Response to consultation on Guidelines on common procedures and methodologies for the supervisory review and evaluation process (SREP)
Go back
In that context, we note that the principle of proportionality is referred to in a selective manner in the Draft Guidelines, i.e. there are sections which specifically refer to its application and others which do not. It is unclear whether this is rather a drafting choice to avoid repetition, or a conscious decision that proportionality will only be applied where explicitly stated so.
In general, we believe that proportionality – as an overarching principle of the Level 1 legislation – should be applied across all topics covered in the Draft Guidelines and there should be no doubt as to its application. More specifically, in relation to Section 4 on the Business Model Analysis, proportionality should be applied across all elements and levels of assessments to be performed by the competent authority. We would propose to include language to this effect in the Draft Guidelines.
More specifically in relation to this Question, we believe each supervisory process should be proportionate to the nature, size, complexity and inherent risk of the relevant investment firm.
Investment firms come in different shapes and sizes, but liquidity providers dealing on own account typically have a low risk profile: businesses are non-complex, portfolios are fully hedged and risk windows are very short (matter of days if not seconds).
We believe balance sheet size is no accurate proxy for a liquidity provider’s inherent risk. The business model of liquidity providers lies in the ‘spread’, i.e. the difference between buying and selling prices. Because this is done in high volumes, their positions on either side of the balance sheet, even for small firms, are usually sizeable (well above EUR 250m). However, the resulting net risk is low, because positions are fully hedged/netted. But because netting is not acknowledged under IFRS/GAAP, and poorly reflected under IFR, the balance sheet size does not reflect the actual risks in positions or the firm.
In this light we are concerned that using a generic balance sheet threshold requires smaller and mid sized firms – as well as their regulators - to apply a very detailed, often irrelevant SREP items, that serve no supervisory purpose. We feel a more sensible and efficient approach for both the regulators and liquidity providers, which focuses on actual risks at hand could be (1) to apply the threshold on a netted basis, e.g. by looking at Net Liquidity (which in itself provides a more accurate threshold) and (2) to permit regulators to apply items in the SREP framework on a case-by-case basis.
We are also concerned about how monitoring will work out in practice. Does COREP/FINREP suffice on a quarterly basis or is more needed? And how does the open standard of determining an ‘Event’ for Class 3 on page 28 of the Consultation work out in practice?
The key ground for the exercise of these powers is that there is activity that is posing excessive risk to the financial soundness of the firm. In Section 10.5 of the Draft Guidelines on common procedures and methodologies for the supervisory review and evaluation process (SREP) under IFD, Application of Other Supervisory Measures, Para 389 sets out guidance for a competent authority exercising its powers under Art 39(2)(e).
“In accordance with Article 39(2), point (e) of Directive (EU) 2019/2034, competent authorities may require the investment firm to make changes to the business model or strategy where:
a. they are not supported by appropriate organisational, governance or risk control and management arrangements;
b. they are not supported by capital and operational plans, including allocation of appropriate financial, human and technological (IT) resources; and/or
c. there are significant concerns about the sustainability of the business model.”
We believe that the current draft of Para 389 of the Draft Guidelines requires clarification to align it with the grounds established in the Level 1 text. The Directive text refers explicitly to excessive risks to the financial soundness of the Firm, but the connection of Para 389(a) to (c) to that basis is not articulated and unclear. The Guidelines suggest that a CA can require changes where the business model or strategy are not supported by the elements (a) to (c). However, this is not the basis of the power in the Level 1 text, which requires the presence of excessive risk to the financial soundness of the firm for the exercise of the power.
We note also that terms such as “operational plans” and “allocation” are undefined in this context, meaning that without clarifying that their assessment is connected to the evaluation of excessive risk to the financial soundness of a firm, the Guidelines might accidentally create alternative grounds for the 39(2)(e) power that are not contained in the Level 1 text.
Alternative proposal: we would therefore propose that the guidance emphasize that the exercise of the power is on the basis of excessive risk to the financial soundness of the firm, and clarify that areas (a) to (c) listed in Para 389 are elements of a firm’s operations that a CA may evaluate in its determination of whether the risk in a part of a firm is excessive, such as to warrant the exercise of the Art 39(2)(e) power.
