We do not believe that the inclusion of profit and loss (P&L) from the change in own credit spreads from derivative liabilities (DVA) should be included in the Pillar 1 framework, but that the current exclusion from Pillar 1 capital remains appropriate.
Whilst we appreciate that in applying this treatment to DVA, there would be some consistency in also excluding positions in own debt that sit within the trading book, we believe that these positions should be included in the internal models approach (IMA). They share characteristics with other trading book positions and from an operational approach it may be difficult to exclude own credit where included as an index, especially for options on the index. The operational challenge for these positions arises both from extracting the underlying position and excluding from time series data for any such indices.
The current practice is in line with the EBA’s Guidelines on Incremental Default and Migration Risk Charge, published in May 2012, mandate inclusion of trading positions in own debt within the IRC for migration only and by extension within VaR and SVaR. DVA, by contrast is excluded according to Article 33 of the Capital Requirements Regulation (CRR).
Excluding own credit from trading positions would be particularly problematic where the own credit component arises in the form of the banks membership of public indices, or other external measures (e.g. the ITraxx index).
In particular, where option structures exist with own credit as part of an underlying reference index, exclusion would require that the implied volatility surface used in Value at Risk (VaR) to be adjusted to reflect the exclusion of the banks own name. This would require maintaining a synthetic time series that is not directly observable in the market. We would therefore support continuing including own-credit exposures of this type in trading VaR.
As per Paragraph 75 of Basel III (regarding application of own credit risk adjustments to derivatives), both day 1 and incremental gains and losses arising from own credit are excluded from core equity tier 1 (CET1) calculations. Therefore including DVA on derivative liabilities in the trading VaR calculation would be asymmetric to this treatment. It is widely regarded by the industry that DVA on derivatives is not only counterintuitive, but is also extremely difficult to monetise through a hedging strategy, and creates perverse incentives for the firm (ie. certain trading desks benefit from the wider firm’s credit deterioration).
Furthermore, with the upcoming IFRS9 standard, DVA will be included with in Other Comprehensive Income (OCI) and not part of ‘above the line’ P&L earnings, in recognition of these points. We therefore do not support inclusion of DVA in VaR, in order to be consistent with these broader and well-established principles.
Under the Fundamental Review of the Trading Book (FRTB), the risk categories of interest rate (IR) and equities split into general and specific risk will cease to exist, with IR split into IR and Credit Risk and equities amalgamated into a single category. We therefore suggest the EBA align its approach with the FRTB.
As noted in the CP, whilst an application for specific risk approval without general risk approval has so far not been recorded, there may be situations where it is reasonable to do so.
Whilst the inclusion of certain products within the IMA scope is an indication of complexity, the principle of proportionality would be better applied by considering the size of the model in question. We would suggest that Competent Authorities (CAs) could consider this in light of the RWAs generated from the IMA as an absolute number and as a proportion to the overall RWAs for the institution. These numbers are already available and therefore easily available for CAs without requiring that new systems by built to assign externally generated categories to positions.
We understand that the proposed assessment for measuring significant share should only be applicable where an institution applies for permission to use IMA on a consolidated or solo basis for a given risk category. As such the proposed measure using Standardised Approach (SA) RWA will only be required for risks previously capitalised under this approach.
The proposed measurement of significant share through standard rules RWAs is consistent with the ongoing requirements for determining significant share using RWA both under the continuing assessment and looking further ahead the overall capital measure under FRTB.
We do not believe that the assessment for significant share should be differentiated across general and specific risks. That said, we agree that it is appropriate to consider the measurement of significance taking into account the effect of positions excluded by a CA in the granting of a permission, but do not agree that the measure proposed (of a reduced percentage) is the appropriate methodology to address this.
There can be reasons as to why an exclusion may manifest itself at a product or position level rather than a risk basis, either as a result of a CA permission, or from the operational ability to process risk categories separately (e.g. under Full Revaluation).
