Answer to the "Public consultation by the working group on €STR-based EURIBOR fallback rates"

Below are my answer to the "Public consultation by the working group on €STR-based EURIBOR fallback rates".

As previously mentioned, I believe that the consultation was deeply flawed with incorrect statements and unproven claims. Some technical analysis of one of the incorrect technical description of one of the conventions is available in a working paper available on a preprint server: Description of overnight floaters with principal adjustment and its advantages. An earlier version of the document has been sent to the official ECB email address related to the RFR working group. No answer has been received yet, which I presume means that the agree with the analysis but do not wish to comment about it.

The answer to the consultation had to be provided in an Excel sheet. This means that the answers are difficult to write and difficult to read. Moreover the mechanism limited the potential answers to some of the questions to a restrictive predetermined set. I have done my best to convert the Excel answers to a readable blog.

Criteria used in the analysis of EURIBOR fallbacks rates (see Section 5.2)

Question 1

Can you identify any additional criteria that should be taken into account?

Question 2

Do you agree with the analysis conducted in Section 5.2.1 and the conclusions of the working group presented in Section 5.2.2 with regard to the evaluation of the €STR-based term structure methodologies on the basis of the selection criteria?

The analysis is missing the best convention, the principal adjustment. The analysis of that convention is deeply flawed. The claims, without explanation, of table 2 are incorrect. In term of operational ease, computation complexity and hedging, the principal adjustment convention is similar to the ones analysed. The major advantage of that convention on the others, the fact that the rate is known at the start of the period, is unfortunately ignored in the document. Detailed analysis available at http://ssrn.com/abstract=3755906.

Corporate lending (Section 5.3.1)

Question 3

Do you agree with the working group’s conclusion that the backward-looking lookback period methodology would be the most appropriate methodology for building a €STR-based term structure that could function as a fallback for most, by value, of the corporate lending linked to EURIBOR?

Another alternative

As mentioned above, "Principal Adjustment" is the most appropriate method.

Mortgages, consumer loans and SME loans (Section 5.3.2)

Question 4.1

Do you agree with the working group’s conclusion that a forward-looking methodology would be the most appropriate methodology for building a €STR-based term structure that could function as a fallback for retail mortgages, consumer loans and SME loans linked to EURIBOR ?

Yes

Forward looking and backward looking compounded method are similar in term of valuation and risk management. A mixture of them should not create any issue for financial institutions.

Question 4.2

If your reply to Question 4.1 was affirmative, would you agree with the proposal to include a term structure built using a forward-looking methodology on the first level of the waterfall structure and, on the second level of the waterfall structure, to include as a backstop, in case a forward-looking term structure methodology is not available, either: a) a term structure built using the backward-looking last reset methodology (up to three-month tenors) or, alternatively; b) a term structure built using the backward-looking lookback period methodology?

Neither

If neither, what alternative would you propose for the second level of the waterfall? (Backward-looking payment delay / Another alternative)

Another alternative

Principal adjustment is the most appropriate method.

Question 4.3

Would you expect your institution to have to cope with any impediments in the case of a rate calculated using the backward-looking lookback period methodology for retail mortgages, consumer loans and SME loans?

Please indicate whether you are (representing) a lender or a borrower.

The consultation allows only (Lender/borrower). Answer is not available in the form: neither.

muRisQ advises both lenders and borrowers.

Question 4.4

Would you expect your institution to have to cope with any impediments in the case of a rate calculated using the backward-looking last reset methodology for retail mortgages, consumer loans and SME loans?

Yes

Please indicate whether you are (representing) a lender or a borrower.

The consultation allows only (Lender/borrower). Answer is not available in the form: neither.

muRisQ advises both lenders and borrowers. Last reset types payment are extremely difficult to price and hedge. We would advise both lenders and borrowers not to use it.

Current accounts (Section 5.3.3)

Question 5

Do you agree that the backward-looking payment delay methodology would be the most appropriate methodology for building a €STR-based term structure that could function as a fallback for EURIBOR for current accounts linked to EURIBOR?

Backward-looking base case. There is no reason to add a delay in this case. The account amount is adjusted for interest rate on a regular basis. This is similar to the "principal adjustment" method for bonds/loans, which is the preferred method. Adding a payment delay or a lookback period make the situation significantly more complex operationally and computationally. The current account are the perfect example where any atempt to solve a problem that does not exists make the situation significantly more complex. A current account has, by definition, day-to-day movements. There is no need to delay interest.

Trade finance (Section 5.3.4)

Question 6.1

Yes

Question 6.2

If your reply to Question 6.1 was affirmative, would you agree with the proposal to include: (i) a term structure built using a forward-looking methodology on the first level of the waterfall structure and (ii) a term structure built using the backward-looking last reset methodology on the second level of the waterfall structure as a backstop, in case a forward-looking term structure methodology is not available?

