The benchmark is dead: long live the benchmark
The transition from one set of interest rate benchmarks to another is conceptually simple. But it is also unprecedented and has deeper consequences than many realised when Libor’s abolition was announced in 2017. With contracts worth hundreds of trillions of dollars referencing the disgraced benchmark, even small errors will have vast repercussions. PPI mis-selling? You ain’t seen nothing yet. Richard Kemmish reports
“If you keep 6s unchanged today, I will f-ing do one humongous deal with you… Like a 50,000-buck deal, whatever.”
So gabbled a Libor trader in a chat room exchange referenced in a US court case.
The writing was on the wall as soon as the discussions were made public. Resignations, fines and prison sentences were inevitable, as was the end of a benchmark seen to be so susceptible to manipulation.
Andrew Bailey, CEO of the UK’s Financial Conduct Authority, finally tolled its death knell when he announced in 2017 that banks would no longer be required to contribute reference rates to the Libor calculation after 2021.
But to abolish one set of benchmarks is easier than to propose another, and the five years given for the transition was always going to be a challenging deadline, perhaps too challenging.
“While central authorities have been considering replacement rates for Libor for some time, firms have mobilised their programmes comparatively recently,” according to Nassim Daneshzadeh, partner at PwC, who is advising clients on their transition plans.
What are we changing to?
We can’t even agree on names. Most market participants by now at least know the names of the new benchmarks for the biggest three currencies — Sofr, Sonia and €str (and if you had to google to check which benchmark is for which currency, you are part of the problem). But these are just the big three, do you know which currency will be using Saron or Tona? Answers at the end of this article.
But this is about much more than just remembering names. With the Libor family there was a commonality of methodology and an understanding of what the rates actually meant — real reference rates for term unsecured lending between banks. That commonality is no more: dollars and Swiss francs will reference secured overnight rates; sterling, euros and yen have to make do with unsecured rates. Central banks will calculate some rates, exchanges will calculate others; sterling overnight rates are compounded, Sofr rates use an arithmetic average overnight rate.
But perhaps the biggest problem is that the new rates are not even a like-for-like replacement. Libor is forward looking and available for various terms — anything from overnight to one year. It was designed as a term rate to make cashflow management easier. Backward looking rates, in contrast, are definitively only ever known at the end of the time period that they relate to. The fact that they are overnight adds another level of complexity when you want to use them to fix a bond coupon for the next three months. Converting overnight rates to term rates relies on a derivatives market and, according to Ann Battle, assistant general counsel at ISDA, “there needs to be sufficient volume in the derivatives markets to be able to generate forward long-term rates from derivatives transactions in the overnight rates. No one can predict when sufficient volume will exist”
She goes on to say: “A big focus of ISDA in the next year will be to assist the market in trading the alternatives.”
The three largest benchmarks are all developing at very different speeds. Sonia has been in existence since 1997 so, as Daneshzadeh says, “it is an existing rate and is getting increasing traction in the derivative markets and in some cash markets” but Sofr “as a new rate is developing more slowly and that may become an issue for a 2021 deadline”.
Even though €str was first published only this year, the euro market can afford to be more relaxed. Eonia, which will continue beyond 2021, is already frequently used in contracts and is an overnight rather than a term rate, easing the pressure on firms to transition.
What’s a bank to do?
Even if we can all remember the new names and agree on some market conventions, there is still a lot of wood to chop. Replies to a ‘Dear CEO’ letter from the FCA and the Bank of England published in June highlighted some of the problems individual banks face.
They included finding and quantifying a firm’s exposure to the old benchmark, making plans to address the risks involved in a transition, addressing prudential and conduct risks in the transition and engaging with their peers in the development of the replacement benchmarks.
Even something as simple as identifying and quantifying exposure to Libor isn’t trivial. How do you find all the Libor exposures in hundreds of thousands of loan contracts, not just in their main reference interest rates but as fall-back rates, or penalty interest rate references? Once the exposures are found, can they be quantified? Are alternative reference rates already contained in the legal documents? Are they hedged?
And Libor exposure isn’t always found in the obvious places such as loan contracts. Risk calculation software, collateral valuation rules, even staff performance measures sometimes have the doomed benchmark lurking in them somewhere.
It all adds up to a vast logistical exercise to replace Libor. According to Daneshzadeh at PwC, “some of the largest global banks already have 100-plus staff working on the project. By the time they move to implementation, this could increase 10-fold.”
The risks of getting it wrong
The bond market is a highly visible aspect of the transition but the biggest risks lie elsewhere.
In the US mortgage market — one never shy of litigation — mortgages are still being originated with Libor reference rates. “They shouldn’t but they are, and that could lead to big litigation problems in the future,” says Oonagh McDonald, a consultant who has called for a reprieve for Libor and who has written a book Holding Bankers to Account about the history of the scandal.
Similarly, it is difficult to rewrite water-tight contracts with companies that rely on a term structure of benchmark rates when there is not even an agreed methodology to create those term rates.
Some contracts might even have to revert to fixed rates as they do not contain proper fall-back clauses, according to Paul Richards, head of market practice and regulatory policy at the International Capital Markets Association, writing in the industry group’s quarterly report (The transition to risk-free rates in the bond market, July 2019).
