Jackson Hole was a disappointment for benchmark nerds.
At the annual gathering of central bankers, finance ministers, academics and financial market pros, the Federal Reserve did not announce or hint at any change in policy that may facilitate broader acceptance of the US’s proposed alternative to Libor, despite the theme of this year’s gathering being ‘Changing market structure and implications for monetary policy’.
Some analysts had hoped that the Fed would suggest a transition away from the effective federal funds rate as a tool for monetary policy, toward the secured overnight financing rate (Sofr). It didn’t.
But capital markets appear to be putting in a good early effort on their own.
Sofr is off to a good start, according to Subadra Rajappa, a research analyst at Société Générale in New York.
The alternative reference rate (ARR) endorsed by the US Federal Reserve has recently been used by Fannie Mae, the World Bank, Credit Suisse and Barclays as a benchmark for floating-rate deals in various formats.
Central clearing for Sofr swaps has already begun. Futures linked to the ARR have been launched by one exchange, with another to follow in October. And banks have transacted in a very small handful of over-the-counter swaps as test cases for that market.
“It’s hard to put your finger on why, exactly, Sofr is off to a good start,” Rajappa qualifies.
The development of markets linked to Sofr are indeed outpacing the Alternative Reference Rate Committee’s transition timeline. The trading of OTC swaps and futures linked to Sofr was projected by the ARRC to begin before year-end, and cleared swaps were projected to follow in the first quarter of 2020.
And, of course, there are the deals. Fannie Mae’s bond, which came in six-, 12- and 18-month maturities, was bought up largely by money market funds (MMFs), accounting for around two thirds of the deal, according to Mark Cabana, head of short rates strategy at Bank of America Merrill Lynch.
From the perspective of MMFs, there’s a lot to like, he says: short-weighted average maturity, so they’ll always trade around par with little volatility; daily reset rates; plus the bonds were seen as cheap compared with where Fannie could have issued over Libor.
That natural demand has manifested itself strongly in the two bank deals that came in late August. Credit Suisse sold a six-month $100 million certificate of deposit tied to Sofr, the first of its kind and the first Sofr-linked instrument from a bank.
Barclays, which acted as a lead on Fannie’s Sofr bond, soon followed CS with a whopping – for the asset-backed commercial paper market, and in late August at that – $525 million three-month deal off its Sheffield Receivables Corp ABCP conduit to demand that surprised the Barclays team behind it. Investors that had never bought from Sheffield were ringing up the bank to get into the deal.
These cash, front-end deals are indeed a measure of support for the market. But buyers of commercial paper, certificates of deposit and short-dated floating rate notes (MMFs and central banks, among them) don’t tend to trade them. And longer-dated instruments, cash and derivatives, are needed to build a discounting curve.
But short-dated cash deals are the starting point for a market with no history, argues Barclays’ Chris Conetta, managing director of US rates trading.
“Market participants, both investors and issuers, are in various stages of ensuring that their trading and risk systems can properly account for the new Sofr benchmark,” he says.
“There’s not much historical data on Sofr yet, and there is not much experience trading derivatives linked to it. So one would expect the initial issuances to be generally shorter-dated as the market gains experience in trading and hedging with this new benchmark. As that progress happens, it won’t be long before you see longer-dated cash instruments, provided there is a functioning derivatives market to complement it.”
It’s that “derivatives market to complement it” where the catch lies. It might not be too difficult for buy-side institutions to prepare their systems to deal with a backward-looking overnight rate. But many borrowers, especially in long-term debt, need cash-flow certainty, and an overnight rate doesn’t provide it, since a borrower only knows the rate it will pay for any given period at the very end of that period.
So, for the longer-dated cash market to take off, a term rate is needed. For a term rate, there needs to be a robust longer-dated derivatives market. But derivatives markets only come into existence as the need to hedge cash instruments arises.
“That’s where you have circularity,” says SocGen’s Rajappa. “The ARRC is talking about using a derivatives market to tease out a term rate, the cash market is looking for a term rate to issue bonds and loans, and the derivatives market is waiting for a cash market to develop.”
Earlier in August, the World Bank sold the first, and so far only, deal with a maturity longer than 18 months. While the bond was also an apparent success, two blocks of swaps into Libor out to two years traded the day before the issuance of the bond, as reported by the CFTC’s Swaps Data Repository, meaning the World Bank probably swapped its Sofr exposure, and is now paying Libor and receiving Sofr.
“This goes to the core of the problem,” Rajappa says. “As long as there are liabilities linked to Libor, there isn’t really an incentive to increase Sofr exposure.”
