Response to consultation on Regulatory Technical Standards on the method for identifying the main risk driver of a position and for determining whether a transaction represents a long or a short position
1. Do you agree with the general method for identifying the main risk driver of a non-derivative position and for determining its direction?
General comments
The main risk driver is not decisive for the trading or banking book allocation. A table of this kind merely creates effort on the part of the institution and offers no added value from a supervisory perspective. Determining the material risk driver is not necessary for the capital requirements of the FRTB and only serves reporting purposes here. We therefore suggest deleting the corresponding cells.
Furthermore, the template cannot be filled in automatically, as the allocation to the trading book or the banking book is usually included in the portfolio data. There is no provision for storing individual CRR allocation articles on the financial instrument. In addition, there may be different allocation reasons for each financial instrument type within a portfolio, e.g. to the trading book.
In general, we request an assessment of the implications of the indicated one-year postponement of the FRTB implementation date in accordance with CRR III.
Banks that apply the SA-CCR entirely use the method to identify the main risk drivers by determining sensitivities, which makes sense in principle. Conversely, banks that apply the SA-CCR in full but are not FRTB banks can only use a simplified procedure for non-derivatives. This means that different approaches are applied within a bank, which can lead to different treatments in the allocation of the risk category and in the allocation of long/short (a problem in the case of hedging).
There is no indication that a fall-back approach can be applied in the case of multiple risk drivers, as provided for in the RTS for SA-CCR (Commission Delegated Regulation (EU) 2021/931). The current proposal for the simplified method leads to a high implementation effort, especially for the calculation cores. For the introduction of a simplified approach in the allocation, additional information is required, such as the sector of the issuer, liquidity of the currency, etc.). It is therefore proposed that the simplified method also includes the option of using a fall-back approach, as the results would not be improved by deriving sensitivities.
Overview of questions for consultation
1. Do you agree with the general method for identifying the main risk driver of a non-derivative position and for determining its direction?
In general, the methodology for determining the primary risk driver and, based on this, determining the long/short classification is sufficiently clearly described and comprehensible.
The requirements for determining the primary risk driver are generally based on existing regulations in the SA-CCR context (Art. 277/EBA RTS on SA-CCR). The EBA RTS already contains complexity-reducing regulations, e.g. a simplified blanket definition of the "FX" risk factor for cross-currency swaps in Article 2.
The new regulation for "non-derivatives" not covered by SA-CCR (including securities) leads to significant implementation costs. In particular, as long/short classifications differ from existing classifications for accounting purposes and from the requirements of past EBA stress tests.
The incorporation of the "FX" risk factor in bond categorization is conceptually sound, providing a clear understanding of the process. However, it does lead to complex behaviour, particularly evident in the numerous case distinctions of the simplified approach. Furthermore, the main risk driver and thus the long/short classification of an instrument can change during the lifetime of an instrument, potentially leading to cliff effects and reduced transparency. Undesirable netting effects may also arise, such as in the case of short USD refinancing of longer-dated USD bonds.
Given these potential issues, it is crucial to consider the exclusion of the 'FX' risk factor in cases of pure translation effects. This could significantly reduce complexity and prevent undesirable netting or cliff effects. Implementing a similar simplifying requirement as is already in place for SA-CCR for cross-currency swaps could be a solution.
2. Do you agree with the analysis proposed in the background section and with the inclusion of this simplified method for fixed-rate bonds, floating-rate notes and stocks?
Providing a simplified approach is generally welcomed, as many banks do not apply the FRTB standardized approach or the SA-CCR.
We refer to question no. 1 regarding the strong complexity-increasing effect of “FX” as a risk factor.
3. Do you think that other non-derivative instruments should be included in the simplified method? If yes, please provide rationale and proposed treatment.
The simplified approach should cover the most common non-derivatives in smaller banks where possible. Fixed- and floating-rate bonds and equities are currently covered.
We recommend clarifying that the bond approach also applies to liability positions with a corresponding change of sign and money market and repo transactions. Similar to the treatment of equity and equity indices in the derivatives area, the presentation for equities should be extended to funds and ETFs. As the simplified approach for derivatives also considers commodity derivatives, a corresponding provision should also be derived for commodity positions. FX spot transactions are also relevant, and their classification should be similar to FX forwards.
Under the proposed amendments, banks approved for the “simplified method” by Art. 4 RTS will have the flexibility to derive the primary risk factor and the long/short classification for transaction types that are not explicitly mentioned. This empowerment avoids a costly relapse into the standard procedure.
The proposal also includes carefully considering bond products with options (e.g. reverse convertibles) under the simplified method. If these products do not account for a significant proportion of the portfolio, the main risk drivers (interest rates or equities) can be identified.
4. Do you agree with the general method for identifying the main risk driver of a derivative position and for determining its direction?
The close alignment with the existing Delegated Regulation 2021/931 to ensure consistency is understandable. However, it needs to be clarified whether the same approach is applicable for derivatives as for non-derivative instruments (the highest weighted delta sensitivity results in the material risk factor) or whether an analogy to the approach in the SA-CCR is mandatory. It would be helpful and sufficient if determining the material risk factor using the highest weighted sensitivity was also the method of choice for derivatives. A general alignment with SA-CCR would involve considerably more effort and have no added value.
We also refer to our general points on the general methodology under question 1.
5. Do you agree with the analysis proposed in the background section and with the inclusion of this simplified method for futures, options and swaps?
The provision of a simplified approach is generally welcomed. The derivation is comprehensible, and the resulting mapping rules are expressly welcome in their simplicity.
6. Do you think that other derivative instruments should be included in the simplified method? If yes, please provide rationale and proposed treatment.
The simplified approach should cover the most common derivatives in smaller banks. In the area of interest rate derivatives, the following products should be included in addition to the swaps already addressed:
- Forward rate agreements,
- swaptions,
- caps/floors and
- collars.
Interest rate options, interest rate warrants, caps, floors and collars are currently excluded. Standard hedging instruments cannot be determined using the simplified approach.
Restricting the simplified method's application to certain products means that a non-FRTB bank or banks that do not apply a full SA-CCR must introduce the risk-weighted sensitivity approach (“non-plain vanilla” product). For small and less complex banks, this means introducing sensitivity-based methods, which would entail increased expense.