If not, what alternative methodology would you propose for the second level of the waterfall? (Backward-looking payment delay / Backward-looking lookback period / Another alternative)

Another alternative

As correctly mentioned in the document, in this case a forward-looking rate is necessary. If no forward looking benchmark is available, one could rely on lender offer and optional physical delivery. Mechanism would be something like: 1) the lender (bank) propose a forward looking rate for each transiaction (not a benchmark, but an actual rate, like 1.25%. 2) The financed entity has the option to accept the rate as is or to accept the rate with an optional delivery of an OIS where he or a party he designates receive the same rate (1.25%). Lender would not undercharge, to obtain a fair rate and would not overchage as their undue profit would be lost into the OIS. The convention contains a self-correcting mechanism. The option on the party is important for the borrower to be able to use a third party to compensate for the unfair rate.

Export and emerging markets finance products (Section 5.3.5)

Question 7.1

Do you agree with the working group’s conclusion that a forward-looking methodology would be the most appropriate methodology for building a €STR-based term structure that could function as a fallback for the majority of EURIBOR-linked products used for export and emerging markets finance products?

Yes

Question 7.2

Do you agree with the working group’s conclusion that for some export and emerging markets finance products – those involving sophisticated counterparties and developed markets – an in arrears methodology might be preferable and, in that case, a backward-looking lookback period methodology would be the most appropriate methodology for building a €STR-based term structure that could function as a EURIBOR fallback for such export and emerging markets finance products?

If not, what alternative methodology would you propose? (Forward-looking / Backward-looking payment delay / Backward-looking last reset / Another alternative)

Another alternative

The questions is ill-posed as it contains several parts that should be dealt separately. Part 1: In-arrears method prefereable: Yes. Part 2: backward-looking lookback period: No. Alternative method proposed: Principal adjustment.

Debt securities (Section 5.3.6)

Question 8

Do you agree that the backward-looking lookback period would be the most appropriate methodology for building a €STR-based term structure that could function as a fallback for EURIBOR-linked debt securities?

Backward looking rates may be appropriate for new issuance, but not so much for existing EURIBOR-linked debts. For those debts the strucuture of the deal, with the rate known at the start of the period, should be preserved as much as possible. The Principal Adjustment method is the method that offers that possibility, al least for all coupons but the last. The risk is similar to an overnight composition in arrears (with variable notional); this approach is compatible with other approaches and asset classes. The name is not the same but the "philosophy" is the same and there is no issue in term of risk management and hedging.

Securitisations (Section 5.4.7)

Question 9.1

Do you agree that for those securitisations that will include underlying assets for which the working group has identified the backward-looking lookback period as the most appropriate methodology for building a €STR-based term structure that could function as a EURIBOR fallback (e.g. syndicated loans, business loans and debt securities), it would be advisable to include the same EURIBOR fallback measure?

All fallback measures proposed in our answers are compatible in term of valuation and risk management with compounded in-arrears approach. They are compatible but not equal to. They all can be hedged in the same way. The different approaches (forward-looking, principal adjustment and compounded in-arrears) can be combined without risk, valuation or hedging complexity issue. The hedging of those securitisation where several of those measures have been used does not create a problem for any finance professional in charge of the securitisation. The fallback measures can be different in name, as long as they are compatible from a risk perspective.

Question 9.2

Do you agree that for those securitisations that will include underlying assets for which the working group has identified the forward-looking methodology as the most appropriate methodology for building a €STR-based term structure that could function as a EURIBOR fallback (e.g. mortgages and SME loans), it would be advisable to include the same waterfall structure as a EURIBOR fallback measure?

If not, what alternative methodology would you propose? (Backward-looking payment delay / Backward-looking lookback period / Backward-looking last reset / Another alternative)

Another alternative

Same answer than the one to Question 9.1

Transfer pricing models (Section 5.3.9)

Question 10.1

Do you agree with the working group’s conclusions that a forward-looking methodology would be the most appropriate methodology for building a €STR-based term structure that could function as a EURIBOR fallback for transfer pricing models for non-financial companies?

Yes

Question 10.2

Do you think that the backward-looking lookback period would be the most appropriate methodology for building a €STR-based term structure that could function as a EURIBOR fallback for transfer pricing models for financial companies?

Yes

Question 10.3

If your reply to Question 10.1 was affirmative (and/or your response to Question 10.2 was negative), would you agree with the proposal to include (i) a forward-looking term structure methodology on the first level of the waterfall structure and (ii) the backward-looking last reset term structure methodology on the second level of the waterfall structure as a backstop, in case a forward-looking term structure methodology is not available?

