ISDA LIBOR fallback protocol - better late than never?

So it’s finally here!

Expect increased pressure from the official sector to get all major market participants to adhere to the protocol, and thus implement robust fallbacks for their existing derivative positions. Irrespective of whether you adhere for your backbook, new derivatives booked from the effective date in January onwards will all apply the new fallback language regardless.

We will be sending a more formal operational comms explaining the mechanics of the protocol following the official publication date in a couple of weeks (and prepare for your email inbox to fill up with other banks doing much the same), but for now below we make some broader observations about how this plays into ongoing discussions about pre-cessation announcements and the possible effect on LIBOR/RFR[3] basis.

The Basics

Before we delve into the triggers, prices etc, a few practical points on the protocol itself (general background on ISDA site):

  • It will be published on 23 October 2020, effective 25 January 2021. There is a facility for large organisations to sign up ‘in escrow’ before publication date, meaning on 23 October you will already see a host of names (including us) on ISDA website that have already adhered
  • When both sides to a derivatives contract adhere to the protocol, then all open trades between them will adopt the fallback language that was agreed through numerous consultations, meaning when LIBOR ceases to be published contracts will fall back to compounded RFR in arrears plus a spread adjustment
  • The spread adjustment will be calculated as the median value of the basis between LIBOR and the respective RFR for the given tenor over the 5 year look back period from the date of the announcement that the LIBOR rate will be considered non-representative (more on that later)
  • Clearing Houses (CCPs) have already said they will incorporate the ISDA fallback language into their rulebooks, meaning all cleared trades will apply robust fallbacks from the effective date

Any drawbacks?

  • The standard ISDA fallback language may not be suitable for some non-linear products. The Non-Linear Working Group is working through those issues, and it is likely counterparties will come to bilateral agreements on alternative fallbacks in such cases
  • The protocol may not be a suitable solution where swaps are tightly linked to other transactions, such as loan-linked swaps, and it may make more sense to transition the combination of transactions together at the same time, or document matching fallbacks (and triggers) across the package to ensure the components remain aligned
  • On that last point, depending upon how the discussions on Tough Legacy pan out, we might see the FCA4 use their new powers to ensure the LIBOR01 screen is redefined to Term SONIA5 + a spread (based on ISDA methodology) for use (only) by Tough Legacy; that would mean such transactions fell back in a manner consistent with the fallback protocol, just leaving a basis between realised and projected SONIA fixings

However for the vast majority of swaps the protocol will provide a simple and effective mechanism to ensure that existing derivatives will no longer end in contract frustration when LIBOR ceases to be published, and provide certainty about rates that will apply.

It does not mean that everyone downs tools on transition though....regulators still expect the majority of exposure to transition to alternatives proactively ahead of cessation rather than wait for the ‘insurance policy’.

That said there may be a growing contingent who will look to the fallback protocol not just as an insurance policy, but as a perfectly valid transition strategy, especially if it is combined with greater clarity on Tough Legacy. Expect more from clearing houses and compression providers to enable this eventual bulk fallback efficiently.

What’s the difference between the “announcement” and cessation?

This all goes back to pre-cessation triggers. The FCA can at any point announce that they consider that as of a certain date in the future that LIBOR (for a certain currency and tenor) is no longer representative. At the point it makes that announcement the clock stops on the 5 year lookback period – i.e. the spread adjustment will be calculated as the median of LIBOR/RFR basis for the 5 year period up until the announcement date. The spread adjustment will then be fixed as ‘X’, and that number used when cessation finally occurs...which might be a week or a year later (though not before January 2022).

And the noises from the official sector seem to indicate increasingly that this might be more like a year in the FCA might say in December 2020 that it expects LIBOR to be non-representative from 3 January 2022. The 5 year period for fixing the adjustment spread would then be December 2015 to December 2020.

We think it is possible the initial success (or otherwise) of the protocol, and greater clarity on tough legacy, might be influencing factors on whether the FCA make such an announcement early. The thinking being that an early announcement and a long lead in will give the market certainty on pricing and timing, and drive forward the pace of transition.

See the diagram below for an illustration of how the timings might work:

  • 1A - Announcement this year, in between ISDA protocol launch and effective date
  • 1B - Announcement early 2021, post ISDA protocol effective
  • 1C - Announcement end of 2021 once clarity that panel banks will stop submitting

What will happen to basis when announcement is made?

At the time of writing, the market is pricing a 25year 3month Libor vs Sonia Basis swap starting in 5years time as a 12.2bp middle, consistent with an early announcement.

In the graph below we show the historic realised 5 year median number up until today (in basis points), and going forwards the projected 5 year median number using a combination of historic and projected values. So market is suggesting if we get an announcement in December this year the spread will be fixed at around 12bp, and if it comes end of next year will be around 8bp.

Note other than through the timing of the announcement, it’s not easy to shift the 5 year median number....even if the pre-cessation announcement does not come for another year (i.e. end 2021), that only makes one more year of ‘new’ fixings, the rest are already in the past. So even with extreme volatility in LIBOR/RFR basis in the coming 12 months, it would have limited impact on the final spread adjustment numbers.

In graphs below we show one scenario where all the basis between the LIBOR/SONIA remaining fixings (in 2021) were way up at extremes we have not seen in last 5 years, and another where they are at or around zero, and the range this moves the median is from 15 to 8bp.  

So what’s going to happen next?

It’s still uncertain whether any announcement will come sooner rather than later...though market pricing would suggest sooner as you can see above.

But we can be sure there will be considerable pressure for the market as a whole to back the fallback protocol. Come two weeks’ time there will be a significant number of institutions that will have signed up ‘in escrow’, and we already know the CCPs are on board. Expect widespread encouragement for others to adopt the protocol too.

And that will at least gives some certainty in these uncertain times.

Phil Lloyd, NWM Sales
Dave Halstead, NWM Trading  
John Stevenson-Hamilton, NWM LIBOR Transition




International Swaps & Derivatives Association



London Inter-Bank Offered Rate



Risk-Free Reference Rates



Financial Conduct Authority



Sterling Overnight Interbank Average Rate

scroll to top