On 26 February 2020, the Bank of England published a discussion paper Supporting Risk-Free Rate transition through the provisions of compounded SONIA, February 2020 (the Discussion Paper) which requested views from sterling market participants in relation to:
- the Bank of England’s intention to publish a daily SONIA Compounded Index; and
- the usefulness of the Bank publishing a simple set of SONIA Period Averages (being the compounded rate for a specific period (e.g. 30 days or 1 month) preceding the relevant day of publication) and how these periods should be defined.
The Discussion Paper followed an announcement by the NYFed that it would publish a daily SOFR Index and SOFR Averages from 2 March 2020. There are also currently proposals to publish similar compound rate indexes for risk free rates for other currencies.
The publication of official compounded indices is a significant milestone in facilitating the adoption of risk free rates, particularly in the loan market. The publication of a daily SONIA Compounded Index is likely to result in SONIA-linked bonds moving from using the “lag approach” to the “shift approach” to determine the applicable compounded SONIA rate for a period.
The rationale for a daily compounded index
The publication of an official compound index is intended to address concerns of participants in the loan market as to the complexity of using compounded risk free rates in arrears as the basis for determining the applicable interest rate. In particular:
- calculation of the compounded rate for a period may be determined by either (i) a simple calculation using the value of the compounded index on the first and last day of the period or (ii) where available, by reference to the relevant published period average; and
- the publication of an index by a trusted official source addresses concerns as to whether the relevant compounded rate for the period has been properly determined (i.e. if the correct daily rates have been used and all weightings for weekends and public holidays falling within the relevant period have been properly made).
From a regulatory perspective, it is hoped that the availability of compounded indices will promote a market consensus across multiple products to use risk free rates and the adoption of the “shift approach” for determining the applicable compound rate as a replacement to LIBOR.
Specific issues for SONIA-linked bonds
As noted in the Discussion Paper and the subsequent statement by the Bank of England’s Risk Free Rate Working Group, the SONIA Compounded Index can be used to determine the applicable compounded SONIA rate in any product that uses the “shift approach”.
The proposed SONIA Compounded Index cannot be used to determine the applicable compounded rate for products that use the “lag approach”, which applies to virtually all SONIA-linked bonds to date. Further, there is no intention to publish a SONIA compounded index that would be compatible for the purposes of using the “lag approach”. This is unlikely to change because (i) as noted in the Discussion Paper, while it is in theory possible to produce a compounded index for the “lag approach”, it would be necessary to prepare a separate index for each lag period (e.g. a 3 day lag compounded index, a 5 day lag compounded index, etc.) and (ii) regulators are unlikely to desire or encourage the publication of multiple indices for the same risk free rate given that it would conflict with the wider policy aim of promoting standardisation across different products.
In conclusion, new issuances of SONIA-linked bonds are likely to use the “shift approach” so that the SONIA Compound Index (if published) and SONIA Period Averages can be used to determine the applicable compounded SONIA rate for an interest period. The position for existing SONIA-linked bonds using the “lag approach” is unchanged – the relevant compounded SONIA rate will continue to be calculated using the daily SONIA rate to determine the rolling compounded SONIA on each day in the relevant interest period.
shift approach v lag approach
Under the “shift approach” both the weighting of the rate and day (the latter being necessary to account for weekends and public holidays) is shifted to the earlier observation period.
Under the “lag approach” the relevant observation period lags the interest period by a fixed number of days (typically 5 business days), but the day weighting is not shifted.
While the difference in day weighting to take account of weekends and holidays under the “shift” and “lag” approaches does not generally result in materially different compounded rates for the same period, the calculations are different and clearly unlikely to give the same result.