Market participants broadly expect to see more deals linked to ARRs offered soon, but they won’t be out of pure economic necessity. While issuers have claimed that their Sofr deals have been priced in line with where they would fund to Libor, many onlookers to the World Bank’s bond argue it came relatively cheap – something the World Bank has staunchly denied.
“The deals have catalyzed markets to make sure that their systems [for dealing with Sofr-based instruments] are up and running,” says Cabana.
But what is really needed, he says, is for banks to start making retail loans linked to Sofr. More bond issuance is good – the derivatives market will only develop as there is a greater need to hedge Sofr exposures. But the pivotal event will be banks making Sofr-based loans, which would require Sofr hedging right at the foundation of bank activity.
The use of Libor boomed originally because a derivatives market was created to hedge cash exposures to Libor. Now with hundreds of trillions of instruments linked to Libor, there’s a certain inertia to overcome in switching away from it.
Cabana argues retail borrowers would likely be indifferent to a change to Sofr as the base rate for their loans, since it at least doesn’t comprise an increment of bank credit risk – indeed Sofr rates should decrease as bank credit becomes riskier. But banks will want to hedge loans for that very reason: their cost of funds moves in tandem with the income they receive from loans priced over Libor.
To make matters a bit more complicated, Sofr is more volatile than the Fed funds rate, partly because of month-, quarter- and year-end spikes in repo rates due to post-crisis leverage requirements.
That causes Sofr to trade cheap to the effective Fed funds rate in the run-up to month-end before gradually dipping below into mid-month, and then spiking again in the run-up to the next month’s end.
Some causes of this dynamic are one-off phenomena, but the spread between Sofr and Fed funds also partly relies on Treasury bill supply, as bills are used as repo collateral. Since February, when the US debt ceiling was suspended, bill supply has increased, and is expected to continue at least until March 2019, when the extension ends.
If you talk to end investors, they’re really only using Fed funds and Eurodollar futures. As long as those are still around, there’s no reason to switch to Sofr futures- Subadra Rajappa, Société Générale
For another half a year, Sofr will likely generally trade above Fed funds, despite being collateralized, says Rajappa.
All this said, derivatives markets in ARRs have already come a fairly long way in a short period of time.
As of late August, trading in the CME’s Sofr futures contracts had steadily grown since its May inception, with total volume exceeding 200,000 contracts and open interest of more than 27,000, across 70 global market participants.
ICE is set to launch one- and three-month Sofr futures in October. This all a good sign, as the creation of term structures for Sofr, which is a backward-looking rate, requires a deep and broad futures market.
But there is still some way to go on that front: open interest and volume traded in one-month Sofr futures on August 18 were 6,611 and 1,243, respectively. That compares to 234,417 in open interest and 1,558 in volume traded in Fed funds futures.
For Sofr three-month futures, open interest was 1,124 and volume was 266. Eurodollar futures, in contrast, saw 1,277,825 in open interest and 190,755 in volume, according to Société Générale.
“If you talk to end investors, they’re really only using Fed funds and Eurodollar futures,” says Rajappa. “As long as those are still around, there’s no reason to switch to Sofr futures.”
As for Sofr swaps clearing, LCH began providing that in July – four have been executed there, Rajappa says.
LCH has taken the ARRC’s recommended approach: using the effective Fed funds rate for discounting and price alignment interest, or PAI, and eventually transitioning to a Sofr discounting curve and Sofr PAI.
(A technical note: PAI is a calculation meant to compensate the poster of variation margin (VM) on a rate swap for the overnight cost of funding it. Since VM is seen as collateral, not as a settled amount owed to the receiver, the payer on an out-of-the-money swap gets PAI paid back to her to mitigate that funding cost.)
CME, when it begins clearing Sofr swaps on October 1, will begin with the latter approach, which is apparently favoured by market participants. That means execution platforms will need systems to support two different discounting approaches to manage the risks associated with them, Rajappa says, introducing basis risk and inevitable user confusion.
Sonia futures have also made progress. ICE Europe launched one-month £3 million ($3.9 billion) notional contracts in December 2017, and three-month, £1 million notional contracts in June this year. Notional volume has hit some £190 billion, and the exchange says open interest continues to build. As of August 20, notional open interest stood at £37.8 billion.
On August 13, ICE introduced inter-contract spreads for three-month Sonia (the Sterling Overnight Index Average is an overnight rate, so investors don’t know what coupon they’ll be paid until the interest rate is calculated and set only days before the payment date) and three-month sterling. London Stock Exchange’s CurveGlobal launched three-month Sonia futures in April, as well as inter-commodity spreads between those futures and three-month sterling futures.
