IBOR Transition – A more in-depth look

Ganesh Melatur is a Fellow of the Chartered Institute of Management Accountants and a Fellow of the Association of Corporate Treasurers.

Sunil Kansal is a Chartered Accountant and a Fellow of the Institute of Chartered Accountants in England and Wales.

Abstract

As is well-known by now, the Financial Stability Board (FSB), the Bank of England (BOE), the Central of Bank of Japan, the Federal Reserve of the USA, the Swiss Central Bank, and the European Central Bank have been leading the initiative to move away from the London Interbank Offered Rate (‘LIBOR’) to alternative overnight risk-free rates (‘RFR’), starting from 2022. This will have an impact on financial transactions worth over 350 trillion dollars. We have decided to write a series of papers on this topic to share an understanding of the underlying issues, its impact on accounting, on the financial market, financial instruments, and market liquidity, fallback language, risk models, hedging strategies, and a whole range of allied issues. In this first paper, we provide a background on the LIBOR transition (typically called ‘IBOR Transition’) and its fundamentals.

LIBOR

The London Inter-Bank Offered Rate (LIBOR) refers to a series of daily interest rate benchmarks. Previously calculated and offered across 10 currencies and 15 borrowing periods, when administered by the British Bankers’ Association (BBA), they serve as a series of indices of the average cost to banks for unsecured borrowing for a given currency and time period. The discussion in this paper focuses primarily on the interbank market.

There are two main factors which have expedited the transition away from LIBOR: first, the financial crisis in 2008 reflected that the LIBOR is not an appropriate risk-free rate and indicated that continuation with the LIBOR may allow the credit risk bubble to build-up again. As a result, subsequent regulatory measures have been implemented to strengthen bank balance sheets that have reduced the utility of unsecured interbank borrowing in the money markets. Second, the LIBOR fell into disrepute following revelations of manipulations by traders in 2012 and is presently administered by the ICE Benchmark Administration (IBA) in a diminished configuration of 5 currencies and 7 borrowing periods. It may completely cease to exist after 2021, when the Financial Conduct Authority (FCA), pulls back from persuading or compelling panel banks to support the index.

These two factors have ensured that continuing with the LIBOR benchmark is neither feasible based on evolving market practice nor from the market trust point of view. At risk are likely to be legacy financial contracts from floating rate borrowers’ notes to swaps and investments, involving thousands of market participants. Underlying the issue of transition is the most basic and pressing contention - finding appropriate alternative benchmarks to LIBOR/IBORs. They must not only provide adequate liquidity, but also act as the basis for plenty of financial products linked to new RFRs, in the same way that LIBOR has traditionally offered to the financial markets.

LIBOR rates were originally begun in the mid-1980s by the BBA for just 3 currencies (GBP, USD and JPY) and 12 monthly maturities (like 1 month, 2m, 3m, … and 12m, for annual). 3 maturities were added later - 1 week (from 1998), 2 weeks and overnight (from 2001). Under the current IBA administration at present, 5 currencies (GBP, USD, EUR, CHF and JPY) and 7 maturities (overnight, 1w, 1m, 2m, 3m, 6m and 12m) are supported.

During the period that it was BBA-administered, LIBOR was calculated and published daily by Thomson Reuters. A panel consisting of between 6 and 18 of the largest market-participating banks in London was asked to submit a daily response to the following question for each currency and maturity:

“At what rate could you borrow funds, were you to do so by asking for and then accepting inter-bank offers in a reasonable market size just prior to 11 AM GMT?”

Each submission was the lowest rate at which a bank estimated it could fund in the London interbank market. They were not necessarily based on actual transactions. The submissions were then arranged in descending order, with the highest and lowest submissions discarded, and the remaining submissions averaged as the arithmetic mean of the two middle quartiles to create LIBOR for a given day. For some currencies, more outliers were discarded if there were a higher number of contributing banks.

LIBOR to IBORs

Even though LIBOR was calculated in London, it was based on daily submissions from a number of international banks and it was used as a benchmark globally. LIBOR also benefited from a combination of the rise of the euro markets and the convenient time-zone in which London sits. With the increasing global integration of financial markets, contracts that converged to a single, internationally recognised benchmark, LIBOR, were easy to establish and trade, leading to additional popularity and growing use in a raft of financial agreements and products.