If not, what alternative methodology would you propose for the second level of the waterfall? (Backward-looking payment delay / Backward-looking lookback / Another alternative)

Another alternative

Last reset types payment are extremely difficult to price and hedge. Like for cash accounts and mortgages, the transfer pricing is likely to be long term with principal fluctuating through time. For those type of financial requirements, the "Principal Adjustment" method is the best approach.

Investment funds (Section 5.3.10)

Question 11

hich methodology – forward-looking or backward-looking lookback period – would be most appropriate for building a €STR-based term structure that could function as a EURIBOR fallback provision for benchmarking purposes for investment fund? (Forward-looking / Backward-looking lookback period / Another alternative)

Another alternative

Backward looking, base case, no lookback period. Benchmarking investment funds does not need immediate payment. The benchmarking is in anyway reported in arrears, when the fund investment performance is available. There is not reason to add a lookback period.

Spread adjustment (see Chapter 6)

Question 13

Please indicate whether you agree with the conclusion of the working group that the historical mean/median spread adjustment methodology should be the preferred approach for cash products.

If not, please rank the approaches discussed in Section 6.2. (dynamic spread adjustment methodology / forward spread adjustment methodology / spot spread adjustment methodology)

Dynamic Spread Adjustment

Original contracts refer to EURIBOR rate, which by definition has a dynamic credit spread. That important feature should be preserved as much as possible.

Question 14

Do you believe that having the same spread adjustment methodology for EURIBOR-linked cash products and other IBOR-linked cash products (the ISDA five-year historical median recommended by the ARRC and by the working group on sterling risk-free reference rates) is: a) essential; b) highly desirable; c) useful; d) unimportant.

An error decided somewhere are not a good reason to reproduce them somewhere else. Each fallback methodology should be as good as possible without imposing a cap on its quality at the mediocrity of other choices.

Question 15

Some cash products may fall back on backward-looking term rates fixing in arrears, while others may fall back on a forward-looking term rate or a backward-looking term rate fixing in advance. Therefore, do you agree that the spread adjustment value for each tenor should be the same, irrespective of whether the products fall back on a forward-looking or a backward-looking rate?

Yes

Also spread should be computed based on forward looking rate versus forward EURIBOR. Not backward looking ON versus forward looking EURIBOR. This exclude the ISDA-style incoherent computation methodology.

Question 16

With regard to whether the historical €STR market data are sufficient to compute any adjustment spread, do you agree that, even though there might not be sufficient €STR historical market data, data can be obtained by using historical EONIA market data with a fixed spread of 8.5 bps between the two indices, given that EONIA has been recalibrated to €STR + 8.5 bps?

The proposal is based on a logical fallacy. EONIA has been recalibrated to ESTR + 8.5 bps after 1 October 2019. Before that there is no fixed spread link between the two rates. There is no logical reason of such a link before that date.

Question 17

Do you think it is useful that for some cash products a one-year period would be applied for transition to the historic mean/median spread adjustment methodology? (yes / no / no opinion)

A transition period recreates a spot spread mechanism. Such spot spread mechanism has already been refused in previous consultation for good reason. There is no reason to reintroduce it. It is easy to manipulate and does not represent a present value neutral estimate.

Calculation methodologies and conventions (see Chapter 7)

Question 18

Do you agree with the working group’s conclusion that it would be useful for market participants to have access to a publication of the spread adjustment and/or an all-in rate that consists of (i) compounded €STR rates with an observation shift as proposed in Chapter 5, and (ii) a spread adjustment as proposed in Chapter 6? (yes / no / no opinion)

Compounded rats are product dependent. There is no way to produce "one" compounded rate that this coherent with the term sheet of a product that is unknown to the compounded rate provider. The publication of the spread make sense only of the spread is a dynamic spread.

Question 19

Do you agree with the working group’s view that if a floor were included, it should be on the sum of the €STR compounded rate plus the spread adjustment? (yes / no / no opinion)

Yes

Question 20

Do you agree that, in general, compounding the rate is the best calculation methodology? (yes / no / no opinion)

Yes

Composition is the best and most natural approach for interest rates. The composition can be achieved in direct or indirect ways. Indirect ways include forward looking based on composition instruments (OIS) and inclusion of the composition in the principal (Principal Adjustment method).

Question 21

Do you agree that the backward-looking lookback period term structure methodology with an observational shift is the preferable calculation methodology? (yes / no / no opinion)

Do you agree that the lag approach is a viable and robust alternative to the observation shift? (yes / no / no opinion)

The lookback period is not always necessary. If lookback actually means "lookback with non-zero shift", the answer is definitively no. If a lookback of zero day is part of the question, the anwser may be yes in some cases. The lag approach is in all cases a bad approach. This is basically unhedgeable and should be excluded in any financially meaningful method.

The answers have been sent to ECB on 6 January 2021.

Answers to more questions will be added later.