Additionally, in Para 84 under c, the assessment of the investments firm’s risk appetite seems to be based upon the average risk appetite levels of a peer group. This setup comes with three concerns: (i) This relative measurement of this criterion is contrary to the measurements of the criteria in in Para 84 under a and b where an investment firm is being assessed on its own merits and not on those of its competitors. Based upon game theory, it would be possible that the best performing investment firm could become under severe scrutiny because his competitors, based upon their aberrant risk appetite, underperform. Putting the better performing investment firm on the spot creates a business incentive. This would be an unintended consequence of the setup of the measurement. (ii) It jeopardizes the required EU supervisory convergence because CA’s have the competence to consider per jurisdiction whether the investment firm’s risk appetite is high or transcend the average. (iii) Investment firms dealing on own account have a business model incomparable with other investment firms serving clients or customers. Based upon the trading and hedging strategies, the risk related to dealing on own account, is limited, only lies on the firm and not on any other market participant. The risk-to-market is small because if a dealer on own account might leave the market, competitors will instantly take this (released) position. Based upon the above and taken into account the essential link to the other criteria and the characteristics of the individual investment firm, we suggest to rephrase the criteria under Para 84 under c as follows: “risk appetite: competent authorities may consider whether the investment firm’s business model or strategy relies on a risk appetite that is considered not appropriate related to the assessment of the return on equity and cash-flow structure of the investment firm.”
Terminology: we also note that in contrast with the Directive text, which refers to “business, operations or network of a firm”, the guidance refers to “business model or strategy”. We would also propose that the guidance is amended to reflect the Directive text, replacing the terms “business model or strategy” with “business, operations or network”.
Regulators should be in the position to apply proportionate reviews given the nature, size, complexity and risk profile of the firm. Most liquidity providers are relatively small, low-risk firms, regardless of balance sheet size. In practice these firms have simple governance models as they are founder/employee owned, which is fit for purpose given the nature of their firms. Therefore, we believe the governance and control rules, that appear to be inspired by those for large banking groups, deserve a proportionate application.
The definition of DTF as introduced in level 1 is not limited to only some types of operational risks. Indeed DTF is referenced in recital 26 as covering operational risks to an investment firm with a trading activity with no explicit limitation of the type of operational risks. It is the industry understanding that DTF covers all operational risks arising in market makers and firms dealing on own account from their trading activity (which is the only activity that these firms have). If this were not the case, the legislation in level 1 would have made a specific reference to the type of operational risks covered by DTF or to the type of operational risks excluded by DTF for firms with a trading activity. The wording of the level 1 text evidences the policy option taken by the co-legislators, which these RTS should be in line with.
Paragraph 263 provides further evidence that this part of the RTS contradicts the level 1 text in this regard by including risks not covered by Pillar 1 risk relating to models, e.g. valuation, pricing models, models for algorithmic trading. The pricing model is a clear example of an internal system used for trading, a failure of which could result in a large volume of trades being concluded. If these model risks are not covered by DTF, it puts into question what DTF does cover.
We note further that the current approach, besides being in contradiction with level 1, would also create uncertainty for firms and create an unlevel playing field where some NCAs may treat some operational risks as covered by Pillar 1 and some not covered by Pillar 1. Should our reading or our understanding not be correct, we would kindly ask if the EBA can provide details on how the calibration for DTF was done at level 1.
In summary, the EBA methodology seems to presuppose that the risks in paragraph 257 are not covered by K-DTF. However, in line with the framework as envisaged by Level 1, investment firms will include these risks when making their assessment and compare it with DTF. Only in cases where Pillar 1 does not adequately cover these risks would additional P2R measures be warranted. As such, we recommend that paragraph 256 to 264 inclusive should be moved from the existing section ‘Operational risks not covered by Pillar 1’, and instead inserted within the ‘Daily Trading Flow’ section (paragraphs 225 to 226). In addition, paragraph 256 should be removed as K-DTF covers all Operational risk as substantiated above.
The risks identified in the discussion paper as candidates for a Pillar 2 addition are already covered by K-NPR, K-CMG and K-DTF. NCAs and firms should only have to identify Pillar 2 requirements in case risks are not, or inadequately, covered by Pillar 1.
Justification for the introduction of this floor seems to be given in recital 10 of the Consultation Paper on the draft Regulatory Standards on the Pillar 2 add-ons, that states “To ensure that additional own funds requirements remain adequate over time, it is necessary to adjust them proportionally to any significant increase of own funds requirements […]”. This ‘floor’ is then to remain in place at least until a new SREP assessment has taken place.