The need to take into account excluded positions, therefore, should cover any risk category under the IMA permission, not just specific risk. We propose that the significant share calculation should consistent for both general and specific risk categories, based upon only the positions for which permission is granted.
The term “significant share” used in CRR 363.2, as acknowledged in this CP, is designed to avoid empty models which were given permission, but are not used.
We question the EBA’s interpretation of the tem “significant share” in arriving at a value of 90-95%, which does not seem consistent with the usual meaning of the term. A “significant share” would suggest that a notable share or share worthy of attention, rather than the vast majority or materially all.
We would compare the significance threshold directly to the minimum modelled threshold in the FRTB, which is set at 10% of Market Risk capital requirements coming from IMA approved desks, when considering the eligibility of a bank to calculate using Internal Models.
Implementing a threshold of 90-95% under the current regulatory requirements, when future requirements will use the 10% threshold, implies a divergence of opinion between regulators which could call into question the applicability of any short term requirement. As such we believe that it is appropriate to adopt a consistent definition of terms where possible throughout European regulation and impose a 10% level as appropriate.
The own fund RWAs are measures which are already reported, are well defined and understood. We agree that this represents an applicable measure for a share of significance. This is also consistent with the direction of travel under the FRTB, where the overall level of RWAs included under IMA vs. Standardised Approach is measured in this way.
The daily P&L measure whilst derived from existing metrics, may require additional operational and technical development in allocating P&L explicitly to the six risk categories.
We do not believe that the proposed additional P&L-based measure would result in a materially different approach from banks to positions that are excluded from the IMA. As a result we do not believe that implementing this measure on a temporary basis ahead of FRTB represents a reasonable cost-benefit decision.
We have concerns relating to the prescriptive nature of the text across articles 17, 18 & 21.
Effective governance within an organisation frequently includes committees as part of the structure, in order to ensure that decisions take into account various stakeholder views. It should be noted, however, that committees can also be bureaucratic and not well suited to day-to-day running of an institution.
The text mandates that the management body, or a committee designated by it, is required to approve overall committee structures, including: function, membership and frequency of any such committee(s).
Prescribing this as a Management Body requirement prevents an institution from assigning any such responsibilities to a function or individual in charge of such a function and allowing their discretion regarding whether a committee is required to address particular elements of risk management.
Whilst certain committees are likely to be incorporated into this structure, a direct designation from the Management Body may not always be appropriate.
Consequently, we propose that this requirement be amended so that a function or member of the Management Body can be given responsibilities for designating the responsibility to an individual or committee where required, rather than this being required from the Management Body or related committee directly.
An example of this might be where the Head of Market Risk is given the responsibility by the Chief Risk Officer (CRO) (as management body member) for ensuring appropriate governance around IMA model development and usage (excluding independent validation). A decision over whether a committee is appropriate for this purpose and the governance around any such committee should therefore fall within the remit of the Head of Market Risk.
We propose that a distinction should be made within the category of validation between the following categories:
Validation which is the exclusive purview of the independent validation function;
Activities in the context of ongoing monitoring, which can be performed by other functions (for example backtesting, which is used for capital and ongoing model assessments; calculating the impact of missing risk factors; impact of different correlation factors as mentioned in Articles 46.2(b) and article 60.2(c) in the consulted RTS) and the output then incorporated in periodic independent validation; and
Reviews which are performed elsewhere, for example adequacy of the implementation of the IT systems, which could be covered by change functions and audit.