Since 2000, many interbank benchmarks similar to LIBOR have been established in different financial markets around the world, and a few such as EURIBOR have been able to build up comparable market share. As with LIBOR, they are used as a reference rate in loans, bonds, derivatives contracts and as benchmark reference rates used in their respective currencies (e.g. EURIBOR for euro denominated contracts), and these interbank rates exhibit characteristics similar to LIBOR. Subsequent to the manipulation crisis that engulfed LIBOR, all of these rates have come under increasing scrutiny. These interbank benchmarks are now commonly referred to as ‘IBORs’ and there are fears that they could cease to exist beyond 2021, when regulators stop asking banks for submissions and the banks themselves turn reluctant to contribute, from a palpable anxiety of falling foul of tough rules on financial benchmarks and conduct.

A different question

Though related to LIBOR, each IBOR is somewhat differently constructed and the daily submission question asked by the index distributor in the context of an IBOR may be different from that used for others. In case of EURIBOR for instance, the question posed is;

“At what rate do you think interbank term deposits will be offered by one prime bank to another prime bank within the EMU zone?”

The question intends to arrive at an estimated rate for transactions between prime banks, but without delineating a prime bank’s characteristics. Further, EURIBOR allows important smaller and weaker banks to contribute rates that are not based on their own situation, and which may be lower than what they would have to pay, and this has led to observed divergences between actual cost of interbank borrowing in several countries in the EU and the EURIBOR.

LIBOR failings

A very important, observed, collective failing of the IBOR is that they are not based on actual transactions but are estimates of what could constitute a likely interbank market. This was perhaps a central determinant in the ultimate collapse of LIBOR, as banks tried to outvie each other in their perceived credit worthiness and strove to provide better estimates for their borrowing costs than their actual situation. In other cases, bank estimates of borrowing rates were provided with a view to profit from their pre-existing trading situations. Last but not the least, the interbank markets for some LIBOR rates were perceived to be nearly non-existent, as in one quoted instance there were just 15 wholesale lending transactions throughout an entire year for a particular currency-tenor, lending credence to the now common belief that daily LIBOR rates are based solely on expert judgement.

The two most significant blemishes - of falsely signalling creditworthiness and private profiteering from the provision of rate estimates - are of primary interest to regulators and help to clinch their argument that these rates are not truly reflective of the industry’s cost of funding. The regulators have cited many other failings, such as a few of the following: risk of manipulation due to conflicts of interest, lack of a standard, regularly employed procedure to corroborate individual submissions, difficulty of corroborating individual submissions, relatively small panel size (and which shrank further due to regulatory risk concerns for banks following the LIBOR-rigging scandal) in relation to the observed market as to be not fully representative, and the lack of subsequent transactions that corroborate the estimated rates. Furthermore, the management of these banks are not required to ensure that they have robust processes in place to guarantee submissions of high quality, and as LIBOR-setting is not a regulated activity, individual employees of banks involved in the process do not have to be “approved persons.” These drawbacks restrict the regulator’s ability to take disciplinary action against errant individuals or banks. The regulatory community on either side of the Atlantic, and also elsewhere, are increasingly adamant that the LIBOR must and will be replaced by new benchmarks that are tethered to actual transactions.

LIBOR and CAPM

Employing the technical parlance of financial markets, the LIBOR may be viewed as being constructed on lines of the capital asset pricing model - with a key, basic rate and added adjustments for counterparty (CP) and tenor (liquidity) risks. Continuing with the CAPM, the key rate could be based on or reflective of the systemic, risk-free rate in a particular market. When viewed in this way, even though their application in inter-bank transactions remains unaffected, the current, endemic use of LIBOR in loan and mortgage contracts however, becomes questionable, particularly when components for term and credit risks are once again added in individual investment and borrowing contracts, revealing double counting practice, since LIBOR itself is a construction containing adjustments for CP and liquidity risks. Presently, this awareness has led to market expectations that on substitution of the LIBOR by new benchmark rates as replacement, finance costs for all contractual arrangements could trend lower. In fact, the replacement rates recommended by the Financial Stability Board (FSB) are the so-called risk-free rates (RFR), that are basically overnight rates without any supplements for CP or liquidity risks.