This is not in line with the practice of own funds requirements at investments firms and, we in our view, would not contribute to more effective prudential supervision. EBA seems to assume that own funds requirements can only ever increase for an investment firm. That is not necessarily the case for investment firms, and in particular for principal trading firms. In particular trading firms with market making obligations may see their own fund requirements both increase and decrease in relative short periods of time, depending on market circumstances – in particular trading volumes and volatility.
By way of example – during the 2020 Covid induced market volatility many market makers remained in the market to provide liquidity. The increase in trading activities of these firms led to a significant increase in own fund requirements from k-factors such as K-DTF, K-NPR and K-CMG. If a floor for additional pillar two requirements would have been in place in the Spring of 2020 these market making firms would have been faced with much higher capital requirements, until long after the markets calmed and own fund requirements from these K-factors would have come down again.
We do not support the approach that the risks listed in the paragraph 257 are “operational risks not covered by Pillar 1”. The purpose of P2R is to ensure supervised entities adequately capture the risks that are in Pillar 1, and to account for additional risks. As set out in Recital (26) to IFR, K-DTF covers all operational risks pertaining to trading. This is further stated in paragraph 225 of the RTS “Competent authorities should form a view on the degree of operational risk related to trading on own account”. Because market makers have no other activities, K-DTF sufficiently covers for operational risk. Also, paragraph 226 correctly includes risks pertaining to algorithmic trading in the section Daily Trading flow.
The definition of DTF as introduced in level 1 is not limited to only some types of operational risks. Indeed DTF is referenced in recital 26 as covering operational risks to an investment firm with a trading activity with no explicit limitation of the type of operational risks. It is the industry understanding that DTF covers all operational risks arising in market makers and firms dealing on own account from their trading activity (which is the only activity that these firms have). If this were not the case, the legislation in level 1 would have made a specific reference to the type of operational risks covered by DTF or to the type of operational risks excluded by DTF for firms with a trading activity. The wording of the level 1 text evidences the policy option taken by the co-legislators, which these RTS should be in line with.
Paragraph 263 provides further evidence that this part of the RTS contradicts the level 1 text in this regard by including risks not covered by Pillar 1 risk relating to models, e.g. valuation, pricing models, models for algorithmic trading. The pricing model is a clear example of an internal system used for trading, a failure of which could result in a large volume of trades being concluded. If these model risks are not covered by DTF, it puts into question what DTF does cover.
We note further that the current approach, besides being in contradiction with level 1, would also create uncertainty for firms and create an unlevel playing field where some NCAs may treat some operational risks as covered by Pillar 1 and some not covered by Pillar 1. Should our reading or our understanding not be correct, we would kindly ask if the EBA can provide details on how the calibration for DTF was done at level 1.
In summary, the EBA methodology seems to presuppose that the risks in paragraph 257 are not covered by K-DTF. However, in line with the framework as envisaged by Level 1, investment firms will include these risks when making their assessment and compare it with DTF. Only in cases where Pillar 1 does not adequately cover these risks would additional P2R measures be warranted. As such, we recommend that paragraph 256 to 264 inclusive should be moved from the existing section ‘Operational risks not covered by Pillar 1’, and instead inserted within the ‘Daily Trading Flow’ section (paragraphs 225 to 226). In addition, paragraph 256 should be removed as K-DTF covers all Operational risk as substantiated above.
The risks identified in the discussion paper as candidates for a Pillar 2 addition are already covered by K-NPR, K-CMG and K-DTF. NCAs and firms should only have to identify Pillar 2 requirements in case risks are not, or inadequately, covered by Pillar 1.
In case, despite these considerations, some form of floor is still considered required, as an alternative we propose the NCA may either decide to apply a fixed amount or a fixed ratio to the additional own fund requirements of a firm. While, in our view, neither of these alternatives would be appropriate from a prudential point of view, the latter of these two would be the most workable.
Finally, no information is given on the possibility and frequency of adjustments/improvement of the own fund requirement in case of an early remediation or closure of the actions identified by the CA. We feel such adjustments should be possible on a continuous basis through the common supervision dialogue, to stimulate to remediate defects as soon as possible and ahead of a new SREP assessment. Therefore, we suggest that investments firms might request an updated SREP assessment when as remedial actions have been implemented.