Most recently, CME announced that, pending regulatory approval, the exchange will launch its own Sonia contracts.
Market participants and associations generally find the timeline scenarios for transition at best challenging or too tight- Jaap Kes
In the UK’s Sonia cash market, there has been some similar progress, though less marked. The European Investment Bank sold a £1 billion five-year floating-rate bond linked to Sonia, the regulator-endorsed favourite to replace Libor, bringing in £1.175 billion in demand – no small feat in the sterling market. The deal was hailed as a big first step for that ARR.
M&G’s Bond Vigilantes blog says: “The deal may very well serve as a benchmark for future issuance in the Libor-less world.”
A banker on the deal told Euromoney in July that the bond was trading well, despite concerns from outside parties that most traders do not have the systems in place to value a bond referenced to a backward-looking rate.
He also said that sovereigns, supranationals and agencies would, “of course”, adopt it, as well as financial institutions, and quickly.
But others are less convinced. One treasury official at a European institution tells Euromoney: “I’m sure there were bank treasuries that bought [the EIB bond] because the regulator wanted to see it, not because [the bank treasuries] wanted it right now. Not all of them would have been able to put it into their systems.”
This official even questions if there ever would be a term Sonia they could use as a base rate.
Enter the EBRD, which, despite a veiled warning the Bank of England gave on July 23 that, really, all future sterling floating-rate issuance should be linked to Sonia, sold a Libor-linked sterling bond on July 24. The deal documents specify that, should Libor become “unavailable”, the calculation agent will decide whether to use a successor or substitute rate.
The deal documents imply some heavy scepticism about how widely used Sonia will be by the end of 2021, when banks are expected to stop submitting Libor.
If the calculation agent decides there is no “industry accepted successor base rate” for instruments linked to three-month sterling Libor, it will just ask big banks to give it quotes for three-month sterling unsecured deposits to prime banks in the London interbank market. In other words, it will revert to some kind of private Libor.
Given that banks don’t really lend to each other anymore and certainly not in three-month tenors (the whole reason regulators believe Libor to be vulnerable to manipulation), selling a bond that could pay based on a not-even-published Libor rate would be a remarkable move to make in the face of regulators that expressly desire issuers to stop using Libor.
Early September saw a step forward for Sonia, though, when Lloyds Bank sold a sterling covered bond linked to the benchmark, followed shortly after by RBC selling a sterling, Sonia-linked senior bond.
Jaap Kes, ING
RBC’s was a one-year bond, attracting the MMF crowd. Lloyds’ three-year covered bond, interestingly, had a three-day execution in order to allow time for investors to “assess their systems capability,” according to Peter Green, Lloyds’ head of public senior funding and covered bonds.
In Europe, the process of transition is only just beginning, with the ECB releasing the results of a consultation regarding its preferred ARR, Ester (euro short-term rate), in August.
Ester is an unsecured rate, so in one important aspect more like Libor.
In another important aspect, it is unlike Libor: it doesn’t exist yet. The ECB will begin publishing it by October 2019 – just two years before the Libor death-knell.
The ECB paper concluded that: “Respondents generally concurred with the working group’s conclusion that the euro short-term rate is the most reliable and robust – and consequently the most appropriate – unsecured candidate for the euro risk-free rate.”
Respondents generally “agreed” or “concurred” with almost everything the working group proposed.
But that the publication of the rate will only begin in the final quarter of 2019 presents serious risks to existing instruments benchmarked to Euribor and Eonia (Euro Overnight Index Average). Europe is some way behind its US, UK, Swiss and Japanese peers in preparing for a possible end to Libor publication in 2021.
Minutes of the ECB’s working group in July show concern among stakeholders that “there are no (or few) workable fallback provisions in place that would cater for the permanent cessation of existing reference rates.”
On top of that, “for legacy contracts, legal frameworks and market legal practices do not ensure a smooth transition to a new reference rate. Hence, the transition to a successor benchmark with legal certainty is not ensured.”
One representative in the group, Jaap Kes, ING’s head of money market & funding and interest rates, noted at the meeting that “market participants and associations generally find the timeline scenarios for transition at best challenging or too tight.”
Eonia is not compliant with the EU’s Benchmark Regulation, which becomes fully effective in January 2020. An overhaul of Euribor to make it compliant is not yet finished, and its eventual success still uncertain. And, according to the July minutes, while the US’s ARRC has a paced transition plan, the ECB group would only present a high-level plan “for all scenarios around Eonia and Euribor” at the September meeting.