The Future of LIBOR
For some years now, work has been ongoing to reform or replace LIBOR. Readers will be aware of the underlying factors of abuse of the LIBOR fixings by banks, but also the decline of liquidity in the markets supposedly underpinning LIBOR and the reality that, in a world where bank credits are more doubted than they once were, an inter-bank rate with a term structure has conceptual limitations as a general “risk free” reference point.
The G20 and Financial Stability Board concluded some years ago that some sort of replacement “nearly risk-free reference rate” was needed. At the end of July, the head of the Financial Conduct Authority, Andrew Bailey, made a concrete announcement on the topic, to the effect that LIBOR will be phased out in 2021. The thinking behind this move is summarised neatly in the observation from Mr Bailey that “We do not think we will complete the journey to transaction-based benchmarks if markets continue to rely on LIBOR in its current form.”
For anyone looking at a loan agreement referencing LIBOR with twenty years left to run, it is important to understand what this will mean and whether potentially painful negotiations with lenders will be required. There are other important issues also, such as practical questions of daily treasury management and hedge accounting.
The front-runner in the UK market appears to be SONIA, the Sterling Over-Night Index Average. This is a rate produced by the Bank of England based on overnight unsecured transactions between banks and building societies. It has been around since 1997 (LIBOR is also quite recent, it dates from 1985 and was created to support derivatives standardisation). The Bank is in the process of reforming and improving SONIA partly with its future as a replacement for LIBOR in mind. This is alongside developments in the Eurozone where the ECB has suggested in May that it might replace EURIBOR with a new reference rate, and in the U.S. where 15 banks voted in June to replace USD LIBOR with repo rates, which are backed by US treasuries.
One important limitation relates to term structure – it is an overnight rate published every day but there is at present no such thing as a “3-month SONIA”. There are technical arguments for leaving it that way and not trying to create a series of longer-term rates based on the OIS (Overnight Index Swap) market, which already exists as a swap market, but it seems likely that many users of LIBOR will want to keep the idea of a three or six-month rate which can be set at the start of each period, like with LIBOR-based loans and swaps, rather than have to calculate rates at the end of each period looking back at the daily compounding rate or the average over the period just ended. Perhaps it does not need to be that way, but there are practical systems implications as well as theoretical ones.
So, there is an important point around term structure still to play out. It may be, for example, that the demise of one-month LIBOR as an option for borrowers will be hastened and most borrowers rarely use e.g. 9 or 12-month rolls anyway – but changes of that order would be a lot easier to process than reconsideration of basic treasury processes and hedge relationships.
It will be highly desirable for borrowers if there can be general agreement that existing loans and swaps, as well as new ones, can simply be read on the basis of crossing out “LIBOR” and writing in “SONIA or relevant term equivalent” – with an absolute minimum of additional complexity and frippery if any – that is surely the objective but at the present time the debate is being carried out in somewhat theoretical terms. In due course, it may be useful to have a stronger voice from borrowers and we will be monitoring the situation partly with that in mind.