VoxEU Column Financial Markets Financial Regulation and Banking

Eclipsing LIBOR

Financial firms have paid fines totalling more than $9 billion for manipulating LIBOR, yet this flawed benchmark has not been replaced. This column argues that there are reduced incentives for banks to participate in setting the LIBOR rate, and so the potential of, and incentives for, manipulation remain. Although LIBOR is unsustainable, international regulators are working to produce more robust alternatives and to smooth the transition.

[T]he volume of unsecured wholesale lending has declined markedly… This development makes LIBOR more susceptible to manipulation, and poses the risk that it may not be possible to publish the benchmark on an ongoing basis if transactions decline further.”
Financial Stability Oversight Council (2016).

The manipulation of the London Interbank Offered Rate (LIBOR) began more than a decade ago, when employees of leading global firms began to submit false reports to the British Banking Association (BBA). At first they did this to influence the value of LIBOR-linked derivatives, but later, during the financial crisis, they did it to conceal the deterioration of their employers’ creditworthiness. US and European regulators reported many of the details in 2012 when they penalised Barclays, the first of a dozen financial firms that collectively paid fines of more than $9 billion (Commodities Futures Trading Commission 2015, New York Times 2016). As well as having to settle claims from aggrieved clients, these firms face enduring reputational damage. In some cases, management was forced out, and in others, individuals received jail terms.

You might think that after this costly scandal, and knowing challenges of maintaining LIBOR, market participants and regulators would have quickly replaced LIBOR with a sustainable short-term interest rate benchmark that had little risk of manipulation. You’d be wrong. The current administrator (ICE Benchmark Administration), which replaced the BBA in 2014, estimates that this guide (now called ICE LIBOR) continues to serve as the reference interest rate for “an estimated $350 trillion of outstanding contracts in maturities ranging from overnight to more than 30 years" (our emphasis).

In short, LIBOR is still the world’s leading benchmark for short-term interest rates.

Against this background, Andrew Bailey, the CEO of the UK Financial Conduct Authority, recently called for a transition away from LIBOR before 2022 (Bailey 2017). In this column we briefly explain LIBOR’s role, why it remains an undesirable and unsustainable interest rate benchmark, but why it will be difficult to replace (even gradually over several years) without disruption.

Why we need benchmarks, and why LIBOR doesn’t work

As articulated by Duffie and Stein (2015), there are three rationales for financial benchmarks, ranging from the S&P 500 to the Brent oil price.

They lower transaction costs by providing an agreed basis for settling trades.

  • By focusing trades on a particular instrument, benchmarks foster competition and improve market depth. While competition squeezes per-trade margins, the transparency provided by a benchmark can boost market activity sufficiently to compensate the traders receiving smaller margins. This explains why private coordination efforts have led to benchmarks in the past. BBA LIBOR, which arose in the 1980s, is just such a case (Hou and Skeie 2014).
  • Benchmark-linked assets provide a mechanism for hedging common risks, increasing the overall risk-bearing capacity of the financial system.

Unfortunately, unlike the S&P 500 index or the Brent oil standard, LIBOR is not based on transactions, making it vulnerable to manipulation. In its pre-2014 incarnation, the BBA calculated LIBOR by surveying a panel of large London banks every business day at 11am to estimate their cost of uncollateralised borrowing in a specific currency, at a specific maturity. In the case of dollar LIBOR, for example, the 2005 panel included 16 banks, and the BBA reported a trimmed mean measure that excluded the top two and bottom two submissions (Eisl et al. 2015). Overall, the BBA reported 150 LIBOR benchmarks, covering 10 currencies at 15 maturities, ranging from overnight to 12 months.

Potential and incentives for manipulation persist

Wherever submissions rely on expert judgement, rather than transactions, there is the potential for manipulation. ICE has implemented reforms since 2014 designed to reduce the risk of this manipulation (ICE Benchmark Administration 2017a) – but there are not many transactions at some currency-maturity pairs, and so it is impossible for participating banks to eliminate expert judgement when they make their daily submissions. As of the second quarter of 2017, ICE reported that slightly more than one-third of submissions for the US dollar three-month LIBOR – the most important tenor for derivatives markets – were based on transactions (ICE Benchmark Administration 2017b).