RFR and OIS

Overnight rates, generally fixed by a country’s central bank for its reserve currency, find considerable employment in Overnight Index Swaps (OIS), that are agreements to swap a fixed rate against a daily overnight reference rate. The OIS could be a fixed or floating interest rate swap with the floating leg tied to a published index of a daily overnight rate reference. The period could range from one week to two years. The two parties agree to exchange at maturity, on an agreed notional amount, the difference between interest accrued at the agreed fixed rate and interest accrued through averaging the floating index rate. This means that the floating rate calculation replicates the accrual on the amount rolled over (Principal plus Interest) at the index rate every business day over the term of the swap. If cash can be borrowed by the swap receiver on the same maturity as the swap and at the same rate and lent back every day in the market at the index rate, the cash pay-off at maturity will exactly match the swap pay out: the OIS acts as a perfect hedge for the funded instrument. Since indices are generally constructed on the basis of the average of actual transactions, the index is generally achievable by borrowers and lenders alike. Economically, receiving the fixed rate in an OIS is like lending cash. Paying the fixed rate in an OIS is akin to borrowing cash.

The interbank markets allow banks to borrow wholesale deposits from other banks over a shorter term at a relatively low cost compared to bond or equity funding. However, the biggest issue with the OIS rate being its short-term nature, borrowers will have to ensure long term LIBOR-linked funding by paying a premium over the OIS rate. Thus, borrowers opting for OIS-linked funding would have to rollover their funding on a daily basis at the OIS rate, with the premium paid by them, or the spread between the LIBOR and OIS being reflective of the borrower’s credit and liquidity risks. Significantly enough, the LIBOR-OIS spread became a very suggestive indicator of credit risk during the 2008 financial crisis.

LIBOR and OIS

The LIBOR–OIS spread has historically been around 10-15 basis points (bps), except during the financial crisis, when it reached an all-time high of 364 bps following the collapse of Lehman Brothers in 2008, that indicated a severe credit crisis. Higher LIBOR/OIS spreads are thus a meaningful indicator of worsening credit conditions. Alan Greenspan, the former Federal Reserve Chairman has stated that, “The LIBOR-OIS spread remains a barometer of fears of bank insolvency,” and that he wouldn’t consider credit markets back to “normal” until the LIBOR-OIS spread was at 25bps.

The chart on the next page captures the LIBOR-OIS spread during the financial crisis in 2008.

(Source: Bloomberg and Financial Times)

Subsequent to the financial crisis, several measures undertaken by the major central banks around the world to withdraw excess liquidity from the financial markets. The banks have reduced unsecured lending between banks significantly, thereby lowering the LIBOR-OIS spread to around 1 – 20 bps now.

RFR/OIS vs IBORs

RFR as overnight rates are based on independently sourced transaction data and are anchored in actual transactions. Hence, there will never be a risk of transaction deficiency. Similarly, overnight rates have been commonly used in cash instruments and OIS have a long history of trading going back 20 years. This needs to be contrasted with the scarcity of underlying transactions in the term interbank and wholesale unsecured funding markets from which IBORs are constructed. However, as against the forward-looking bias of the IBORs, the most important drawback of RFR is their backward-looking nature. This means that the interest to be paid and received on contracts with an RFR basis, will not be known in advance to the parties to the transaction who will hence be unable to assess their exposures and will remain at a disadvantage in planning likely cashflows and suitable hedges for their contracts. An additional handicap is that RFR, due to their inherent overnight nature, do not include term components for 1m, 3m or a year, etc. which deter their application in corporate floating rate notes, loans, mortgages and securitisations. This hurdle needs to be crossed if the proponents of RFR are keen for them to be used as the next generation of benchmark rates.