We propose that some of the mandatory requirements listed are not required. For example validation triggers as outlined in Art. 23 3. (d), (g), (h) as IT system changes should not necessarily trigger validation activities, nor large losses in stress testing results with extremely severe scenarios, or decrease stressed VaR/VaR ratios which may be caused by a worsening of the market environment driving an increase of
Whilst we agree with the principle of “independent” validation and separateness from the developers of a model, the proposal in Article 22 of the proposed RTS that an Independent Validation unit must be separate from the Risk Control Unit (department), is neither consistent with the CRR, nor a practical approach. CRR Article 368.1 (b) states that:
“The institution shall have a risk control unit that is independent from business trading units and reports directly to senior management. The unit shall be responsible for designing and implementing any internal model used for purposes of this Chapter. The unit shall conduct the initial and on-going validation of any internal model used for purposes of this Chapter, being responsible for the overall risk management system. The unit shall produce and analyse daily reports on the output of any internal model used for calculating capital requirements for position risk, foreign exchange risk and commodities risk, and on the appropriate measures to be taken in terms of trading limits;”
It is clear from this, that the validation is the responsibility of the Risk Control Unit.
Interpretation of the proposed RTS suggests that the functional reporting to the CRO would no longer be permissible. As such, an entirely separate line to the Management Body would be required, which appears both impractical and calls into question the implicit ability of the CRO to acknowledge and prioritise required changes effectively. Validation is a core element for a model risk management function which should report to the CRO.
We propose that an appropriate level of independence is one where the Independent Validation unit, which would be responsible for validating all models across the institution, including those used for IMA capital, should report directly to the CRO. This would be compliant with CRR Article 368.1(b) and ensure both appropriate expertise at the level of the Management Body and an appropriate level of independence with respect to models both within, and outside of, the Risk Control unit, developed by different units within Risk (for example Market Risk). In addition, for global systematically important institutions (GSIIs), senior management should be interpreted as the level below the management board which would comply with the desired independence as per Article 22.1(b) of the proposed RTS.
We agree with the principle that limits should be set and monitored, and breaches resolved through a formally documented process. We also agree with the proposals to document the cause and size of any limit breach and that usage and limit breaches should be considered when setting limits.
We do not agree, however, with some of the proposals within the CP, as a committee is unlikely to:
Meet on a daily basis in order to approve Traded market Risk limits where breaches occur;
Review and approve the full suite (often thousands) of limits across different measures and levels; and
Given all these concerns, we recommend that Articles 18, 21 and 25-28 are re-written to provide a pragmatic and workable proposal.
We would additionally like to highlight the following:
1. Requirement for a Committee to own limits and the day to day management of those limits. Whilst we agree that the top of the house VaR limit should be approved by a Committee designated by the Management Body, we do not believe that the Committee must enact the day-to-day approval of limit breaches, but should be able to delegate limit authority to an appropriate individual (e.g. Head of Market Risk) to grant temporary limits up to a documented amount.
2. Sub-limits owned by Management Body or designated committee
We understand the rationale for the same committee responsible for the Group VaR as having responsibility for the level directly below the Group VaR. We would appreciate clarification that this can be interpreted as either asset-class or business level. We recommend that the day-to-day management of this level of limits also has the ability to be delegated in a similar fashion to the Group level limit.
3. Jurisdiction level limits
We do not believe that it is appropriate to mandate limits to be set at the level of different jurisdictions and CAs or that these must follow the same governance structure as Group level limits.
Where material market risk exists, and local regulation requires, limits should be in place and the control around these limits comparable to the Group level. However, in many cases these limits may be owned by the local management body or delegate rather than being required to revert to the responsible Group committee. The magnitude of risk and number of entities that this may come from is likely to be inappropriate for ownership at the senior level of the institution in many cases. We instead recommend that legal entity or local jurisdiction limits should follow similar principles and the approach formally documented.
4. Approval of all limits by committee
Where limits of a more granular level are set, and non-VaR limits are set, we do not believe that it is appropriate to mandate a that committee approves these limits or that a business should be the proposer of these limits. We believe that it is reasonable for the CRO and Head of Market Risk to take responsibility and to delegate authority to independent Risk Managers (within the Risk Control Function), provided that these authorities are clearly documented and limit setting, tracking and breaches are recorded and tracked.
5. Consistency between VaR and other risk limits
Different limit types are used to measure different elements of risk (for example single name exposures measure concentrations, whereas VaR measures portfolio risk). If limit types were directly scaled without consideration for the risk each is trying to measure, there would be little need to establish a range of limits to measure different elements of the risk profile.