As we already explained in our answer to Q6 EBA seems to assume that own funds requirements can only ever increase for an investment firm. That is not necessarily the case for investment firms, and in particular for principal trading firms. In particular trading firms with market making obligations may see their own fund requirements both increase and decrease in relative short periods of time, depending on market circumstances – in particular trading volumes and volatility.
- Liquidity providers hold their cash and positions at banks that act as General Clearing Members (i.e. clearing banks). These require their clients to maintain a level of cash and ‘net liquidity’(i.e. a positive balance between long and short positions, including cash) to comply with their internal risk models as well as their own prudential requirements. These assets are mandatorily subject to a right of pledge, which will not be waived by GCMs.
- Additionally, it makes little sense to use only cash as eligible for meeting liquidity requirements. Positions in financial instruments held at a GCM are highly liquid and can be liquidated/converted into cash on very short notice. Therefore IFR/IFD permits firms to use liquid financial assets to cover their liquidity requirements.
- Finally, firms dealing on own account have no clients that could demand liquidity.
Therefore, we believe a more comprehensive liquidity assessment for market makers is not necessary.
Regarding section 1.2 we have to conclude that some definitions conflict with the principle of methodology that the definiendum should not appear in the definiens. As an example we point to the definition of ‘reputational risk’ where there is no circumscription of ‘reputation’. It also somewhat confusing that there is no single definition of risk. Risk is always used in relation to a form of risk, like risk-score, without attributing any thought when something is a risk/at risk.
Question 1: Do you agree with the proposed categorisation and the proportionate approach to the application of the SREP to different categories of investment firms?
To start with a more general remark on the whole Draft Guidelines, the 23 members of the Association of Proprietrary Traders APT, based in the Netherlands, welcome the fact that the Guidelines underline the principle of proportionality throughout the document. This principle is clearly set out in the Level 1 legislation and should be diligently pursued in the secondary legislation. We firmly believe that, in view of the variety of investment firms whether in terms of nature, size, complexity, risk or otherwise, proportionality is key to ensuring effective and efficient supervision which fulfils the objectives of the legislation. While the categorisation of the firms addresses proportionality to a certain extent, a multitude of levels remains within each category which, if not addressed by taking a proportionate approach, would create an unnecessary burden for the investments firms and the competent authority.In that context, we note that the principle of proportionality is referred to in a selective manner in the Draft Guidelines, i.e. there are sections which specifically refer to its application and others which do not. It is unclear whether this is rather a drafting choice to avoid repetition, or a conscious decision that proportionality will only be applied where explicitly stated so.
In general, we believe that proportionality – as an overarching principle of the Level 1 legislation – should be applied across all topics covered in the Draft Guidelines and there should be no doubt as to its application. More specifically, in relation to Section 4 on the Business Model Analysis, proportionality should be applied across all elements and levels of assessments to be performed by the competent authority. We would propose to include language to this effect in the Draft Guidelines.
More specifically in relation to this Question, we believe each supervisory process should be proportionate to the nature, size, complexity and inherent risk of the relevant investment firm.
Investment firms come in different shapes and sizes, but liquidity providers dealing on own account typically have a low risk profile: businesses are non-complex, portfolios are fully hedged and risk windows are very short (matter of days if not seconds).
We believe balance sheet size is no accurate proxy for a liquidity provider’s inherent risk. The business model of liquidity providers lies in the ‘spread’, i.e. the difference between buying and selling prices. Because this is done in high volumes, their positions on either side of the balance sheet, even for small firms, are usually sizeable (well above EUR 250m). However, the resulting net risk is low, because positions are fully hedged/netted. But because netting is not acknowledged under IFRS/GAAP, and poorly reflected under IFR, the balance sheet size does not reflect the actual risks in positions or the firm.
In this light we are concerned that using a generic balance sheet threshold requires smaller and mid sized firms – as well as their regulators - to apply a very detailed, often irrelevant SREP items, that serve no supervisory purpose. We feel a more sensible and efficient approach for both the regulators and liquidity providers, which focuses on actual risks at hand could be (1) to apply the threshold on a netted basis, e.g. by looking at Net Liquidity (which in itself provides a more accurate threshold) and (2) to permit regulators to apply items in the SREP framework on a case-by-case basis.
We are also concerned about how monitoring will work out in practice. Does COREP/FINREP suffice on a quarterly basis or is more needed? And how does the open standard of determining an ‘Event’ for Class 3 on page 28 of the Consultation work out in practice?