The inherent advantages accruing to IBOR are not lost to their many advocates, however. In fact, it appears plausible that IBOR may continue well after 2021 and IBA is trying its best to ensure the achievement of such an outcome. It has planned a root-and-branch reform to eliminate many of the identified vulnerabilities of the market-favoured benchmark. For starters, rates submitted to IBA by the panel banks will have to be based on actual transactions, necessarily trade-weighted, as against the simple rates used in the past. IBA will accept bank submissions only on the basis of an evolved, well-defined standard and regularly employed procedure. And it proposes to substitute the submission question that was previously quoted in this paper, by means of a 3-level test akin to the fair value hierarchy first introduced through IFRS 7 and familiar to accounting professionals.

First level submissions are those offered by panel banks that have actually borrowed in the wholesale markets from at least two different sources over the past 24 hours, and they must submit a weighted average of the transaction. Second-level submissions are from panel banks that have borrowed 2 to 10 days prior, and their estimate should be the weighted average of those transactions after adjusting it to reflect changes in the RFR. Finally, banks making a level 3 submission, may do so in accordance with an IBA-approved framework, requiring considerable expert judgement but still based on market data.

The process objective is to come up with a daily IBA LIBOR rate that is to the greatest possible extent anchored to actual, wholesale funding transactions. To provide more focussed and reliable estimates, the IBA intends to produce LIBOR rates only for those currency-tenor pairs that are most critical to its clients, while staying away from thinly traded forecasts. Not surprisingly, its efforts seem to be somewhat well supported by panel banks that have confirmed their submission to a reformed benchmark, well past the 2021 deadline date.

LIBOR Transition

The RFR camp on the other hand, counts on the well-publicised intention of the regulatory authorities of many countries to transition the market away from LIBOR, primarily due to its stigmatised record. Andrew Bailey, previous CEO of the UK FCA and currently Governor, BoE, speaking at a conference in July 2019, emphasised “misplaced confidence,” in the survival of LIBOR and stipulated that, “firms should treat it as something that will happen,” as predicted by the end of 2021. US Federal Reserve authorities and other US banking institutions regulators have been similarly insistent that the discontinuation of LIBOR is a certainty.

Regulatory concerns over the index, even as it has moved away from financial misconduct and manipulation for private ends, still remain linked to its lack of transactional basis. That may seem somewhat paradoxical for RFR, which are supported by regulators primarily because they are solidly transaction-based, but their pervasive market presence is yet to evolve into tangible application in financial contracts and products employing them as the primary basis, chiefly because markets continue to cling to LIBOR-linked products. This market preference does not arise from a lack of attempts. In year 2018, the first GBP 1 billion five-year bond was launched by the European Investment Bank (EIB), referencing SONIA (a GBP LIBOR replacement RFR). Though first planned for a GBP 500 million size, it was more than doubled following strong investor-backing, to the order of GBP 1.55 billion. It was priced at SONIA + 35 bps, though secondary trading lowered the spread to sub-30 bps as was expected that rates based on replacement benchmarks may well lower finance costs.

This first attempt was followed in 2018 by a World Bank sale (through IBRD) of GBP 1.25 billion bonds also of five years, with a SONIA spread of just 24 bps. Private-sector banks such as the Lloyds Banking Group (3-year debt GBP 750 million with SONIA spread of 43 bps), and Santander have also been successful in their early attempts. Credit Suisse and World Bank have priced deals using SOFR, a US RFR that however is a secured rate as against SONIA and LIBOR that are unsecured rates. National Express became the first UK company to take a loan priced off SONIA, in July 2019.

However, this has not resulted in an exodus away from the stigma-filled IBOR, even as market participants disregard regulatory advice and continue to ink deals referencing LIBOR and EURIBOR (is the advance obituary notice for EURIBOR’s death in 2020 “greatly exaggerated”, much like Mark Twain’s famous cable from London to the US press in 1897?) The FCA warned in July 2018 about the risks of inertia and the transition not being fast enough, but the RFR could be pointless unless the private sector starts using them. Other private sector observers have opined that the RFR market would evolve only when borrowers request products based on the new benchmark, or that banks were not yet in a position to offer RFR-based products since they were waiting achieve comparable liquidity to LIBOR-based products. But to reach that desired liquidity, RFR-based products need to be available in the first place. Not surprisingly, this has led to views that a two-tier market could perhaps evolve offering both LIBOR and RFR-based products.