Requiring a link between limits of different types assumes that appetite to different risk events is demonstrable and consistent. Subjecting banks to a requirement of linking across numerous limits is likely to prove a significant bureaucratic overhead, with no clear benefits and which could have the adverse effect of reducing the number of limits an institution employs in response. As such we recommend removing Article 25.1(e) from the proposed RTS. 9
6. Authorisation of products by trader
Article 25.1(f) proposes that trader level authorities must be approved by the committee responsible for setting limits. We do not believe that these two controls (risk appetite and individual trader product mandates) need be considered by the same body, but that the establishment of Trading Mandates can be governed through alternative channels (e.g. alongside New Products and trading controls). We recommend removing this sub-article from the text.
Where a process for approving individual trades is clearly documented in policy and restrictions around size and number of trades are set, then we agree that it can be appropriate for approvals where certain new product approval (NPA) requirements are outstanding.
In line with our answer to Question 9, we do not believe that requiring committees as the method of governance for new products is appropriate.
Whilst many institutions may use a committee to perform governance, alternative process and designated departmental authorities can be equally valid as a method of managing and controlling the process. As per RTS Article 29, the new product policy would form the basis of requirements which need to be fulfilled.
We agree that the stability of the model is an important factor in the models suitability for risk management and capital calculations. However, we contend that a 250 day period is unnecessarily long when considering the overall model application process, the ongoing backtesting multiplier in the case of VaR and the length of time a new model may be under consideration by the competent authority.
We propose that a minimum of 90 days at the time of application would be appropriate in demonstrating stability of the model:
90 days is long enough to ensure that any positions the institution is likely to take would have passed through the model, not being missed through short term unusual periods of (in)activity;
A longer lead time to model application, jeopardises the ability to succeed in planning and roll-out of new models due to the extended lead time, with approvals potentially taking three-six months, an 18 month lag to implementation from completion of model implementation is unrealistic to bring a risk model into the capital framework;
A longer lead time increases the likelihood of further changes being required to a model in order to contend with new market conditions, products or regulations, without providing a great increase in significance of the stability demonstrated; and This shorter requirement does not prevent the competent authority requiring continued monitoring of the appropriateness of the model and the CA has the ability to provide an increase in the multiplier for a probationary period of time if stability remains a concern at the conclusion of the competent authorities review.
We agree that the results of the portfolios can be a useful indicator for the variability against industry results. However, given that results from previous benchmarking exercises have not provided definitive conclusions, but rather indicated particular areas which may require further investigation, and that the EBA has not yet completed an annual exercise mandated under Article 78 of Capital Requirements Directive (CRD), we do not believe that this should be a requirement for an IMA permission application.
We would also highlight that in order for a new model to provide results, it would have to be able to run the hypothetical portfolio exercise (HPE) at the point of publication. Given that the HPE is mandated to take place annually, we do not believe that this provides sufficient justification for delaying the consideration of an application for a new model permission.
We agree with the requirements to conduct stress tests on at least a monthly basis, and that ad hoc and hypothetical scenarios should be used as well as historical stress scenarios.
However, there are certain elements across Articles 32 and 33 of the RTS, which we believe are not appropriate to mandate through regulation. In particular we would like to highlight the following:
Article 32.2 states that stress scenarios must be used to assess the reasonableness of VaR. VaR is already subject to validation and review through P&L backtesting, hypothetical backtesting and, model validation. In addition, VaR looks at a particular confidence interval and time horizon. Stress tests & scenarios are frequently not calibrated to a given confidence interval, often moving only certain parameters and often hypothetical. As such they are unlikely to be useful in assessing the accuracy of the VaR model; and
We are concerned about both the requirement and potential read across in Article 33 paragraph 2(b)(ii) where institutions shall address event risk for equities and jump-to-default risk for credit positions by considering eight instantaneous defaults with zero recovery of the four specific interest rate risk long positions in the current portfolio with the largest exposure and the four largest equity long positions in the current portfolio. Alternatively, the event risk stemming from a sharp rise in equity prices should also be considered for the four largest short positions."
