Basis risk looms for insurers in Libor transition

UK insurers may need to pay more and run basis risk to hedge interest rates after the transition

  • The Financial Conduct Authority announced in July that banks will no longer be compelled to publish the interbank lending rate Libor beyond 2021, with Sonia the preferred successor rate.
  • That leaves insurers potentially having to rewrite legacy swap contracts, which will be costly, time-consuming and an operational headache, says EY’s Gareth Mee.
  • Insurers will be hedging one rate and exposed to another if European authorities do not change the methodology of Libor-based discounting curves used in Solvency II.
  • Such basis risk incurs capital charges and may be unavoidable whatever the transition period, as switching hedges will be gradual.
  • Andrew Smith, a professor at University College Dublin, thinks insurers may still be forced to discount at a Libor-linked rate, if the Bank of England does not seek a change in discounting rates.
  • Potential illiquidity in the developing Sonia swaps market could load insurers with rate risk and frustrate a change in Solvency II discounting curves.

There may be no buy-side firms on the Bank of England’s working group on sterling risk-free reference rates, but scrapping Libor may prove to be as much of a headache for UK insurers as for banks.

Insurers could find their hedges cost more, stop working and incur capital charges as they run basis risk over the years the market transitions from Libor to a new rate benchmark, most likely to be the sterling overnight index average (Sonia) in the UK.

Just like banks, insurers will face the lengthy and costly process of rewriting a large back book of interest rate derivatives linked to Libor. Many of the old contracts favour insurers, and newly drawn-up ones will come with a liquidity premium – up to 10 basis points feasibly – if the market in them is underdeveloped.

Insurers may find their hedges stop working as they hedge against moves in a new rate benchmark while maintaining exposure to Libor. Insurers will have to discount their sterling liabilities using Libor-based swap curves until European authorities change the methodology in Solvency II.

“Eiopa is closely monitoring the developments linked with Libor,” says a spokesperson for the European Insurance and Operational Pensions Authority. “But currently, no amendment of the legislation, nor material changes of methodology Eiopa uses to derive the curves, is planned.”

“Ultimately, on the switch date – and I don’t think anyone has the exact date – if Eiopa hasn’t updated the definition of the risk-free rate, insurers are going to be running basis risk that day. You’re going to need to hold capital for that basis risk,” says Michael Eakins, co-head of UK rates sales at Goldman Sachs.

Those Risk.net sought for comment see no way around the creation of basis risk, as the rewriting of hedging contracts will occur over a lengthy period, while a change in the discounting curve would be a discrete event.

Ultimately, on the switch date… if Eiopa hasn’t updated the definition of the risk-free rate, insurers are going to be running basis risk that day. You’re going to need to hold capital for that basis risk

Michael Eakins, Goldman Sachs

Insurers may be reluctant to switch their hedges until the discount rate changes, but waiting too long may mean liquidity in legacy Libor swaps drops off, opening insurers to price increases if all other swaps users have switched.

The BoE could encourage insurers to switch their hedges by setting out when and how the discount rate curve will be recalibrated, however, the setting of the discount rate is in the hands of the European Commission.

Andrew Smith, a professor at University College Dublin, thinks basis risk could be a permanent new problem for insurers if the BoE decides not to push for a change in Solvency II, which he thinks likely after the UK leaves the EU in 2019.

Solvency II legislation requires interest rate swaps to be used for discounting in all currencies “unless markets for them are not liquid”, says an Eiopa spokesperson. It is not clear what happens if markets for either Libor or Sonia swaps – or both – past a certain duration are illiquid.

“An awful lot of life insurers have cashflow between five and 25 years out, so it’s simply not practical to say we’re just going to use the market up to five years,” says Smith.

Many insurers Risk.net tried to contact for this article did not want to comment at such an early stage. “I haven’t started looking at it,” says the European regulatory chief for a large US insurer.

“Insurers are starting to think about alternatives to Libor, but it is quite early in the process. They are keen to minimise balance sheet volatility,” says Iain Forrester, head of insurance investment strategy for Aviva Investors.

Out with the old hedges

The UK’s Financial Conduct Authority (FCA) announced in July that banks will no longer be compelled to publish the interbank lending rate Libor beyond 2021. The Bank of England will take over the end-to-end administration of a revised Sonia in April 2018. This is the preferred alternative risk-free rate (RFR) of its working group – composed mostly of banks.

Elsewhere, the US has chosen the secured overnight financing rate (SOFR) as the successor to dollar Libor, and a Swiss working group has chosen the Swiss average rate overnight (Saron) to succeed Swiss Libor. Europe has yet to agree on a replacement for Euribor, with a number of secured and unsecured rates still in the running.

While the Libor administrator, Ice Benchmark Administration, says it intends to keep publishing the benchmark beyond 2021, banks are thought to have little desire to keep contributing to the panels.

If there are not enough banks to publish a viable Libor rate, many swap contracts in their current form do not specify what rate to use instead.

