Press articles related to LIBOR fallback and new benchmarks
Two recent press articles have attracted my attention
Not a surprise there. My opinion, expressed for several years on this blog, is that "Change of benchmark [...] index is a difficult task" (Sep 2014).
One possibility for the Fallback, would be to write "multi-stage fallback" procedures. Some of them are described in my quant perspective on fallback. Related to the potential chaos for some loans and bonds, one could imagine the following approach: fallback triggers for derivatives that are more sensitive than the extreme "discontinuation of publication". One such trigger could be a "FCA verdict" as described in FCA: ‘We can be Libor fallback trigger’ (subscription required).
Once we know that the derivatives will trigger in a meaningful way, we can deal in a different stage with the extreme case of contract with no fallback and loose definition of LIBOR. For those extremes, it could be to agreed with the benchmark administrator (and the regulator) that instead of hard discontinuation, the administrator will continue to publish a figure, for example a RFR term rate plus a spread, but that rate would not qualify anymore as a benchmark for the new BMR. It is there only for "chaotic" legacy trades. In that way you avoid the chaos of the extreme cases due to the absence of fallback and at the same time leave the opportunity to the cases where a new fallback can be introduced to trigger it in a meaningful way with the replacement a meaningful alternative agreed by the parties.
A similar approach has already been discussed in the overnight benchmark question of transition from EONIA to ESTER. Some have proposed to continue EONIA beyond 2020 as ESTER plus a spread, with the new EONIA to be used only for legacy contracts and not for new ones.
The volatility of the SOFR had to be expected. EFFR is a rate "between friends" without bid/offer and with very limited market constraints. It is almost completely flat between FOMC meetings. The real lending market, like SOFR, has a bid/offer (around 10bps) and regulation driven intra-month seasonality. Those intra-month seasonality include month-end regulation-driven balance sizes and 15-of-the-month tax related demands.
I mentioned in a previous blog that this intra-month seasonality will have an important impact on valuation when moving from EFFR to SOFR collateral (and PAI). Which means more work to calibrate precise collateral discounting curves (and more related advisory engagements?).
It is not clear yet how much of this daily volatility extend to the monthly volatility (for example of the forward-looking term rates or the backward looking compounded rates). Some will certainly be present but how much of the daily spikes will reproduce on a monthly, quarterly or yearly basis has still to be seen.
Obviously instead of complaining about the volatility and developing complex models to deal with it, one could instead change the cause of this unhealthy behavior. For example, instead of having month-end and year-end measurements of balance sheet, why not a daily averages over a period. We have now real-time and one-line everything; would it be finance fiction to ask a bank to compute the size of its balance sheet (e.g. the notional of its repos in this case) on a daily basis? We are not talking here about traveling to the Moon and back, only to communicate to the regulators simple balance measures — that I hope the banks are already computing on a real-time basis for (risk) management purposes — on a daily basis. On the other side the regulators need to update their rules from a quarterly snapshot to an average daily snapshot; that should be feasible also.
Risk: Libor may linger as regulators ‘change tune’
International regulators appear to be coming round to the idea that Libor may have to be kept alive to avoid chaos in some loan and bond markets after the end of 2021.
Not a surprise there. My opinion, expressed for several years on this blog, is that "Change of benchmark [...] index is a difficult task" (Sep 2014).
One possibility for the Fallback, would be to write "multi-stage fallback" procedures. Some of them are described in my quant perspective on fallback. Related to the potential chaos for some loans and bonds, one could imagine the following approach: fallback triggers for derivatives that are more sensitive than the extreme "discontinuation of publication". One such trigger could be a "FCA verdict" as described in FCA: ‘We can be Libor fallback trigger’ (subscription required).
Once we know that the derivatives will trigger in a meaningful way, we can deal in a different stage with the extreme case of contract with no fallback and loose definition of LIBOR. For those extremes, it could be to agreed with the benchmark administrator (and the regulator) that instead of hard discontinuation, the administrator will continue to publish a figure, for example a RFR term rate plus a spread, but that rate would not qualify anymore as a benchmark for the new BMR. It is there only for "chaotic" legacy trades. In that way you avoid the chaos of the extreme cases due to the absence of fallback and at the same time leave the opportunity to the cases where a new fallback can be introduced to trigger it in a meaningful way with the replacement a meaningful alternative agreed by the parties.
A similar approach has already been discussed in the overnight benchmark question of transition from EONIA to ESTER. Some have proposed to continue EONIA beyond 2020 as ESTER plus a spread, with the new EONIA to be used only for legacy contracts and not for new ones.
Wall Street Journal: The Benchmark Set to Replace Libor Suffers Volatility Spike
The volatility of the SOFR had to be expected. EFFR is a rate "between friends" without bid/offer and with very limited market constraints. It is almost completely flat between FOMC meetings. The real lending market, like SOFR, has a bid/offer (around 10bps) and regulation driven intra-month seasonality. Those intra-month seasonality include month-end regulation-driven balance sizes and 15-of-the-month tax related demands.
I mentioned in a previous blog that this intra-month seasonality will have an important impact on valuation when moving from EFFR to SOFR collateral (and PAI). Which means more work to calibrate precise collateral discounting curves (and more related advisory engagements?).
It is not clear yet how much of this daily volatility extend to the monthly volatility (for example of the forward-looking term rates or the backward looking compounded rates). Some will certainly be present but how much of the daily spikes will reproduce on a monthly, quarterly or yearly basis has still to be seen.
Obviously instead of complaining about the volatility and developing complex models to deal with it, one could instead change the cause of this unhealthy behavior. For example, instead of having month-end and year-end measurements of balance sheet, why not a daily averages over a period. We have now real-time and one-line everything; would it be finance fiction to ask a bank to compute the size of its balance sheet (e.g. the notional of its repos in this case) on a daily basis? We are not talking here about traveling to the Moon and back, only to communicate to the regulators simple balance measures — that I hope the banks are already computing on a real-time basis for (risk) management purposes — on a daily basis. On the other side the regulators need to update their rules from a quarterly snapshot to an average daily snapshot; that should be feasible also.
Comments
Post a Comment