Question 2: Do you agree with our proposal regarding business model analysis? Are there any other drivers of business model/strategy that you believe competent authorities should consider when conducting the investment firms’ business model analysis?
NB. This answer has been coordinated with FIA EPTA. To start with, in this context we would like to respond to paragraph 389 of 10.5 on the Basis for Use of Other Supervisory Measures. Art 39(2)(e) IFD creates two ways in which a CA can exercise its intervention powers; “to restrict or limit the business, operations or network of investment firms or to request the divestment of activities that pose excessive risks to the financial soundness of an investment firm”.The key ground for the exercise of these powers is that there is activity that is posing excessive risk to the financial soundness of the firm. In Section 10.5 of the Draft Guidelines on common procedures and methodologies for the supervisory review and evaluation process (SREP) under IFD, Application of Other Supervisory Measures, Para 389 sets out guidance for a competent authority exercising its powers under Art 39(2)(e).
“In accordance with Article 39(2), point (e) of Directive (EU) 2019/2034, competent authorities may require the investment firm to make changes to the business model or strategy where:
a. they are not supported by appropriate organisational, governance or risk control and management arrangements;
b. they are not supported by capital and operational plans, including allocation of appropriate financial, human and technological (IT) resources; and/or
c. there are significant concerns about the sustainability of the business model.”
We believe that the current draft of Para 389 of the Draft Guidelines requires clarification to align it with the grounds established in the Level 1 text. The Directive text refers explicitly to excessive risks to the financial soundness of the Firm, but the connection of Para 389(a) to (c) to that basis is not articulated and unclear. The Guidelines suggest that a CA can require changes where the business model or strategy are not supported by the elements (a) to (c). However, this is not the basis of the power in the Level 1 text, which requires the presence of excessive risk to the financial soundness of the firm for the exercise of the power.
We note also that terms such as “operational plans” and “allocation” are undefined in this context, meaning that without clarifying that their assessment is connected to the evaluation of excessive risk to the financial soundness of a firm, the Guidelines might accidentally create alternative grounds for the 39(2)(e) power that are not contained in the Level 1 text.
Alternative proposal: we would therefore propose that the guidance emphasize that the exercise of the power is on the basis of excessive risk to the financial soundness of the firm, and clarify that areas (a) to (c) listed in Para 389 are elements of a firm’s operations that a CA may evaluate in its determination of whether the risk in a part of a firm is excessive, such as to warrant the exercise of the Art 39(2)(e) power.
Additionally, in Para 84 under c, the assessment of the investments firm’s risk appetite seems to be based upon the average risk appetite levels of a peer group. This setup comes with three concerns: (i) This relative measurement of this criterion is contrary to the measurements of the criteria in in Para 84 under a and b where an investment firm is being assessed on its own merits and not on those of its competitors. Based upon game theory, it would be possible that the best performing investment firm could become under severe scrutiny because his competitors, based upon their aberrant risk appetite, underperform. Putting the better performing investment firm on the spot creates a business incentive. This would be an unintended consequence of the setup of the measurement. (ii) It jeopardizes the required EU supervisory convergence because CA’s have the competence to consider per jurisdiction whether the investment firm’s risk appetite is high or transcend the average. (iii) Investment firms dealing on own account have a business model incomparable with other investment firms serving clients or customers. Based upon the trading and hedging strategies, the risk related to dealing on own account, is limited, only lies on the firm and not on any other market participant. The risk-to-market is small because if a dealer on own account might leave the market, competitors will instantly take this (released) position. Based upon the above and taken into account the essential link to the other criteria and the characteristics of the individual investment firm, we suggest to rephrase the criteria under Para 84 under c as follows: “risk appetite: competent authorities may consider whether the investment firm’s business model or strategy relies on a risk appetite that is considered not appropriate related to the assessment of the return on equity and cash-flow structure of the investment firm.”
Terminology: we also note that in contrast with the Directive text, which refers to “business, operations or network of a firm”, the guidance refers to “business model or strategy”. We would also propose that the guidance is amended to reflect the Directive text, replacing the terms “business model or strategy” with “business, operations or network”.
Question 3: Do you agree with the proposed criteria for the assessment of internal governance and firm-wide controls?