The persistent decline in uncollateralised short-term borrowing, and the associated reduction in the number of transactions, is only making matters worse, as the US Financial Stability Oversight Council stated above. As an example, compare unsecured and secured (collateralised) short-term funding in the Eurozone. Figures 1 and 2 plot the trajectory of these two types of funding, using the latest available data from the ECB’s money market survey (European Central Bank 2015). Since 2003, secured borrowing has gradually replaced unsecured lending. As Figure 1 shows,

unsecured funding declined by 85%, while secured funding grew by nearly 90% (Figure 2). The two charts highlight the relative scarcity of transactions longer than one month in maturity.

Figure 1 Euro money market: Cumulative quarterly turnover in unsecured cash borrowing, 2003-2015

Source: Adapted from Chart 16, ECB (2015).

Figure 2 Euro money market: Cumulative quarterly turnover in secured cash borrowing, 2003-2015

Source: Adapted from Chart 30, ECB (2015).

Even a purely transaction-based benchmark will fail when there are few transactions, because traders could manipulate it with a few small, well-timed transactions.

Duffie and Stein (2015) show that there remains an incentive to manipulate. They report that daily transactions at the key three-month maturity in LIBOR-based US dollar derivatives markets are roughly 1,000 times larger than in the cash market for unsecured bank funding, and the stock of derivatives affected by price movements is about 100,000 times larger, creating the possibility of enormous derivatives-market gains or losses in response to tiny changes in LIBOR. The temptation for manipulation may be overwhelming, even in the face of strong compliance oversight.

LIBOR is unsustainable

Faced with an environment where there are few unsecured transactions, fewer banks are willing to participate in the survey panels. From their perspective, participating in the LIBOR survey with its increased compliance obligations (or in surveys for Euribor, HIBOR, SIBOR and other IBORs) is costly, with large legal and reputational risks. Those risks only grow as the submissions rely increasingly on expert judgment.

LIBOR is on life support. To keep it alive, the UK authorities have had to cajole leading banks to remain ‘voluntary’ participants (Enrich 2013). In his July speech, Bailey, the CEO of the FCA, explained that his agency’s leadership has “spent a lot of time persuading panel banks to continue submitting” to avoid “the market disruption that would be caused by an unexpected and unplanned disappearance of LIBOR.” Bailey also warned that – under the new European Benchmarks Regulation that takes effect on 1 January 2018 – the FCA’s authority to compel banks to make LIBOR submissions will expire before the currently targeted four to five-year transition away from LIBOR is complete.

Coordinating the transition from LIBOR

A range of international regulators is working hard to make this transition as smooth and rapid as possible. In the US, in 2014 the Federal Reserve convened the Alternative Reference Rates Committee (ARRC), a group of market participants who have been assessing transactions-based alternatives to LIBOR. In June 2017, they chose as the best-practice short-term interest rate benchmark a new ‘broad Treasuries repo financing rate’. This will be published by the Federal Reserve Bank of New York and the Treasury Office of Financial Research (Alternative Reference Rates Committee 2017, Bayeux et al. 2017). On 24 August, the Federal Reserve Board sought public comment on the proposal to produce the new Secured Overnight Financing Rate (SOFR), and two other narrower measures of secured financing costs.

SOFR has four key advantages over LIBOR:

  • It is based on a deep, liquid market, with average daily transactions of several hundred billion dollars, making it difficult to manipulate.
  • It is calculated as a volume-weighted median (rather than a trimmed mean), making it less prone to bias (Eisl et al. 2015).
  • As a secured financing rate, it is more consistent with post-crisis short-term financing practice.
  • As a near-default-free rate, it provides investors who currently use LIBOR-linked derivatives to hedge default-free interest rate risk with a better mechanism for doing so.

That said, SOFR lacks two elements that have been integral to LIBOR. It does not report a range of maturities, or the funding risk premium of banks. With regard to the first, given that Treasury bills, notes and bonds allow investors to manage maturity risk, this seems like a minor issue. The loss of LIBOR-linked derivatives would, however, pose a much bigger challenge for creditors wishing to manage fluctuations in risks associated with bank borrowing – what is normally referred to as ‘funding risk’.

