We are seeing more new loans referencing SONIA from day 1, as more lenders get their IT systems ready to handle compounded RFR rates, and the publication by the LMA of a form of “switch agreement” (for a loan based on LIBOR on day 1 but ready to switch to SONIA some time during 2021) has provided impetus. That does not however mean the LMA document is simply being copied out: there are several ways of calculating compounded SONIA, and in particular two key variables: whether to compound cumulatively or not, and whether to have an observation shift or not, technical issues which make little or no difference in outcome – and so do not concern most borrowers - but which nevertheless need to be determined one way or the other. Non-cumulative compounding is more complex and most banks consider it unnecessary for their bilaterals or club deals – and we have seen one large syndicated RCF that dispensed with it too. The derivatives market has moved significantly to using SONIA rather than LIBOR for new contracts, and is in the process of converting existing contracts onto a SONIA basis – a “protocol” exercise being undertaken at present by ISDA in this respect will probably have, by the time you read this, over a thousand signatories. The Bank of England has started publication of a backwards-looking SONIA compounding index. To date we have, surprisingly, seen little use made of it.
An issue where consensus is building (but remains fluid) concerns the “credit adjustment spread” – the add-on required to SONIA where a loan based on LIBOR is to be converted to a SONIA basis, so that it represents what the corresponding LIBOR would have been. Different banks have different preferences for this. In the derivatives market, there is already a single convention – taking the median value of the LIBOR/SONIA spread over the last five years. The other method, used in several FRNs, is to refer to the OIS swaps market. Over the coming months, we expect that this issue will become more standardised, as the two principal methods of ascertaining it should converge.
As for the "tough legacy", legislative moves are afoot in the UK (in the Financial Services Bill currently before Parliament) to permit the FCA to authorise a so-called “synthetic LIBOR” to continue after LIBOR is phased out under the terms of the UK-onshored version of the Benchmarks Regulation, and which would allow the “tough legacy” contracts to continue to function.
Discussions about having a forward-looking “term” rate based on SONIA have been subdued this year, and the FCA is keen not to distract attention from moving as many users as possible away from LIBOR and only SONIA. Some unsophisticated treasurers may prefer this (and in the US, some regional US banks are very keen on the idea) and discounting activities need one, but in many cases there are alternatives – a fixed rate loan for example. We do not expect much regarding a forward-looking rate until the shift to SONIA is well underway in the cash market.
As regards US$ LIBOR, the position for the cash markets is possibly a little more nuanced, with more demand from US regional banks for a forward-looking rate to replace it (and this remains the first rung of the ladder of preferences of the NY Fed’s working group on LIBOR transition, even though it does not exist, or not yet anyway), and even for US$ LIBOR to continue beyond 2021 on a synthetic basis. This is still developing. The large US G-SIBs are all ready to shift to SOFR for cash products. We expect more clarity during the next few months.
The market is now wondering when the FCA will make its anticipated announcement that, as from 31st December 2021, LIBOR will officially no longer be representative of the underlying reality it purports to reflect. It had been expected before the end of 2020 but Q1 of 2021 now seems more likely.