Libor

Industry groups are working on improving so-called fallback clauses in swap contracts to specify that in this situation, Libor-linked swaps should reference the new, currency-specific RFRs, such as Sonia. But as Sonia lacks the credit risk aspect of Libor, it’s intended for a spread to be applied on top to create a synthetic Libor rate. The FCA has proposed this spread to be based on a set of term bank credit spreads, while other groups are working on a plan to use the basis between Libor and reformed Sonia on the changeover day.

Others, such as Pimco, believe it’s easier to change out of Libor swaps and into those linked to a new RFR ahead of the end of Libor. Either way, this will require a mammoth repapering exercise.

“Longevity swaps and liquidity facilities insurers have in place will be very long-term: they could be 30 years or more. Rewriting all these contracts is going to be extremely inconvenient and time-consuming,” says Gareth Mee, leader of EY’s global investment advisory. “In the derivatives world, at the moment, some of the older contracts allow insurance companies to post certain types of assets as collateral they wouldn’t be allowed to do under the new contracts. There’s a possibility if insurers went back to the banks to renegotiate the contracts, those favourable terms will fall away,”

He also thinks it will be a “real pain” for banks, as they will either be accused of trying to profit from insurers’ woes or there will be political pressure for them to bear some of the operational costs.

The question for insurers is how quickly will we get the liquidity on the new rate and how fast is liquidity on Libor going to run off… Will they leave old contracts on Libor? What does that mean for their liquidity?

Marc Storan, EY

“The question for insurers is how quickly will we get the liquidity on the new rate and how fast is liquidity on Libor going to run off… Will they leave old contracts on Libor? What does that mean for their liquidity?” says Marc Storan, risk and actuarial adviser at EY.

He thinks the buy side will need to start trading Sonia swaps before sufficient volume is reached, and doing so will mean higher transaction costs in that period. Lower liquidity in the swaps market could also mean insurers having to take interest rate risk when they don’t want to.

A liquidity crunch in either Libor swaps, as insurers exit the market, or the still nascent Sonia swaps market could feasibly exact 10 basis points in transaction costs, which would amount to many millions in frictional costs for the insurance industry, thinks Mee.

William de Leon of Pimco
Bill de Leon, Pimco

“The worst-case scenario is that for some reason this product doesn’t really pick up liquidity and after the period for regulators to enforce publication of Libor, the banks stop publishing it. That’s the nightmare scenario. I don’t think that is likely,” says Bill de Leon, head of portfolio risk management at Pimco.

Laura Brown, head of liabilities-driven investment at Legal & General Investment Management, is more upbeat about Sonia swap trading volume as pension funds start trading Sonia swaps opportunistically.

“In terms of liquidity, we’re expecting it to be able to clear to 50 years before the end of the year, which will continue to help build this market. Our trading experience so far when implementing Sonia swaps has been very positive, with pricing just as competitive as Libor,” she says.

Liquidity at the long end of the Sonia swap curve may rise with buy-side firms’ trading them and some think those pension funds that do not discount their liabilities using Libor swap rates will have an incentive to switch first. “If you’re a pension fund, you’d want to change as soon as possible,” says Max Verheijen, pension fund adviser at Cardano, citing a drop-off in liquidity of back-book Libor swaps.

“Now this has been signposted by the BoE, the likely direction of travel is that existing portfolios are effectively earmarked as to become more illiquid,” says Andrew Kenyon, insurance solutions actuary at NatWest Markets.

“If with-profit funds have existing large Libor swap portfolios, it might end up being much more expensive to rebalance those portfolios, for example when reviewing things like guaranteed annuity option hedges,” he says.

University College Dublin’s Smith, who co-devised the Smith-Wilson RFR extrapolations Eiopa uses, is pessimistic about liquidity of overnight indexed swaps (OIS), which reference Sonia, with durations more than five years.

He thinks Eiopa could devise a new ultimate forward rate (UFR) to construct a UK discount curve, as used in Solvency II to extrapolate rates, but this could cause its own problems, not least the potential to be a source of ongoing arguments between regulator and industry, as it has been for German insurance policymakers.

“Although everybody recognises Libor has its failings, I don’t think many people would argue it was better to replace Libor with a made-up number between five and 25 years,” he says.

In with the new liabilities

Insurers will also face higher liabilities from a lower discount rate.

Solvency II RFR curves used for discounting sterling have two components: an interest rate swap rate tied to Libor, minus a credit risk adjustment to account for bank risk.

“If a sufficiently liquid [Sonia] curve emerged, it would be a bit lower than the current risk-free curve on average. The existing credit risk adjustment is crudely speaking 50% Sonia and 50% Libor, so there would potentially be a hit to insurers,” says Paul Fulcher, head of asset-liability restructuring for Nomura.

Smith, on the other hand, thinks the BoE probably won’t seek to change Eiopa’s discount rates after the UK leaves the EU in 2019, leaving insurers with the problem of basis risk for some years.