“Permanent discontinuation of Libor does not appear to have been considered when these contracts were originally written, [so] there is a risk of market disruption if nothing is done to pre-empt fall-backs to a fixed rate.”
If the entire objective of Libor reform is to protect the innocent, could the very process create new risks for them? The Bank of England has already highlighted the risk of conduct issues during transition. Does a bulk transfer of retail mortgage contracts, for example, from Libor to a new rate raise questions of fairness, conflict of interest and information asymmetry?
“In the case of conduct regulation, it is important that the change of benchmark does not give rise to mis-selling or other conduct risks,” says Richards at ICMA.
Tax losses and capital gains
Then there are the accounting and taxation aspects. A material change to a debt instrument or a derivative could in some jurisdictions trigger a tax loss for the investor or, more often, given the recent downward path in interest rates, a capital gain.
Or, from an issuer’s point of view, there is a real risk that switching both the bond that they have issued and the swap that they used to hedge it could be deemed to be a material change, which would trigger a loss of hedge accounting. Given the (frequently unrecognised) importance of hedge accounting to the entire bond new issue market, the P&L swings could be enormous. Most expect the accounting standards boards to come to the rescue; if they rule that the benchmark transition is not a material change to the contract, and therefore an accounting event, tax authorities will follow. But those rulings will depend on the nature of the transition — and time is running out for them.
If the process of transitioning to the new rates revives the manipulation concerns that started the whole process, so does the most radical proposal so far: not doing it.
Replace or reform?
Despite the traders’ chat rooms and their colourful evidence of manipulation, and despite the manifest flaws in the entire Libor structure, not everyone agrees that Libor has to go, or “at the very least, banks should be pressing for an extension,” according to McDonald.
The problem of legacy contracts could be addressed by a stay of execution. “A potentially complementary option would be for the FCA... to allow the continued publication of Libor for legacy products... when the usage of Libor is restricted for new transactions,” according to Richards. But Battle says that “market participants have shown some concern about a ‘non-representative’ rate possibly being published.”
It is not as if regulators have been sitting on their hands since the scandal first erupted. In 2012 the UK government commissioned a report from Martin Wheatley into the crisis which concluded that Libor needed substantial reforms but, crucially, that it should continue to be published. The reforms were comprehensive and, in the opinion of many, perfectly adequate to restore the benchmark’s integrity and, crucially, reputation. As McDonald points out, “the press always refer to Libor as a scandal-ridden benchmark, but that view is at least six years out of date”.
The changes to the disclosure rules, with individual bank submissions now being kept confidential for three months, reduce the manipulation risk — if you don’t know who helped you manipulate the fixing, who are you going to give that 50,000-buck deal to?
The three month embargo also removes one of the other problems seen in Libor fixings. The whole scandal first came to light because of rumours that a bank was struggling to borrow in the inter-bank market at the height of the financial crisis. Willem Buiter, a former member of the Bank of England’s Monetary Policy Committee, famously quipped that Libor was “the rate at which banks don’t lend to each other”. If no one can see the individual submissions, no one knows if any of their peers are struggling to access interbank funding.
Other reforms included taking the administration of Libor away from the British Bankers’ Association and making it a regulated activity for the first time, making manipulation of the benchmark a criminal offence, shifting the emphasis to actual transactions rather than indicative prices and abolishing the less liquid currencies and maturities that were more amenable to foul play.
The reforms should also be seen in the context of much wider changes to bank compliance which would have captured some of the more egregious examples of Libor-rigging. Looking at the colourful language used in some of the traders’ chat rooms, the adjectives alone would now see the culprits summoned to the compliance department, let alone the behaviour they described.
But the reforms were insufficient. When he announced the death of Libor, Bailey said that the problem was no longer about the method but about the thing being measured.
“The market for unsecured wholesale term lending to banks is no longer sufficiently active,” he said. And if the market is illiquid “... how can even the best run benchmark measure it?”
The use of secured rates — from the repo market — is supposed to improve the quality of the data but, as McDonald points out, “repo rates are frequently very volatile, as we saw last September — this hardly accords with the principles set for the new benchmarks”.
The FCA has highlighted that not enough banks fully appreciate this and instead consider Libor replacement to be a voluntary, if belated, response to the well-documented abuses. As they said in response to a consultation earlier this year, “Some responses demonstrated limited understanding of the inherent weaknesses in Libor, instead attributing the need to transition solely to historic compliance issues... a small number of responses still presented transition from Libor as a choice rather than a necessity”.
The dollar repo spike in September may have been a one-off — or it may not. But with trillions of dollars referencing the new benchmark and with time fast running out to iron out such glitches, let alone input the resultant benchmarks into millions of contracts and thousands of IT systems, the possibility of a stay of execution for Libor beyond 2021 has to be very real.
The chat rooms where the problem started may have been almost comically crass and frequently obscene, but finding the right antidote? Far more complex than most people realised.GC
Saron: Swiss Average Rate Overnight
Tona: Tokyo Overnight Rate Average
Richard Kemmish is a guest contributor