Response to consultation on Guidelines on limits on exposures to shadow banking
Go back
Quite apart from the lack of a mandate, we do not see any issues of substance or risk aspects that would support the need for separate Pillar II requirements relating to the setting of specific internal limits for shadow bank exposures. We are therefore highly critical of and reject the additional qualitative requirements explicitly for shadow banks. The qualitative requirements for Pillar II arising from CRD IV have already been comprehensively transposed into national law (in Germany, for example, through section 25a of the KWG in conjunction with the “MaRisk”). In addition to individual limits, these also include overarching, e.g. sectoral limits. Mitigating risks by using limits and other safeguards is thus already an established element of internal risk management at institutions and also covers risks arising from shadow bank exposures. We cannot understand why separate requirements and processes should now be stipulated explicitly for shadow banks in these Draft Guidelines when they must in any case be imposed for all kinds of borrowers. In addition to CRD IV, requirements for risk management by institutions are already anchored in a range of EBA guidelines (e.g. Internal Governance, SREP). The requirements set out in Title II, paragraphs 1 and 2 of the Draft Guidelines would cause unnecessary additional administrative effort, because separate frameworks, policies and reports would have to be developed explicitly for shadow banks, and these would then be subject to separate examination by supervisors and auditors. We cannot identify any corresponding benefits from this.
The limiting rules in existing limit systems under Pillar II are not based on the eligible capital referred to in the CP, but are based on the specific requirements for each institution derived from the borrower-related, sectoral and geographic risk diversification that is necessary or defined in the business policy, based on the credit portfolio model used. In addition, limits are not normally set at the level of the GCCs in Pillar II, but at the level of the individual borrower or counterparty. Overall, it can be said that there is no synchronisation with existing limit systems, which would ultimately lead to a further increase in cost and effort for implementation and the subsequent ongoing processes.
Moreover, the Draft Guidelines give the impression that shadow banks represent their own risk type. This is something we do not understand. As a matter of principle, each shadow bank inherently represents a borrower that can give rise to a range of risks for an institution (credit risk, market risk, operational risk, etc.) that are different for each shadow banking entity. We believe that it is overstepping the mark to generally assume a correlation of 1 and thus a high concentration risk for shadow banks. This would put shadow banks in a worse position per se than other borrowers, which would not be appropriate. It therefore also does not make any sense to require separate risk management mechanisms for shadow banks. In fact, this would not be possible, because the shadow banking sector is far too heterogeneous for it to be managed using a standardised approach.
Of course Pillar II requires institutions to identify, measure and manage credit risk concentrations. Applied to shadow banks, this means that – as set out in Article 81 of CRD IV – the concentration risk that arises from connections between a shadow bank and other borrowers (keyword: groups of connected clients) or from sectoral or geographic concentration must be managed appropriately. In any case, the CEBS Guidelines on the revised large exposures regime from December 2011 already stated that only idiosyncratic risk is analysed in the large exposures regime, whereas geographic and sectoral risk would be addressed under Pillar II. Our understanding is that the requirement to consider interconnectedness is therefore already satisfactorily met in the large exposures regime through the fundamental obligation to test for control or interconnectedness in accordance with Article 4(1) point 39(a) and (b) of the CRR.
The requirements governing the provision of information for setting limits for shadow banks set out in the Draft Guidelines are already very far-reaching, much too detailed and almost impossible to implement fully in practice. We believe that the proposals run the risk that the fallback approach would become the default because of the very detailed information requirements.
We have the following specific remarks on paragraph 1 of Title II of the Draft Guidelines:
Point a) does not make clear which exposure is meant with regard to individual borrowers or the GCCs, and whether the look-through requirements should be considered or not. The definition of the exposure in the Draft Guidelines is not sufficiently precise (see also our specific remarks on Q1 relating to groups of connected clients and look-through requirements). In addition, it should be sufficient to identify all material potential risks. The word “material” should therefore be inserted in front of “potential risks”.
In point b), we are concerned about the requirement to involve the credit risk committee in each decision, as this would undermine existing credit approval arrangements. We believe that it is not necessary to involve the CRC in the case of minor exposures.
