10. FCA: Final messages on LIBOR before end-2021, December 10, 2021SourceBack to text: ↑1↑2↑3
11. FCA: The US dollar LIBOR panel has now ceased, July 3, 2023SourceBack to text: ↑1↑2
12. Federal Reserve: LIBOR Act implementing final rule, December 16, 2022Official releaseBack to text: ↑1↑2
13. Federal Reserve: Adjustable Interest Rate (LIBOR) Act, statutory textOfficial sourceBack to text: ↑
14. Federal Reserve: May 2021 Financial Stability Report, funding risk and LIBOR transitionOfficial sourceBack to text: ↑
15. ARRC: Closing Report, final reflections on the transition, November 2023Official sourceBack to text: ↑1↑2
16. ARRC: The transition to SOFR, timeline and milestonesOfficial sourceBack to text: ↑