Efforts to improve IBOR continue through global reforms from the International Organisations of Securities Commissions (IOSCO) and the G-20 mandated FSB, by means of guidelines on best practice, to make them more robust and trustworthy. Thus, the IBOR compilation process now is much more reliable, robust and well-documented and follows IOSCO guidelines closely. Though that isn’t enough for most regulators, a few may be prepared to tolerate a two-tier market in their jurisdictions, consenting to follow market choice due to many reasons, including that of legacy contracts.

LIBOR to RFR Transition table

So far, the following central banks have announced their plans to move to an alternate RFR.

(Source: Progressed estimate from an initial table contained in The Wheatley review of LIBOR, August 2012)

Two-tier market with IBOR & RFR

The chart below provides a rough breakdown for the total financial values of contracts that depend on LIBOR as a benchmark.

As mentioned previously, floating rate corporate notes, bonds, securitisations and mortgages rely on LIBOR with term and liquidity premia added on. However, these do not constitute more than 12% of the market, based on the table above, and together with forward rate agreements that are another 15% of the market, it appears as if the mainstay for LIBOR in the financial contracts market may be slightly beyond 25%. For interest rate swaps and derivatives, the backward-looking aspect RFR does not pose a problem as these trades overnight and could largely switch to RFR, except for hedging instruments that reference a cash (funded) instrument to which they relate. Other kinds of derivatives fall outside the scope of this discussion.

Alternate scenarios

If it were decided to move forward exclusively with RFR, assuming forward-looking RFR with term and liquidity components were somehow incorporated into contracts, the biggest outstanding problem would be about dealing with legacy contracts, since for most instrument types, LIBOR would have to be switched to RFR on a contract-by-contract basis. Fall-back provisions that are included in loans, bonds and derivatives to ensure a rate is available when LIBOR is not available on any given day due to any reason, are not helpful particularly when the benchmark is about to be discontinued. A new model wording for contracts to provide a migration mechanism would have to be developed on a case-by-case basis. System changes for computation of interest on loans and bonds that reference backward-looking rates such as an RFR, would also be necessary.

Though it appears that derivatives may be easier to change than cash instruments, via so-called protocols for such amendments, commercial negotiations still remain to be undertaken for changing financial agreements in a manner acceptable to both parties to a derivative transaction. Furthermore, as RFR are likely to be lower than IBOR, additional compensations may have to be decided to maintain the transacting parties’ original bargain. But, as most parties to such agreements may lack detailed financial and legal knowledge to negotiate the necessary changes, additional costs including for communication and training may be warranted.

Hence, the biggest question to be answered now is: will it be possible by the end of 2021 to switch from IBORs to the new RFRs? Such a scenario would mean the complete discontinuation of IBORs beyond 2021 and an adequate availability of new financial instruments linked to RFRs together with sufficient market liquidity. Due to the introduction of regulatory measures like Basel III, to better capitalise banks, the IBOR discontinuation would not anyhow be a pressing concern, as interbank borrowing via money markets would be down to a trickle. An alternate scenario is also possible where there may not be an adequate number of new RFR linked products and liquidity in the market, forcing market participants to look for alternate RFRs (such as BOE base rate) due to which regulators may decide to extend the deadline to discontinue IBORs.

In our next paper, we will discuss the impact of IBORs on the accounting of financial instruments and market risk hedging.

References

-  Reports from The Banker

-  Credit Suisse (2001): Material on Overnight Index swaps, www.acisuisse.ch/docs/dokumente/OIS_Note_CSFB_Zurich.pdf

-  Various publications from The Association of Corporate Treasurers (ACT), U.K.

-  A discussion provided by HSBC on the IBOR reforms

-  A discussion provided by RBS on the construct of LIBOR

-  Bank of Japan (2006): Recent developments of OIS market in Japan; Review, www.boj.or.jp/en/type/ronbun/rev/data/rev06e04.pdf

-  BIS (2003): The Euro Interest rate swaps market; Quarterly review, www.bis.org/publ/qtrpdf/r_qt0303f.pdf

-  British Bankers’ Association (2004): Sterling overnight Index Average, www.bba.org.uk, Benchmarks

-  Reserve Bank of Australia (2002), Australian Financial Markets, Bulletin,