In point e), it is doubtful that the requirement is practicable, as we believe that it will not be possible to obtain complete transparency about the interconnectedness between shadow banks, which will also change over time. In consequence, the process of assessing/reflecting risks will also remain unclear. In this case, too, the principle of materiality should play a role and a corresponding materiality threshold should be specified (e.g. similar to the increase proposed by the Basel Committee in its April 2014 large exposures framework in the limit for an in-depth test of interconnectedness to 5% of the aggregate risk exposures to a shadow bank); on the topic of interconnectedness, please also refer to our previous remarks.
In any case, individual limits are set as part of the regular credit processes or can be derived from the criteria defined in the credit risk strategy.
We reject a prudential requirement for an aggregate limit for shadow banks – regardless of whether this is 25% or another defined limit – by the EBA as part of the Pillar II process, as provided for under the fallback approach. In line with the principle of proportionality, and in exercise of their responsibility as managers, aggregate / sectoral limits should be set – where this is sensible and necessary – by the institutions individually to reflect their business model, their risk appetite and the materiality of the exposures. The necessary limits thus depend significantly on the business model and cannot therefore be specified globally. It is then a matter for the competent authorities to examine the appropriateness of the limits set by the institutions in the course of their supervisory activities.
For example, groups of institutions that themselves establish funds have larger volumes of AIFs – because of start-up finance, etc. – than institutions with another business model. On the other hand, small and medium sized institutions with a business model focussed on regional lending use investments in funds or securitisations for risk diversification purposes.
As an alternative, we are proposing to at least examine the possibility of setting lower individual and/or aggregate limits for shadow banks in the large exposures regime. This would mean that individual limits under Pillar II due purely to the classification of the borrower as a shadow bank would be unnecessary. In our view, calibrating an aggregate large exposure limit for shadow banks depends crucially on the definition of what is a shadow bank, and should under no circumstances be set at less than 800% (calibration subject to the QIS and depending on the definition of a shadow bank) of eligible capital so as to reflect the reservations expressed in these comments, in particular as regards the inclusion of all AIFs and MMFs without exception.
Alternatively, a lower individual large exposure limit for shadow banks could be set on condition that the definition of a shadow bank is modified to reflect our reservations above.
We wish to make the following additional remarks on paragraph 4 of Title II of the Draft Guidelines:
The scope of the information to be collected appears to be very substantial – especially if the EBA wishes to stick to its proposed materiality limit of 0.25% of eligible capital. It will therefore be difficult to gather all the information required to set individual limits for exposures to shadow banking activities, and some of an institution’s counterparties might not be comprehensively assessed in accordance with all of the requirements illustrated in the CP.
We believe that points c), d), e) and f) are not practicable because a complete look-through does not appear to be possible, at least to the extent necessary, and the bank would have to rely solely on the assessment by the shadow bank itself. It should be clarified at least in respect of paragraph 4 of Title II of the Draft Guidelines that the requirements should be interpreted such that the information to be provided by a shadow bank is sufficient, and that the institution is not required to obtain information over and above this, or to verify the information that has been provided to it.
If an institution is unable to meet the requirements, the EBA is proposing an aggregate limit of 25% of eligible capital for all exposures to shadow banks (Option 1 preferred by the EBA). We cannot understand what motivates the EBA to already favour Option 1 at this stage. Of course, this approach is the most conservative of all the options.
If the EBA wishes to stick to its plan to impose an aggregate prudential limit for the fallback approach – despite the serious concerns expressed above – the proposed 25% of eligible capital is far too conservative and would have an impact on the business and investment policies of the institutions that is extremely difficult to estimate. The reason for this is that, at present, up to 25% of the institution’s eligible capital can be lent in principle to each shadow bank. Option 1 would considerably exaggerate the risks arising from lending to shadow banks. It therefore has the effect of sharply limiting investments by banks and, in view of the broad definition of shadow banks that underlies the Draft Guidelines, it would effectively restrict transactions in funds, certifications, etc., to an extent that goes far beyond the current large exposures regime. We also believe it would negatively impact portfolio diversification. In our opinion, this would represent an economically unjustified restriction on the institutions’ business lines that would be affected, it would contradict the economic policy objective of encouraging the securitisation market, and would significantly exceed the EBA’s mandate under Article 395(2) of the CRR.
A QIS would have to be performed before such an aggregate limit is set, and the alternative options listed in the CRR (a lower large exposure limit for individual exposures or an aggregate limit for large exposures to shadow banks) would have to be examined.
