Fallback rule book: some comments
Some features of the IBOR Fallback Rate Adjustments Rule Book published recently by Bloomberg.
Presented without order and with very little comments, just as the ideas comes when reading. I will add more features as I go through the document.
The first comment was presented in a previous blog: Fallback and same day payment?
I have a XXX-XIBOR-1M with effective date Friday 2020-03-27. The maturity date is Monday 2020-04-27. The "Accrual Start Date" for the composition has an offset of 2 business days and is Wednesday 2020-02-25, the "End Accrual Date" is 1 month later, i.e. Saturday 2020-04-25, adjusted to Monday 2020-04-27.
The "two days shift" that was promised is actually a 0 day shift with respect to the payment date.
Example 1M-IBOR swap over 5-month period: 2020-03-26 to 2020-08-26. The standard accrual periods for the coupons payments are on dates 2020-03-26, 2020-04-27, 2020-05-26, 2020-07-26, 2020-07-27, and 2020-08-26. The accrual periods for RFR compositions are [2020-03-24, 2020-04-24], [2020-04-23, 2020-05-25], [2020-05-22, 2020-06-22], [2020-06-24, 2020-07-24], [2020-07-23, 2020-08-24].
This means overlaps in overnight composition of 1, 3, -2, and 1 days. Some days are double counted, while other are not counted at all. This is different from a standard OIS. How do you manage such a risk when SOFR has spikes. The excuse of "averaging" does not apply anymore here. It means also the requirement of a very precise curve calibration process, including the intra-month seasonality (see for example Fed Funds to SOFR: Impact of seasonality).
The document indicates that "Rate Adjustments are not themselves separate benchmarks for regulatory purposes." Has this been confirmed by regulators? For me the use of adjustment spread is a use of benchmark in the sense of the EU BMR (see my questions on that here). I have not seen explicit opinion of the contrary by regulators.
What is the legal status of the Rule book document? The new fallback definition will be embedded into the ISDA 2006 definitions. Then it will be those ISDA definitions that will be binding. The wording in the introduction of the rule book seems to imply that the document and even any following update will be "binding". Binding for who?
I'm still uncertain about the "calculate and distribute", does it means that Bloomberg computation will be the official rate or only a service to help computing the rate. What if someone get a result different from the one of Bloomberg, which one is legally binding, the one described in the ISDA definition or Bloomberg version?
The spread is computed between the LIBOR rate and the Adjusted Reference Rate (ARR). The ARR is the compounded in-arrears rate on the composition accrual period, i.e. including the two day offset. So the spread includes 3 parts: the credit spread (what we intuitively foresee with such a name), the misestimation of the market (forward looking v backward looking) and, the new part, the operational shift for payment. I have not yet computed the impact of that last element. We will see when we run the detailed computations.
With the definition of Median Period, it means that the LIBOR spreads for different tenors will not be computed on the same periods.
The "Median Period End Date" is defined by reference to a date that is "Tenor period immediately prior to" another day. In finance, to the best of my knowledge, there is not conventional way to move backward. There are standard definition of moving forward, e.g. Modified Following, but no backward. With respect to moving forward, the functions are not bijective, two dates may end-up on the same forward date.
The Fallback Transformer will be updated soon to include the new fallback definition. Detailed impact on trades risk and valuation will be available in the coming days.
IMM swaps have to be considered as "exotics" and the proposed fallback does not work for them. For example a IMM period starting on Wednesday 2022-12-21 will end on Wednesday 2023-03-15. The composition is done on the period from Monday 2022-12-19 to Monday 2023-03-20 (19 adjusted following). The payment is due 5 days before the amount is known. Hopefully you don't have IMM swaps in your book, if you do, better plan for the fallback of the fallback.
Obviously the fallback will not work for FRA with ISDA FRA discounting settlement, LIBOR in arrears or Range accruals.
Presented without order and with very little comments, just as the ideas comes when reading. I will add more features as I go through the document.
The first comment was presented in a previous blog: Fallback and same day payment?
I have a XXX-XIBOR-1M with effective date Friday 2020-03-27. The maturity date is Monday 2020-04-27. The "Accrual Start Date" for the composition has an offset of 2 business days and is Wednesday 2020-02-25, the "End Accrual Date" is 1 month later, i.e. Saturday 2020-04-25, adjusted to Monday 2020-04-27.
The "two days shift" that was promised is actually a 0 day shift with respect to the payment date.
Example 1M-IBOR swap over 5-month period: 2020-03-26 to 2020-08-26. The standard accrual periods for the coupons payments are on dates 2020-03-26, 2020-04-27, 2020-05-26, 2020-07-26, 2020-07-27, and 2020-08-26. The accrual periods for RFR compositions are [2020-03-24, 2020-04-24], [2020-04-23, 2020-05-25], [2020-05-22, 2020-06-22], [2020-06-24, 2020-07-24], [2020-07-23, 2020-08-24].
This means overlaps in overnight composition of 1, 3, -2, and 1 days. Some days are double counted, while other are not counted at all. This is different from a standard OIS. How do you manage such a risk when SOFR has spikes. The excuse of "averaging" does not apply anymore here. It means also the requirement of a very precise curve calibration process, including the intra-month seasonality (see for example Fed Funds to SOFR: Impact of seasonality).
The document indicates that "Rate Adjustments are not themselves separate benchmarks for regulatory purposes." Has this been confirmed by regulators? For me the use of adjustment spread is a use of benchmark in the sense of the EU BMR (see my questions on that here). I have not seen explicit opinion of the contrary by regulators.
What is the legal status of the Rule book document? The new fallback definition will be embedded into the ISDA 2006 definitions. Then it will be those ISDA definitions that will be binding. The wording in the introduction of the rule book seems to imply that the document and even any following update will be "binding". Binding for who?
I'm still uncertain about the "calculate and distribute", does it means that Bloomberg computation will be the official rate or only a service to help computing the rate. What if someone get a result different from the one of Bloomberg, which one is legally binding, the one described in the ISDA definition or Bloomberg version?
The spread is computed between the LIBOR rate and the Adjusted Reference Rate (ARR). The ARR is the compounded in-arrears rate on the composition accrual period, i.e. including the two day offset. So the spread includes 3 parts: the credit spread (what we intuitively foresee with such a name), the misestimation of the market (forward looking v backward looking) and, the new part, the operational shift for payment. I have not yet computed the impact of that last element. We will see when we run the detailed computations.
With the definition of Median Period, it means that the LIBOR spreads for different tenors will not be computed on the same periods.
The "Median Period End Date" is defined by reference to a date that is "Tenor period immediately prior to" another day. In finance, to the best of my knowledge, there is not conventional way to move backward. There are standard definition of moving forward, e.g. Modified Following, but no backward. With respect to moving forward, the functions are not bijective, two dates may end-up on the same forward date.
The Fallback Transformer will be updated soon to include the new fallback definition. Detailed impact on trades risk and valuation will be available in the coming days.
IMM swaps have to be considered as "exotics" and the proposed fallback does not work for them. For example a IMM period starting on Wednesday 2022-12-21 will end on Wednesday 2023-03-15. The composition is done on the period from Monday 2022-12-19 to Monday 2023-03-20 (19 adjusted following). The payment is due 5 days before the amount is known. Hopefully you don't have IMM swaps in your book, if you do, better plan for the fallback of the fallback.
Obviously the fallback will not work for FRA with ISDA FRA discounting settlement, LIBOR in arrears or Range accruals.
Comments
Post a Comment