The offset will also be applied to the historical spread used in the fallback. It means that the spread used in the fallback will be impacted by three distinct features: credit/liquidity, market misestimation and offset. The first two have been discussed extensively in different blogs and publication, the third one is new. The interest rates used to measure the spread are related to different periods. Depending if the monetary policy is in a rate increase or a rate decrease cycle over the 5-year lookback period, the intuitive spread will be overestimated or underestimated.
The graph below is the time series of the spreads since SOFR/pre-SOFR has been published (August 2014). As usual in my graphs, the dates reported are the LIBOR fixing dates (not the maturity date) and the SOFR related numbers stop 3 months ago because of the "in-arrears" feature.
The second graph is a focus of the offset impact. Two time series of spreads are computed: one with the two business days offset and one with zero business day offset. The spreads with offsets are a little bit bumpier, potentially due to difference in length in the LIBOR and ON periods.
All the graphs in our blogs are produced by production grade libraries. Our libraries include the Fallback Transformers which allow to apply the exact fallback details to actual trades and portfolios. The transformation details include overnight composition, period offset, spread applicable, and discontinuation date. It is suitable for what-if analysis and risk management.
Don't fallback, step forward!
Contact us for LIBOR fallback and discontinuation: trainings, workshops, advisory, tools, developments, solutions.