NB. This answer has been coordinated with FIA EPTA. Each supervisory process should be proportionate to the nature, size, complexity and inherent risk of the relevant investment firm. Investment firms come in different shapes and sizes, but typical liquidity providers dealing on own account have a low risk profile: businesses are non-complex, portfolios are fully hedged and risk windows are very short (matter of days if not seconds).Regulators should be in the position to apply proportionate reviews given the nature, size, complexity and risk profile of the firm. Most liquidity providers are relatively small, low-risk firms, regardless of balance sheet size. In practice these firms have simple governance models as they are founder/employee owned, which is fit for purpose given the nature of their firms. Therefore, we believe the governance and control rules, that appear to be inspired by those for large banking groups, deserve a proportionate application.
Question 4: What are the appropriate methods for the investment firms to analyse the potential impact of cyclical economic fluctuations on their activities and risks? Are they currently used by investment firms in their risk management processes?
NB. This answer has been coordinated with FIA EPTA. Liquidity providers (market makers) have very short risk cycles and their positions are typically hedged. Because they appropriate the spread between bid and ask prices as a business model and not by investing for the longer term, the direction of the market and economic cycles are largely irrelevant. Firms manage their risks well during turbulent markets in order to fulfil their market making obligations. Position risk (which should be almost flat, as these are typically fully hedged with related instruments) and liquidity risk is closely managed during such events in the ordinary course of business. Because the result of a trade is (almost) immediately cleared and settled, profits or losses are not subject to longer term exposures. The result of a trade (P&L) is immediately known and will not change over time. Therefore, cyclical risk and economic fluctuations are largely irrelevant to the risk profile of the firm.Question 5: Do you agree with the proposed criteria for the assessment of risks-to-capital? Does the breakdown of risk categories and subcategories provide appropriate coverage and scope for the supervisory review, having in mind various business models of investment firms?
NB. This answer has been coordinated with FIA EPTA. We do not support the approach that the risks listed in the paragraph 257 are “operational risks not covered by Pillar 1”. The purpose of P2R is to ensure supervised entities adequately capture the risks that are in Pillar 1, and to account for additional risks. As set out in Recital (26) to IFR, K-DTF covers all operational risks pertaining to trading. This is further stated in paragraph 225 of the RTS “Competent authorities should form a view on the degree of operational risk related to trading on own account”. Because market makers have no other activities, K-DTF sufficiently covers for operational risk. Also, paragraph 226 correctly includes risks pertaining to algorithmic trading in the section Daily Trading flow.The definition of DTF as introduced in level 1 is not limited to only some types of operational risks. Indeed DTF is referenced in recital 26 as covering operational risks to an investment firm with a trading activity with no explicit limitation of the type of operational risks. It is the industry understanding that DTF covers all operational risks arising in market makers and firms dealing on own account from their trading activity (which is the only activity that these firms have). If this were not the case, the legislation in level 1 would have made a specific reference to the type of operational risks covered by DTF or to the type of operational risks excluded by DTF for firms with a trading activity. The wording of the level 1 text evidences the policy option taken by the co-legislators, which these RTS should be in line with.
Paragraph 263 provides further evidence that this part of the RTS contradicts the level 1 text in this regard by including risks not covered by Pillar 1 risk relating to models, e.g. valuation, pricing models, models for algorithmic trading. The pricing model is a clear example of an internal system used for trading, a failure of which could result in a large volume of trades being concluded. If these model risks are not covered by DTF, it puts into question what DTF does cover.
We note further that the current approach, besides being in contradiction with level 1, would also create uncertainty for firms and create an unlevel playing field where some NCAs may treat some operational risks as covered by Pillar 1 and some not covered by Pillar 1. Should our reading or our understanding not be correct, we would kindly ask if the EBA can provide details on how the calibration for DTF was done at level 1.
In summary, the EBA methodology seems to presuppose that the risks in paragraph 257 are not covered by K-DTF. However, in line with the framework as envisaged by Level 1, investment firms will include these risks when making their assessment and compare it with DTF. Only in cases where Pillar 1 does not adequately cover these risks would additional P2R measures be warranted. As such, we recommend that paragraph 256 to 264 inclusive should be moved from the existing section ‘Operational risks not covered by Pillar 1’, and instead inserted within the ‘Daily Trading Flow’ section (paragraphs 225 to 226). In addition, paragraph 256 should be removed as K-DTF covers all Operational risk as substantiated above.
The risks identified in the discussion paper as candidates for a Pillar 2 addition are already covered by K-NPR, K-CMG and K-DTF. NCAs and firms should only have to identify Pillar 2 requirements in case risks are not, or inadequately, covered by Pillar 1.