We do not agree with the definition of event risk as the simple sum of the largest four single names JTDzero for each category. If this were to be read across to the requirement under VaR (article 370.f of CRR), the implication for specific risk Pillar 1 capital would be far beyond an appropriate capitalisation requirement.
The potential effect in comparison with the incremental risk charge (IRC), where the four largest names with zero recovery could easily sum to a larger number than the IRC given large high grade sovereign holdings held by banks that either do not default or show significant expected recovery in the modelled approach and which under standardised approaches have a 0% weighting.
We propose that the term “event risk” is removed from this section, in order to remove potential confusion.
Assuming zero recovery and simultaneous defaults without consideration of rating is overly conservative. A more appropriate measure would be to align the number of defaults with rating grades and recovery to be below expected, rather than zero, when considering single name risks."
We have no further comments on this section.
We agree with the proposals set out in Article 36.1 and 36.2(a).
In order for VaR to work as a risk management tool, it is important that P&L and VaR are calculated wherever possible on the same market data. Article 36.3(b), by contrast appears to force a break with this principle, by mandating a single global closing time for individual risk factors in VaR.
For example, a USD swap being traded by desks in different time zones (e.g. Tokyo & New York) will have its P&L based on closing times relevant for the desks in question. As such, the same position can have a different closing price depending upon where it is traded. In order to ensure that the VaR follows this, therefore it can be necessary to have different time series for the USD Swap, appropriate for the time zone in question.
The alternatives suggested of either disallowing any consolidation for affected desks or calculating regulatory VaR without this time zone allocation are respectively capital punitive or operationally inefficient and delinked from risk management practice. The approach as detailed is not proportional in its requirements and there is no clear benefit from the proposal of enforcing a single close.
Please refer to our answer to question 19.
Mandating a daily update for market data would be inappropriate. The principles of minimum requirements around alternative modelling are appropriate to apply without consideration of the model employed. Whilst updating market data may be more straightforward for Historical Simulation models than alternatives, setting a higher threshold for those models and
considering the impact of this as being equivalent across different models, may not always be the case.
For example, a His Sim model with a five year window may well see a smaller impact on its VaR from large single day moves in very recent history than a model with a one year window, which are more reactive. As such we do not believe it is appropriate to differentiate the minimum market data update frequency based on model choice.
Market data volumes are significant and issues of stale, false or absent market data require careful control. In some instances, banks can perform this control post release of the market data to the VaR model. Whilst this allows for daily updates of market data, it increases the risk of incorrect data being used in the VaR, which can affect risk management of the portfolio risks. Alternatively, banks can release the market data periodically, which allows for more time to investigate data and a higher confidence around the accuracy of such data.
Whilst we agree with a minimum frequency of market data updates, we believe that banks are best placed to make the decision over the control environment that they apply to market data release, which will depend upon volume of and number of sources of data, VaR model and sensitivity of the model to single points of data.
We agree with Option 1 under Article 40.4 (c), that only changes in the risk categories for which permission has been granted, and therefore in scope of the VaR for the purposes of calculating capital, should be included in the hypothetical P&L.
Practically, however, we agree that including all risk categories P&L may be pragmatic for some firms given the requirement to strip out P&L from unapproved risk categories. Therefore, we recommend that this is an option granted at the discretion of CAs.
The EBA should opine on the applicability of disregarding backtesting exceptions from the addend calculation where factors driving an exception can be shown to come from risk factors capitalised elsewhere. An example of this is through the standardised approach where a risk category does not have IMA permission.
The appropriateness of the model is measured by its performance in measuring the risks that are within its scope. As such, it is a cleaner test of the model performance to only include the P&L that is in scope.