If, additionally, this requirement were to come into force without a suitable grandfathering arrangement, the institutions would be forced to terminate some of their current exposures before the agreed terms, with unforeseeable consequences for the markets.
Because institutions have already implemented comprehensive analysis and control processes for their significant investments in funds and securitisations, Option 2 of the fallback approach would have considerably less restrictive consequences for the markets.
As a result, Option 2 not only rewards institutions with a potentially higher exposure limit for knowing their counterparties, but in fact more adequately reflects the reality of their risk profile. Hence, Option 2 could potentially provide more incentives to gather information about exposures than Option 1. However, Option 2 will only be attractive for those institutions (1) whose exposures to shadow banking entities currently amount to more than 25% of their eligible capital, and (2) if an exposure assessment according to the requirements of the principal approach results in higher individual and aggregate limits than the proposed 25% in Option 1 of the fallback approach. At first sight, these potentially higher limits if Option 2 is applied are apparently less conservative than the defined 25% limit in Option 1. However, as pointed out above, they reflect the real risk profile of exposures to shadow banking entities better than a fixed percentage, and therefore represent a more prudent approach that should be in the EBA’s interest.
In light of the high level of requirements associated with the principal approach, institutions (particularly smaller and medium-sized institutions) should be free to choose the fallback approach. We do not share the concern that giving institutions such a choice would lead to regulatory arbitrage (see paragraph 29).
In this context, we again call for an EU-wide QIS in particular for the assessment of questions 5 and 6.
In our view, the proposed 25% aggregate internal limit is therefore not consistent with the large exposures regime, because the limit of 25% of eligible capital in the large exposure regime always only refers to a limit on the risk concentration in respect of an individual borrower or a group of connected clients. However, shadow banks as a whole do not represent a group of connected clients because of control or interconnectedness. Consequently, 25% is far too low for an aggregate limit. We wish to recall at this point that CRD II specified a limit of 800% of own funds for all of an institution’s large exposures. In this case too, the limit referred only to those exposures that exceeded the 10% large exposure limit.
Moreover, the fallback approach would probably have to be used in particular by small and medium-sized regional institutions because it is more likely than not that the substantial process requirements of the principal approach cannot be met by banks of this size. From the viewpoint of these regional institutions, investments in funds, including for example in real estate funds with a national scope, help diversifying their portfolio, which is a positive factor from a risk perspective. Limiting these investments, together with the investments in other shadow banks, to an aggregate of 25% of eligible capital, would unreasonably disadvantage small and medium-sized institutions, including in terms of risk diversification.
2. Do you agree with the approach the EBA has proposed for the purposes of establishing effective processes and control mechanisms? If not, please explain why and present possible alternatives.
We wish to stress again at this point our view that the requirements contained in the Draft Guidelines relating to internal risk management processes are not covered by the mandate in Article 395(2) of the CRR. The objective of Article 395(2) of the CRR is clearly to mitigate overarching/systemic risks from interconnectedness between banks and shadow banks by means of suitable additional large exposure limits, not to expand the Pillar II requirements.Quite apart from the lack of a mandate, we do not see any issues of substance or risk aspects that would support the need for separate Pillar II requirements relating to the setting of specific internal limits for shadow bank exposures. We are therefore highly critical of and reject the additional qualitative requirements explicitly for shadow banks. The qualitative requirements for Pillar II arising from CRD IV have already been comprehensively transposed into national law (in Germany, for example, through section 25a of the KWG in conjunction with the “MaRisk”). In addition to individual limits, these also include overarching, e.g. sectoral limits. Mitigating risks by using limits and other safeguards is thus already an established element of internal risk management at institutions and also covers risks arising from shadow bank exposures. We cannot understand why separate requirements and processes should now be stipulated explicitly for shadow banks in these Draft Guidelines when they must in any case be imposed for all kinds of borrowers. In addition to CRD IV, requirements for risk management by institutions are already anchored in a range of EBA guidelines (e.g. Internal Governance, SREP). The requirements set out in Title II, paragraphs 1 and 2 of the Draft Guidelines would cause unnecessary additional administrative effort, because separate frameworks, policies and reports would have to be developed explicitly for shadow banks, and these would then be subject to separate examination by supervisors and auditors. We cannot identify any corresponding benefits from this.
