A Creator of the U.K.'s Libor Replacement Says It's Still Too Complex

(Bloomberg Markets) -- As banks prepare to move away from the scandal-tainted London interbank offered rate, few in the city of London will be watching more closely than William Porter. Now head of European credit strategy at Credit Suisse Group AG, Porter was working on J.P. Morgan’s short-term interest rate and strategy team in the mid-1990s as it searched for a way to hedge short-term floating-rate risks against European currencies.

Their solution was the sterling overnight index average, or Sonia. Earlier this year, the benchmark, which is considered impossible to manipulate because it’s based on actual transactions, was designated Libor’s heir apparent by the Bank of England. Here, Porter describes how Sonia came to life and why it still isn’t the perfect solution to measuring overnight lending rates.

A Creator of the U.K.'s Libor Replacement Says It's Still Too Complex

EMMA HASLETT: Set the scene: It’s 1996, you’re working at J.P. Morgan. How does Sonia come about?

WILLIAM PORTER: It was a desire to replicate overnight swaps that existed in France. There were two swaps traded against the average overnight rate that the Bank of France observed, known as TAM. Effectively, French market participants were hedging short-term, floating-rate risks and funding risks using those swaps.

We were trading a lot of foreign exchange forwards, which have two short-term interest-rate risks, at the time. Everything, by convention, goes back to the dollar. So you’ve got the underlying currency, and you’ve got the dollar, and we wanted a way to strip that out. We were looking at ways to do it using futures, but nothing really worked because of the funding element, and we had the brain wave of starting to trade on [the] fed funds effective [rate] the way they were used to trading in France on this TAM rate. We just started doing that, so that started the dollar overnight index swap market in London.

We filled in gaps in the U.K. and in Japan by appealing to the brokers with a new product to trade and by getting them to publish the overnight average that they observed. So the Wholesale Market Brokers’ Association in the U.K. [now known as the European Venues & Intermediaries Association] started publishing its observed average overnight rate, which was Sonia—and the Bank of England kept a watchful eye. Fast-forward 20 years, and enhanced Sonia is going to be calculated by the bank.

EH: Was it hard to persuade other market participants to get involved?

WP: They were initially skeptical, but there was always some liquidity. As you know from the Libor fiasco, these books are huge, some of them. So it didn’t take much market share to make it worth the brokers’ while.

One of the tricks at the time was how volatile Sonia was. There was no benchmarking. The overnight sterling market was a dark art. That was one area of difficulty. It was difficult to persuade authorities that this was in the public interest. They eventually came round, but there was so much volatility in Sonia. I don’t think they wanted to publicize it.

EH: You later joined the European Banking Federation’s Euribor steering committee as the European Union prepared to introduce the euro. What was your role?

WP: The euro was hoving into view, and Europe was seen as a major threat to the London markets. I was inserted into the Euribor working group that led into the steering committee. As Euribor ramped up, the London International Financial Futures and Options Exchange observed liquidity was going to be greater, so they benchmarked their euro interest rate future to Euribor, not Libor, and traded it in London anyway. That was a huge decision, very public, and seen as a disloyal threat in some circles.

Part of the reason I was put into that committee was because there was no provision for overnight rates. So there we were trading all these currencies that were going to be succeeded by the euro overnight, and there was no provision for an overnight successor rate, so that caused a bit of a panic. Eonia [the euro equivalent of Sonia] was how we put that right—we basically added Eonia to the Euribor structure and just basically mapped across the participants.

EH: Is its implementation being handled well by the Bank of England?

WP: It’s a difficult challenge. I’ve had concerns about the nature of fixings for a long time, and I think Andrew Bailey [chief executive officer of the Financial Conduct Authority, the U.K. financial markets watchdog] is right to point to systemic issues that he wants to get rid of. Maybe there’s some detail in the implementation, but overall I think they’re on absolutely the right track.

EH: Could there be a cliff-edge scenario on the date the Bank of England switches from Libor to Sonia, leaving investors who hold bonds that were pegged to Libor but mature after the transition in an awkward position?

WP: I don’t think at the end of 2021 the world’s going to end, but it’s a bit like Brexit: I don’t think there’s going to be a cliff on 29 March 2019, but it would be imprudent in the extreme not to hedge against it. There’s a date starting to emerge in the future when the way that we’re doing things is going to change. So the most sensible thing to do is to start anticipating that.

EH: Twenty years on, do you have any regrets?

WP: Its complexity is a regret. We regretted this early on. The daily compounding is, in my view, a mistake. I don’t think it’s going to be changed now. It’s analytically pure. We said at the time, Should we do it the other way? And with hindsight we should have just averaged it and then left people to deal with any complexity worries.

The logical thing to do is just to set up a site where you can put in two dates and get the compounding between them, and then click on the number and socialize the calculation rather than everyone doing it in-house, unverified. To have an external, verifiable source would be a good idea.

So that is one regret. We should have done just simple averaging.

Haslett covers European credit markets at Bloomberg News in London.

©2018 Bloomberg L.P.