Question 6: Do you agree with the proposed guidance for the setting and communication of additional own funds requirements?
We are concerned about the fact that paragraph 300 of the draft guidelines proposes to create a floor with respect to the additional Pillar 2 requirements. This is the result of the fact that according to paragraph 299 the additional own funds requirements are to expressed both (1) as an absolute amount, as well as (2) as the ratio of that amount to own funds requirements. Paragraph 300 subsequently states that the level of additional own fund requirements shall at all times be the highest of the absolute amount and the amount necessary to satisfy the ratio.Justification for the introduction of this floor seems to be given in recital 10 of the Consultation Paper on the draft Regulatory Standards on the Pillar 2 add-ons, that states “To ensure that additional own funds requirements remain adequate over time, it is necessary to adjust them proportionally to any significant increase of own funds requirements […]”. This ‘floor’ is then to remain in place at least until a new SREP assessment has taken place.
This is not in line with the practice of own funds requirements at investments firms and, we in our view, would not contribute to more effective prudential supervision. EBA seems to assume that own funds requirements can only ever increase for an investment firm. That is not necessarily the case for investment firms, and in particular for principal trading firms. In particular trading firms with market making obligations may see their own fund requirements both increase and decrease in relative short periods of time, depending on market circumstances – in particular trading volumes and volatility.
By way of example – during the 2020 Covid induced market volatility many market makers remained in the market to provide liquidity. The increase in trading activities of these firms led to a significant increase in own fund requirements from k-factors such as K-DTF, K-NPR and K-CMG. If a floor for additional pillar two requirements would have been in place in the Spring of 2020 these market making firms would have been faced with much higher capital requirements, until long after the markets calmed and own fund requirements from these K-factors would have come down again.
We do not support the approach that the risks listed in the paragraph 257 are “operational risks not covered by Pillar 1”. The purpose of P2R is to ensure supervised entities adequately capture the risks that are in Pillar 1, and to account for additional risks. As set out in Recital (26) to IFR, K-DTF covers all operational risks pertaining to trading. This is further stated in paragraph 225 of the RTS “Competent authorities should form a view on the degree of operational risk related to trading on own account”. Because market makers have no other activities, K-DTF sufficiently covers for operational risk. Also, paragraph 226 correctly includes risks pertaining to algorithmic trading in the section Daily Trading flow.
The definition of DTF as introduced in level 1 is not limited to only some types of operational risks. Indeed DTF is referenced in recital 26 as covering operational risks to an investment firm with a trading activity with no explicit limitation of the type of operational risks. It is the industry understanding that DTF covers all operational risks arising in market makers and firms dealing on own account from their trading activity (which is the only activity that these firms have). If this were not the case, the legislation in level 1 would have made a specific reference to the type of operational risks covered by DTF or to the type of operational risks excluded by DTF for firms with a trading activity. The wording of the level 1 text evidences the policy option taken by the co-legislators, which these RTS should be in line with.
Paragraph 263 provides further evidence that this part of the RTS contradicts the level 1 text in this regard by including risks not covered by Pillar 1 risk relating to models, e.g. valuation, pricing models, models for algorithmic trading. The pricing model is a clear example of an internal system used for trading, a failure of which could result in a large volume of trades being concluded. If these model risks are not covered by DTF, it puts into question what DTF does cover.
We note further that the current approach, besides being in contradiction with level 1, would also create uncertainty for firms and create an unlevel playing field where some NCAs may treat some operational risks as covered by Pillar 1 and some not covered by Pillar 1. Should our reading or our understanding not be correct, we would kindly ask if the EBA can provide details on how the calibration for DTF was done at level 1.
In summary, the EBA methodology seems to presuppose that the risks in paragraph 257 are not covered by K-DTF. However, in line with the framework as envisaged by Level 1, investment firms will include these risks when making their assessment and compare it with DTF. Only in cases where Pillar 1 does not adequately cover these risks would additional P2R measures be warranted. As such, we recommend that paragraph 256 to 264 inclusive should be moved from the existing section ‘Operational risks not covered by Pillar 1’, and instead inserted within the ‘Daily Trading Flow’ section (paragraphs 225 to 226). In addition, paragraph 256 should be removed as K-DTF covers all Operational risk as substantiated above.