We believe that the same approach as for hypothetical P&L is appropriate (please see answer for question 22). There is no differentiation made in CRR Article 366.3 about the scope of the positions or risk categories covered by the backtesting of the portfolio under hypothetical and actual P&L. As such the proposed RTS Article 40.5 (c) should be amended or removed.
The current proposal to include all risk category P&Ls, even where only partial permission is granted, could result in significant numbers of backtesting exceptions, automatically driving the multiplier significantly higher. The result of this is that in such cases, additional IMA capital is held for Risk Categories which are capitalised under standard rules approaches and creating an uneven playing field for these risk factors when comparing to institutions with no IMA permission.
We appreciate that using actual P&L for assessing the backtesting multiplier is a pragmatic, rather than theoretical approach for capitalising intraday trading and new positions.
Whilst we would not be in favour of implementing new methodology to attempt to capture these effects within the IMA, we would point out that the current regulatory text and application by CAs would benefit from addition clarification around discounting exceptions from the multiplier addend.
The text needs to be more specific so that backtesting exceptions whilst initially counted towards the addend, should be discounted by CAs, where appropriate. Examples of this could be where a bank can evidence that additional, sufficient capital resources are held to account for a risk causing the breach or, in the case of actual P&L that the breach is caused by the exiting of positions as part of a structural business change (e.g. divesting non-core positions).
We agree that where valuation adjustments calculated less frequently than daily are included in the backtesting P&L, these should be eligible for discounting from the addend multiplier.
Rather than mandating the inclusion or exclusion of theta, we would recommend that theta should either be included in both VaR and backtesting P&L, or excluded from both.
We do not believe that stochastic correlations should be required for correlation risk arising within pricing models (i.e. from products which contain correlation risk such as options on multi asset baskets).
Given that implied correlations are frequently not observable in the market, capturing in VaR may be not appropriate, either as a risk factor which does not exhibit daily movements or challenging to model, which should be considered against whether the risk is material.
We also note that (in line with our answer to question 5), the presence of products that give rise to these risk factors (per Article 46.1(a)) of the proposed RTS) should not by definition make the risk factor “material”, but that the level of risk must be considered in this regard.
Where the risk is not material, or if the risks are not appropriate to be included in VaR, the regulation should not mandate inclusion in the VaR model, nor the use of stochastic correlations.
We agree with the first requirement to review the correlations with at least monthly frequency.
However we fail to see the benefit of requesting a specific test, e.g. using historically observed high/low correlations within the periodic validation process. A more generic test looking at the sensitivity of the VaR to changes in empirical correlations is deemed more appropriate in this context. It is also worth noting that the impact of correlations from a severely adverse period is already captured within the daily SVaR calculation.
We recommend that the language in Article 48.2 is changed as per the below, so that CAs are given appropriate flexibility to consider the potential impact of the elements under consideration and are not forced to increase the multiplier where this is not warranted:
“Competent authorities shall verify that the multiplication factors mc and ms reflects conservatively at least the following flaws and shortcomings of the VaR and Stressed VaR models related to the risk categories covered by the model scope of application:”
“Competent authorities should consider the following items in relation to VaR and Stressed VaR Models, related to the risk categories covered by the model scope of application in their assessment of the multiplication factors mc and ms:”
Whilst the principle of flooring the SVaR multiplier to the VaR multiplier is appropriate in most situations, we consider that there could be situations where a CA may wish to use a multiplier specifically in relation to market data (e.g. frequency of updates), which would not apply to the SVaR measure using a different historical period.
We therefore propose that if a CA chooses to enforce such a VaR only multiplier, it should be able to justify this choice appropriately.
We are supportive of Article 48.5 (a) which enables CAs to take into account any reserves held by an institution to mitigate any shortcomings of the IMA.
The demonstration that a model is conservative should be taken into account when the CA is reviewing the need for a multiplication factor in excess of the minimum specification.