“There would be several reasons to do that. One is it would make the case for equivalence [with the EU] much easier. Second, it would remove the risk of arbitrage: selectively reinsuring certain trades to benefit from differences in regulatory yield curves,” he says.

bank-of-england-2016
The Bank of England

“You’re going to have to persuade Eiopa. Particularly for sterling, you’d have to do some pretty heroic extrapolation, and I don’t think they’d be comfortable with that either.”

“Why would you go through this numerical operational upheaval for something that would make very little difference? I would have thought they have much bigger things to worry about.”

A difficulty then might be how the BoE might seek to encourage insurers to use overnight indexed swaps such as Sonia under Solvency II.

“It’s going to be a bit chicken-and-egg,” says Fulcher. “Until the Solvency II RFR is set as OIS swaps, life companies will be incentivised to use Libor swaps. If life insurance companies are incentivised to use Libor swaps, liquidity doesn’t develop in the OIS market.”

“The way you break the egg is you mandate something,” says Fulcher. “Pre-Brexit, it would be for Eiopa to have the discussion, but post-Brexit the Prudential Regulation Authority could set a specific rate for the UK.”

Smith thinks the BoE’s PRA has other tools to encourage insurers to switch: “The PRA quite often asks insurers for supplementary regulation. They certainly could call up those insurers and ask them to recalculate their liabilities based on an OIS curve. The largest insurers in the UK – nearly 20 of them – use internal models [that] are subject to PRA approval… The PRA could ask insurers to investigate the possibility of a big move in the OIS curve.”

If the BoE wanted insurers to use the OIS curve and an ultimate forward rate of 4.2%, I suspect insurers would bite their arm off because it would give them a massively higher discount rate for returns between five and 50 years

Andrew Smith, University College Dublin

Smith believes insurers will not get dramatically different liabilities with an OIS curve under current conditions: “The complication is you’ve got a much greater degree of extrapolation in the OIS curve, so potentially many more arguments with the regulator. Things like UFR will suddenly become an enormously important discussion.

“To take an extreme example, if the BoE wanted insurers to use the OIS curve and an ultimate forward rate of 4.2%, I suspect insurers would bite their arm off because it would give them a massively higher discount rate for returns between five and 50 years.”

Problems with a transition

Regulators might try to avoid loading insurers with basis risk between two interest rate benchmarks by enforcing some sort of transition period, but there are difficulties here too.

“If you turn one curve to another curve, you might get an immediate capital hit – you’d think there might be some sort of transitional,” says Fulcher. “In reality, the market transition of hedges would proceed at one pace and the Solvency II transition would proceed at another. It can all get a bit noisy in the short term.”

“Without a transition, one might expect a rush out of one rate and into another, which would distort the market,” says Mee.

Regulators could create a transition period where two different discount curves could be used, he thinks, but this won’t solve all problems. “You can prolong the transition risk to many years, but basis risk will be created one way or the other,” he says.

Another option could be for regulators to allow insurers to discount using government bond curves, as some eastern European insurers do if the market for interest rate swaps is illiquid in their domestic currencies.

UK £10 note

“Some insurers feel like they want to move back to gilts. But the problem with a specified risk-free curve is that any other hedge has basis risk. Gilts are not particularly a good hedge for a swap curve, but nor are swaps a good hedge for a gilt curve,” says Fulcher.

“Giving insurers some flexibility on the choice of risk-free curve would help reduce procyclicality and market distortions. Pre-Solvency II, most UK insurers used gilts, but some such as Aviva used swaps, and similarly in the Netherlands we saw a range of different approaches.”

Pension funds that use swap curves to discount liabilities face similar problems to insurers, but some are also contemplating moving to a government bond discounting, says Verheijen. With Sonia swaps 20 basis points lower than Libor, pension funds would have approximately 4% higher liabilities, he reckons.

One transition route, proposed by Isda, is for the Libor-Sonia spread to be captured on the last days of Libor and used as a proxy for a credit risk adjustment, but many fear high trading on that day could distort pricing.

“Is it right that it is fixed for ever more? Even if it is a number of observation dates. If that is the way it is done, there is potential for distortions. If you had an averaging process to determine the spread, which would be different from the spot value on the last day of transition, there is an arbitrary transfer of value from one industry to another,” says Simon Hotchin, head of strategic solutions at HSBC.

Insurers and pension funds would prefer as tight a spread as possible between Sonia and Libor, but some worry hedge funds will seek to profit from widening the spread, before or during the snap period.

“There is a potential for manipulation if it’s a known event in the market. I’m not sure how you could control for that. If it’s sudden, there could be a general market disruption and you could be locking in those perhaps escalated spreads,” says Courtney Walker, senior portfolio risk manager at Pimco.

Insurers will have a high workload in amending their swaps and discounting – hard enough with regulatory uncertainty, even harder if hedge funds try to make money from it.


Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact info@risk.net or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact info@risk.net to find out more.

The future of life insurance

As the world constantly evolves and changes, so too does the life insurance industry, which is preparing for a multitude of challenges, particularly in three areas: interest rates, regulatory mandates and technology (software, underwriting tools and…

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here