The limiting rules in existing limit systems under Pillar II are not based on the eligible capital referred to in the CP, but are based on the specific requirements for each institution derived from the borrower-related, sectoral and geographic risk diversification that is necessary or defined in the business policy, based on the credit portfolio model used. In addition, limits are not normally set at the level of the GCCs in Pillar II, but at the level of the individual borrower or counterparty. Overall, it can be said that there is no synchronisation with existing limit systems, which would ultimately lead to a further increase in cost and effort for implementation and the subsequent ongoing processes.
Moreover, the Draft Guidelines give the impression that shadow banks represent their own risk type. This is something we do not understand. As a matter of principle, each shadow bank inherently represents a borrower that can give rise to a range of risks for an institution (credit risk, market risk, operational risk, etc.) that are different for each shadow banking entity. We believe that it is overstepping the mark to generally assume a correlation of 1 and thus a high concentration risk for shadow banks. This would put shadow banks in a worse position per se than other borrowers, which would not be appropriate. It therefore also does not make any sense to require separate risk management mechanisms for shadow banks. In fact, this would not be possible, because the shadow banking sector is far too heterogeneous for it to be managed using a standardised approach.
Of course Pillar II requires institutions to identify, measure and manage credit risk concentrations. Applied to shadow banks, this means that – as set out in Article 81 of CRD IV – the concentration risk that arises from connections between a shadow bank and other borrowers (keyword: groups of connected clients) or from sectoral or geographic concentration must be managed appropriately. In any case, the CEBS Guidelines on the revised large exposures regime from December 2011 already stated that only idiosyncratic risk is analysed in the large exposures regime, whereas geographic and sectoral risk would be addressed under Pillar II. Our understanding is that the requirement to consider interconnectedness is therefore already satisfactorily met in the large exposures regime through the fundamental obligation to test for control or interconnectedness in accordance with Article 4(1) point 39(a) and (b) of the CRR.
The requirements governing the provision of information for setting limits for shadow banks set out in the Draft Guidelines are already very far-reaching, much too detailed and almost impossible to implement fully in practice. We believe that the proposals run the risk that the fallback approach would become the default because of the very detailed information requirements.
We have the following specific remarks on paragraph 1 of Title II of the Draft Guidelines:
Point a) does not make clear which exposure is meant with regard to individual borrowers or the GCCs, and whether the look-through requirements should be considered or not. The definition of the exposure in the Draft Guidelines is not sufficiently precise (see also our specific remarks on Q1 relating to groups of connected clients and look-through requirements). In addition, it should be sufficient to identify all material potential risks. The word “material” should therefore be inserted in front of “potential risks”.
In point b), we are concerned about the requirement to involve the credit risk committee in each decision, as this would undermine existing credit approval arrangements. We believe that it is not necessary to involve the CRC in the case of minor exposures.
In point e), it is doubtful that the requirement is practicable, as we believe that it will not be possible to obtain complete transparency about the interconnectedness between shadow banks, which will also change over time. In consequence, the process of assessing/reflecting risks will also remain unclear. In this case, too, the principle of materiality should play a role and a corresponding materiality threshold should be specified (e.g. similar to the increase proposed by the Basel Committee in its April 2014 large exposures framework in the limit for an in-depth test of interconnectedness to 5% of the aggregate risk exposures to a shadow bank); on the topic of interconnectedness, please also refer to our previous remarks.
3. Do you agree with the approach the EBA has proposed for the purposes of establishing appropriate oversight arrangements? If not, please explain why and present possible alternatives.
We refer here to our fundamental comments on Q2, in which we argue that separate requirements for exposures to shadow banks are not needed from a risk perspective for either internal risk management or the governance of the institutions. In addition, imposing such requirements is not covered by the mandate under Article 395(2) of the CRR.4.Do you agree with the approaches the EBA has proposed for the purposes of establishing aggregate and individual limits? If not, please explain why and present possible alternatives.
As we already explained in our specific remarks on Q2, we believe that setting separate limits for shadow banks under Pillar II does not make any sense, because the shadow banking sector is very heterogeneous and setting separate limits would not generate adequate management triggers.In any case, individual limits are set as part of the regular credit processes or can be derived from the criteria defined in the credit risk strategy.