In order to be transparent regarding the driver(s) of any additional multiplier above the minimum, we propose that the text should reflect a requirement for CAs to clearly define the quantity of any additional multiplier which is assigned to any identified shortcoming. In doing so, the capital impact of individual shortcomings will be transparent and subsequent remediation and removal of the multiplier clearly understood by both institution and CA.
The requirements of Article 43.3 are likely to present significant difficulty for compliance without adopting a full revaluation approach. In particular, the capture of all cross gammas (b) would require simultaneous perturbation for multiple risk factors for all but the simplest non-linear products. Point (c) also appears to formalise the need for full revaluation for products in this category. We also believe that as per our answer to question 5, the categorisation of products to determine approach is not proportionate.
It is our view that this requirement is going beyond what is required under the CRR which does not require full revaluation, but only capture of all material risk factors. As such we recommend removing section 3 from article 43.
Whilst we consider that a proxy validated for VaR should be acceptable for SVaR, we note that this should not be a mandatory condition to use the same proxy for the SVaR period, should a better alternative be available.
We agree that the majority of the proposals in the SVaR section are appropriate.
However, we do not agree with the section of Article 49.2, which links SVaR determination methodology to reverse stress testing in article 33.2 (a). SVaR methodology and reverse stress testing are two separate risk measurement approaches and requiring a link between the two has significant potential to push SVaR methodology away from comprehensive formulaic approaches.
We agree that including a position in the IRC sufficiently meets the requirements of Article 370 (f) of the CRR.
We note that the draft RTS, however, only notes compliance for debt instruments under Article 56.4. We recommend that this is extended to all instruments included in the IRC, to cover event risk where equity instruments are also included in the model.
We would appreciate clarification that whilst the proposed alternative route to compliance is for consideration, it should not be considered a mandatory approach for institutions. Whilst in principle we support the approach of allowing an alternative conservative approach to compliance, we believe that the approach suggested is overly conservative.
Rather than calculating a separate VaR based on the largest FX positions over the previous year, we believe that the best approximation of the current FX positions would be more representative of the risk and that the lack of diversification with the remaining portfolio then sufficiently provides for a conservative capital outcome.
We consider the requirements appropriate as they provide more guidance and clarifications on the respective modelling options.
We propose that banks should still have the option to decide whether they prioritise internal or external ratings for usage in their IRC calculation. An issue we see with the prioritisation of internal ratings, is that due to compliance restrictions the public side (e.g. business) cannot have full view of private side information like internal ratings.
This either limits the consistency between IRC internal ratings or limits the transparency of parameters of the IRC charge, which can reduce its application for risk management.
In our view PDs should be based on historical data and a direct conversion of credit spread into PDs. However, using spread-implied ratings which are derived by benchmarking an issue’s or issuers credit spread against spread benchmark curves by rating and category and then further mapping the spread-implied rating to a (historical) PD via the respective master scale, should be permitted.
Whilst we agree on the requirement to make the ratings from different rating sources comparable, we object to mandating the same rating-to-PD masterscale for all rating sources. This is because the underlying rating systems may be different (e.g. point-in-type vs through-the-cycle ratings). Consequently the PDs should align with the usage of internal or external ratings where banks should have the option to prioritise their usage themselves as stated in our answer to Q41.
We agree that a non-zero PD is a reasonable condition; however, we note that IRC models can generate non-zero charges through migration, even where the default state is assigned a zero PD.
We support that no explicit floor of 3bps is prescribed as stated in the FRTB rules for default risk charge (DRC). Such a floor would be highly punitive for high quality sovereigns where positions are held as liquidity reserve or as risk-free (or least risky) asset.
We agree that losses due to default should be based on the difference between current market value and the recovery value as this reflects the actual P&L event more accurately.
We would appreciate a confirmation from the EBA that our reading of Article 64.2.a is appropriate regarding our assumption that the risk control unit establishes a maximum size permitted for the individual positions with inferred ratings" requires size limits to be set on issuer level (as opposed to security or portfolio level)."