Response to consultation on draft technical standards in relation to credit valuation adjustment risk
Go back
The industry view is that proxying of the spread of the subsidiary by the spread of the parent company is a reasonable approach in most cases, to the exception of certain holding companies. In the case of holding companies, subsidiaries often have very different characteristics and risk profiles from the parent company, a more granular assessment of the subsidiary’s rating would therefore be appropriate.
Differences in characteristics (e.g. rating) between the two counterparties may be reflected by applying a multiplier. In situations where ratings are not available, a case-by-case review and computation of the multiplier should be used. We believe that the EBA RTS should consider such an approach, subject to regulatory approval, as this could produce a more accurate proxy credit spread than using a generic spread based on rating, industry and region.
Although the majority of proxied counterparties are generally mapped to credit spreads derived using the rating, industry and region attributes, some proxied counterparties may instead be mapped to other single name counterparties (or combinations of single name counterparties) with which they share similar characteristics. We believe that this may produce a closer match to the credit spread which the counterparty would trade in the market than would be possible using generic spreads. The examples, proposed in paragraphs 3 and 4, are broadly used indeed. Additionally, proxying a counterparty with a proven interest may be considered, subject to regulatory approval: e.g. if company A is linked to company B via an explicit guarantee.
Furthermore, although we understand the rationale for imposing maximum materiality (i.e. size) thresholds for non-IMM trades, we do not believe that a threshold on the number of transactions is needed. The capital charge that firms calculate is not driven by the number of transactions but rather by the size and type of transactions. Firms may have large volumes of small transactions in non-IMM portfolios that attract small capital charges. Firms should not be penalized by having to adopt the standard CVA approach for portfolios that do not attract large capital charges. We also believe that basing the threshold logic on EADs would be a sensible approach; however the exclusion of collateral from the calculation seems contradictory to the wider calculation of exposure for capital requirements purposes. If collateral is applied primarily against non-IMM exposures then collateral should be considered as part of the threshold calculation as it is a valid mechanism of exposure reduction.
We do not expect the thresholds to have a material impact on larger institutions.
In addition, we would like to clarify:
- the draft RTS require firms to calculate and report the arithmetic average of at least monthly observations of the ratio of ‘the individual size of each non-IMM portfolio to the total size of all portfolios’ to the local regulator. Does this mean that the arithmetic average for every non-IMM portfolio needs to be reported to the local regulator (this could result in thousands of lines of data from each firm)?
- Cost: as stated in the response to previous RTS, new proxy logic leads to a misalignment between the accounting CVA and the regulatory CVA frameworks. This inconsistency could force firms to first build a strategic infrastructure to address new requirements and afterwards create new operational units having the sole function of dealing with regulatory CVA in addition to the existing units. This would lead to significant incremental costs to the industry for no risk management benefit.
- Benefit: although we do agree that imposing harmonised criteria for the calculation of the CVA risk could support creation of a level-playing field, our view is that the reasonability and the easiness of implementation of those criteria that is vital. The current proposal, as per last formulation, would not necessarily help in achieving this.
Additionally, should the RTS call into question the way credit spread shocks are determined within internal VaR models, then the impact in terms of implementation cost and risk analysis would be drastic. Concretely, a large majority of banks would face the overwhelming challenge to revisit in depth their credit VaR models within a very tight timeframe (rules become live on January 1st, 2014) without prior assessment of the impacts both in terms of capital charge and risk management. VaR indicators are a corner stone of use test requirements and imposing banks to comply with prescriptive rules that come at odds with the way positions are monitored and risk managed will undoubtedly severely undermine the necessary consistency between capital requirement and risk management.
Finally, the cost/benefit assessment currently ignores the burden to switch between the advanced CVA charge and the standard CVA charge as a consequence of the requirement introduced in CRR Article 383(6) to fallback to standard method in case proxy spreads are deemed not compliant. Indeed, in major institutions, standardized methods are under the Finance function responsibility while advanced methods are under the Risk function responsibility. As a result, switching from a method to the other will be burdensome in terms of workflow and aggregation of results.
Please provide information and data concerning the availability of CDS data with respect to the minimum categories for ‘rating’, ‘industry’ and ‘region’ defined in points (b), (c) and (d).
Please see the attached document for a full industry response with relevant background information.Paragraph 3 allows for the proxying of the spread of the subsidiary by the spread of the parent company. Where no rating is available for the subsidiary or the parent undertaking or both, should the entities be considered equal in terms of the ratings attribute? Do you think that this treatment is appropriate? Please state the reason(s) in favour and/or against.
Please see the attached document for a full industry response with background information. The below extract is a summary of the response.The industry view is that proxying of the spread of the subsidiary by the spread of the parent company is a reasonable approach in most cases, to the exception of certain holding companies. In the case of holding companies, subsidiaries often have very different characteristics and risk profiles from the parent company, a more granular assessment of the subsidiary’s rating would therefore be appropriate.
Differences in characteristics (e.g. rating) between the two counterparties may be reflected by applying a multiplier. In situations where ratings are not available, a case-by-case review and computation of the multiplier should be used. We believe that the EBA RTS should consider such an approach, subject to regulatory approval, as this could produce a more accurate proxy credit spread than using a generic spread based on rating, industry and region.