Put differently, it would be useful to have two benchmarks: one like SOFR to hedge changes in risk-free rates, and another tailored to hedging changes in the level of bank funding risk. The Financial Stability Board’s Market Participants Group has offered a proposal for an exclusively transactions-based LIBOR+, which seems well-suited to the second purpose. To ensure an adequate volume of transactions, LIBOR+ would be based on all available wholesale-unsecured bank-funding rates (including bank commercial paper and certificates of deposit) at the relevant maturity (see Financial Stability Board 2014, especially Section 2.2.1, which identifies alternative reference rates and discusses a transition).

The challenges ahead

How quickly would the legacy LIBOR-linked contracts disappear if everyone were immediately to substitute SOFR for LIBOR in all new debt and derivatives contracts? In some markets, such as exchange-traded derivatives, it would probably take less than five years (see FSB 2014, Figure 2, p. 244). Many over-the-counter derivatives would likely linger for longer. And, some LIBOR-based debt instruments – like the 30-year floating-rate mortgages included in some mortgage-backed securities – could be with us for decades.

Moreover, unless the authorities step in, not all market participants would shift instantly to using SOFR-linked financial instruments, even if it would be in their collective interest to do so. To smooth the adjustment, regulators should at least require that all new LIBOR-linked contracts include a fall-back clause to substitute a sustainable reference rate if a reliable version of LIBOR was no longer available.

If LIBOR were to disappear, courts would have to resolve the contract frustration over legacy LIBOR instruments. These legal risks would be limited if there was a comparable, sustainable replacement rate. At short maturities the MPG is confident that LIBOR+ captures enough transactions to achieve a seamless transition. But we don’t know how much legal risk would linger, or what would happen if the volume of uncollateralised funding continues to shrivel. Presumably, contract holders facing losses from a benchmark substitution would challenge its validity in court, with the potential for market disruption.

The good news is that the international regulators and leading market participants – including the largest banks that dominate the derivatives markets – are aware of the risks arising from the need to replace LIBOR. They are working together to create a path to develop, test and agree practical substitutes. While there is no way to guarantee a disruption-free transition, the partial eclipse of LIBOR already is well underway. It is only a matter of time until we reach totality.

Authors’ note: An earlier version of this column appeared at www.moneyandbanking.com.


Alternative Reference Rates Committee (2017), “The ARRC Selects a Broad Repo Rate as its Preferred Alternative Reference Rate,” press release, 22 June.

Bailey, A (2017), “The Future of LIBOR”, speech at Bloomberg London, 27 July.

Bayeux, K, A Cambron, M Cipriani, A Copeland, S Sherman, and B Solimine (2017), “Introducing the Revised Broad Treasuries Financing Rate,” Liberty Street Economics, 19 June.

Commodity Futures Trading Commission (2015), Order Instituting Proceedings Pursuant to Sections 6(C) and 6(D) of the Commodity Exchange Act, as Amended, Making Findings and Imposing Remedial Sanctions, 20 May.

Duffie, D and J C Stein (2015), “Reforming LIBOR and Other Financial Market Benchmarks,” Journal of Economic Perspectives 29(2): 191-212.

Eisl, A, R Jankowitsch and N G Subrahmanyam (2017), “The Manipulation Potential of Libor and Euribor", European Financial Management.

Enrich, D (2013), “Banks Warned Not to Leave Libor,Wall Street Journal, 13 February.

ECB (2015), Euro money market survey. September.

Financial Stability Board (2014), “Final Report of the Market Participants Group on Reforming Interest Rate Benchmarks,” 2 July.

Financial Stability Oversight Council (2016), 2016 Annual Report. United States Department of the Treasury.

Hou, D and D R Skeie (2014), “LIBOR: origins, economics, crisis, scandal and reform,” Federal Reserve Bank of New York, Staff Report No 667.

ICE Benchmark Administration (2017a), “Summary of ICE LIBOR Evolution,” 24 January.

ICE Benchmark Administration (2017b), “ICE LIBOR Quarterly Volume Report,” Q2.

New York Times (2015), “Tracking the Libor Scandal,” Dealbook, 23 March, 23 March 2016.

1,470 Reads