The risks identified in the discussion paper as candidates for a Pillar 2 addition are already covered by K-NPR, K-CMG and K-DTF. NCAs and firms should only have to identify Pillar 2 requirements in case risks are not, or inadequately, covered by Pillar 1.
In case, despite these considerations, some form of floor is still considered required, as an alternative we propose the NCA may either decide to apply a fixed amount or a fixed ratio to the additional own fund requirements of a firm. While, in our view, neither of these alternatives would be appropriate from a prudential point of view, the latter of these two would be the most workable.
Finally, no information is given on the possibility and frequency of adjustments/improvement of the own fund requirement in case of an early remediation or closure of the actions identified by the CA. We feel such adjustments should be possible on a continuous basis through the common supervision dialogue, to stimulate to remediate defects as soon as possible and ahead of a new SREP assessment. Therefore, we suggest that investments firms might request an updated SREP assessment when as remedial actions have been implemented.
Question 7: What are your views regarding the interactions between SREP and internal processes of investment firms (such as recovery planning or ICARAP)?
/Question 8: Do you agree with the proposed guidance for the setting and communication of P2G? Would you consider it appropriate to express P2G not only as an absolute amount of own funds but also as a percentage of Pillar 1 own funds requirements? Please provide rationale for your views.
We do see the benefit of setting the P2G not only as an absolute amount, but also as a percentage of own funds requirements. We would however not support the creation of a de-facto ‘floor’ for the P2G via a requirement to always use the highest of both the absolute and the relative amount.As we already explained in our answer to Q6 EBA seems to assume that own funds requirements can only ever increase for an investment firm. That is not necessarily the case for investment firms, and in particular for principal trading firms. In particular trading firms with market making obligations may see their own fund requirements both increase and decrease in relative short periods of time, depending on market circumstances – in particular trading volumes and volatility.
Question 9: Do you agree with the proposed criteria for the assessment of liquidity risk? Should investment firms that deal on own account, in particular market makers, be subject to more comprehensive liquidity risk assessment?
During IFR/D’s Level 1 phase this topic has been extensively discussed. While maintaining unencumbered cash may seem relevant for banks under CRR, liquidity for investment firms (dealing on own account) is fundamentally different:- Liquidity providers hold their cash and positions at banks that act as General Clearing Members (i.e. clearing banks). These require their clients to maintain a level of cash and ‘net liquidity’(i.e. a positive balance between long and short positions, including cash) to comply with their internal risk models as well as their own prudential requirements. These assets are mandatorily subject to a right of pledge, which will not be waived by GCMs.
- Additionally, it makes little sense to use only cash as eligible for meeting liquidity requirements. Positions in financial instruments held at a GCM are highly liquid and can be liquidated/converted into cash on very short notice. Therefore IFR/IFD permits firms to use liquid financial assets to cover their liquidity requirements.
- Finally, firms dealing on own account have no clients that could demand liquidity.
Therefore, we believe a more comprehensive liquidity assessment for market makers is not necessary.
Question 10: Do you agree with the proposed guidance for the setting and communication of specific liquidity requirements?
We would appreciate the supervisor (1) to give clear qualitative and quantitative criteria that will be taken into account in the setting of the liquidity requirement and (2) to provide precise information upfront on the frequency of the review of the liquidity requirements. As improving Liquidity Risk Management and Reporting may not be reflected until the next Supervisory Assessment, we suggest that investments firms might request a updated SREP assessment when the required remedial actions have been implemented.Question 11: Do you have any views or suggestions with regard to appropriate incorporation of ESG risks within SREP, including any proposed methods or criteria for the assessment of ESG risks within SREP?
Proprietary trading firms are aware of the need to take into account the potential impact on ESG factors in business operations. However, not all applicable legislation is clear and (upcoming) legislation on this topic increasingly diverges and lacks consistency. The ESG component in the SREP Consultation is not clearly defined, guidelines are not conclusive and there is lack of detail. This creates uncertainties and makes it very difficult for investment firms to work with these rules in practice. We would like to urge EBA to take this into account when incorporating additional ESG risk criteria or requirements in the SREP assessment.Regarding section 1.2 we have to conclude that some definitions conflict with the principle of methodology that the definiendum should not appear in the definiens. As an example we point to the definition of ‘reputational risk’ where there is no circumscription of ‘reputation’. It also somewhat confusing that there is no single definition of risk. Risk is always used in relation to a form of risk, like risk-score, without attributing any thought when something is a risk/at risk.