Please indicate other particular cases in which single named proxies might be appropriate.
Please see the attached document for a full industry response with background information.Although the majority of proxied counterparties are generally mapped to credit spreads derived using the rating, industry and region attributes, some proxied counterparties may instead be mapped to other single name counterparties (or combinations of single name counterparties) with which they share similar characteristics. We believe that this may produce a closer match to the credit spread which the counterparty would trade in the market than would be possible using generic spreads. The examples, proposed in paragraphs 3 and 4, are broadly used indeed. Additionally, proxying a counterparty with a proven interest may be considered, subject to regulatory approval: e.g. if company A is linked to company B via an explicit guarantee.
Do the proposed thresholds of [15] % for the number of non-IMM portfolios, of [1] % for each individual non-IMM portfolio, and [10] % for the total size non-IMM portfolios, together with the definitions, provide an incentive for institutions to limit their portfolio exposures not covered by the IMM? Will the defined thresholds of [15] %, [1] % and [10] % cause any impact for your institution?
The industry does not believe that the size of the thresholds will be a material consideration when firms assess whether portfolios should be covered by the IMM or standardized approach. The standardized approach is less sophisticated and risk-sensitive than the IMM approach, which means that it typically delivers much more conservative capital requirements. This provides an incentive for firms to use the IMM approach wherever possible. However, the main reasons that some firms use the IMM for the majority of their exposures and the standardized approach for a subset of portfolios, is driven by systems, modeling and data constraints. These issues may mean that certain portfolios remain on the standardized approach, even if there is a capital incentive to move them to the IMM approach.Furthermore, although we understand the rationale for imposing maximum materiality (i.e. size) thresholds for non-IMM trades, we do not believe that a threshold on the number of transactions is needed. The capital charge that firms calculate is not driven by the number of transactions but rather by the size and type of transactions. Firms may have large volumes of small transactions in non-IMM portfolios that attract small capital charges. Firms should not be penalized by having to adopt the standard CVA approach for portfolios that do not attract large capital charges. We also believe that basing the threshold logic on EADs would be a sensible approach; however the exclusion of collateral from the calculation seems contradictory to the wider calculation of exposure for capital requirements purposes. If collateral is applied primarily against non-IMM exposures then collateral should be considered as part of the threshold calculation as it is a valid mechanism of exposure reduction.
We do not expect the thresholds to have a material impact on larger institutions.
In addition, we would like to clarify:
- the draft RTS require firms to calculate and report the arithmetic average of at least monthly observations of the ratio of ‘the individual size of each non-IMM portfolio to the total size of all portfolios’ to the local regulator. Does this mean that the arithmetic average for every non-IMM portfolio needs to be reported to the local regulator (this could result in thousands of lines of data from each firm)?
The EBA expects that only a limited number of counterparties/names will receive a proxy spread. Do you agree with this conclusion? If not, could you explain why and state how many of your names will require a proxy spread?
No. Our analysis indicates that due to the current slowdown in activity on the OTC market, there will be a significantly larger number of counterparties with no observable credit spread (and thus requiring a proxy spread) than there will be counterparties which would benefit from receiving a direct mapping. This is particularly true for smaller and/or non-financial counterparties, as well as for private banking clients. For larger institutions, more than two thirds of the names will require a proxy spread.Do you agree with the above analysis of the costs and benefits of the proposals? If not, please provide any evidence or data that would further inform the analysis of the likely cost and benefit impacts of the proposals.
We believe that the cost-benefit ratio for this particular change is disproportional. The main reasons for this are:- Cost: as stated in the response to previous RTS, new proxy logic leads to a misalignment between the accounting CVA and the regulatory CVA frameworks. This inconsistency could force firms to first build a strategic infrastructure to address new requirements and afterwards create new operational units having the sole function of dealing with regulatory CVA in addition to the existing units. This would lead to significant incremental costs to the industry for no risk management benefit.
- Benefit: although we do agree that imposing harmonised criteria for the calculation of the CVA risk could support creation of a level-playing field, our view is that the reasonability and the easiness of implementation of those criteria that is vital. The current proposal, as per last formulation, would not necessarily help in achieving this.
Additionally, should the RTS call into question the way credit spread shocks are determined within internal VaR models, then the impact in terms of implementation cost and risk analysis would be drastic. Concretely, a large majority of banks would face the overwhelming challenge to revisit in depth their credit VaR models within a very tight timeframe (rules become live on January 1st, 2014) without prior assessment of the impacts both in terms of capital charge and risk management. VaR indicators are a corner stone of use test requirements and imposing banks to comply with prescriptive rules that come at odds with the way positions are monitored and risk managed will undoubtedly severely undermine the necessary consistency between capital requirement and risk management.
Finally, the cost/benefit assessment currently ignores the burden to switch between the advanced CVA charge and the standard CVA charge as a consequence of the requirement introduced in CRR Article 383(6) to fallback to standard method in case proxy spreads are deemed not compliant. Indeed, in major institutions, standardized methods are under the Finance function responsibility while advanced methods are under the Risk function responsibility. As a result, switching from a method to the other will be burdensome in terms of workflow and aggregation of results.