We reject a prudential requirement for an aggregate limit for shadow banks – regardless of whether this is 25% or another defined limit – by the EBA as part of the Pillar II process, as provided for under the fallback approach. In line with the principle of proportionality, and in exercise of their responsibility as managers, aggregate / sectoral limits should be set – where this is sensible and necessary – by the institutions individually to reflect their business model, their risk appetite and the materiality of the exposures. The necessary limits thus depend significantly on the business model and cannot therefore be specified globally. It is then a matter for the competent authorities to examine the appropriateness of the limits set by the institutions in the course of their supervisory activities.
For example, groups of institutions that themselves establish funds have larger volumes of AIFs – because of start-up finance, etc. – than institutions with another business model. On the other hand, small and medium sized institutions with a business model focussed on regional lending use investments in funds or securitisations for risk diversification purposes.
As an alternative, we are proposing to at least examine the possibility of setting lower individual and/or aggregate limits for shadow banks in the large exposures regime. This would mean that individual limits under Pillar II due purely to the classification of the borrower as a shadow bank would be unnecessary. In our view, calibrating an aggregate large exposure limit for shadow banks depends crucially on the definition of what is a shadow bank, and should under no circumstances be set at less than 800% (calibration subject to the QIS and depending on the definition of a shadow bank) of eligible capital so as to reflect the reservations expressed in these comments, in particular as regards the inclusion of all AIFs and MMFs without exception.
Alternatively, a lower individual large exposure limit for shadow banks could be set on condition that the definition of a shadow bank is modified to reflect our reservations above.
We wish to make the following additional remarks on paragraph 4 of Title II of the Draft Guidelines:
The scope of the information to be collected appears to be very substantial – especially if the EBA wishes to stick to its proposed materiality limit of 0.25% of eligible capital. It will therefore be difficult to gather all the information required to set individual limits for exposures to shadow banking activities, and some of an institution’s counterparties might not be comprehensively assessed in accordance with all of the requirements illustrated in the CP.
We believe that points c), d), e) and f) are not practicable because a complete look-through does not appear to be possible, at least to the extent necessary, and the bank would have to rely solely on the assessment by the shadow bank itself. It should be clarified at least in respect of paragraph 4 of Title II of the Draft Guidelines that the requirements should be interpreted such that the information to be provided by a shadow bank is sufficient, and that the institution is not required to obtain information over and above this, or to verify the information that has been provided to it.
5. Do you agree with the fallback approach the EBA has proposed, including the cases in whichit should apply? If not, please explain why and present possible alternatives. Do you think that Option 2 is preferable to Option 1 for the fallback approach? If so, why? In particular: Do you believe that Option 2 provides more incentives to gather information about exposures than Option 1? Do you believe that Option 2 can be more conservative than Option 1? If so, when? Do you see some practical
It is understandable that the EBA wishes to create incentives to collect as much and as comprehensive information as possible about shadow banks. However, this masks the aspect of materiality, which is a part of every lending decision. In our view, there is no need for any “technical” fallback approach because any deficiencies in setting internal limits – even if they relate to shadow banks – would be addressed as part of the SREP and could be sanctioned by additional capital requirements.If an institution is unable to meet the requirements, the EBA is proposing an aggregate limit of 25% of eligible capital for all exposures to shadow banks (Option 1 preferred by the EBA). We cannot understand what motivates the EBA to already favour Option 1 at this stage. Of course, this approach is the most conservative of all the options.
If the EBA wishes to stick to its plan to impose an aggregate prudential limit for the fallback approach – despite the serious concerns expressed above – the proposed 25% of eligible capital is far too conservative and would have an impact on the business and investment policies of the institutions that is extremely difficult to estimate. The reason for this is that, at present, up to 25% of the institution’s eligible capital can be lent in principle to each shadow bank. Option 1 would considerably exaggerate the risks arising from lending to shadow banks. It therefore has the effect of sharply limiting investments by banks and, in view of the broad definition of shadow banks that underlies the Draft Guidelines, it would effectively restrict transactions in funds, certifications, etc., to an extent that goes far beyond the current large exposures regime. We also believe it would negatively impact portfolio diversification. In our opinion, this would represent an economically unjustified restriction on the institutions’ business lines that would be affected, it would contradict the economic policy objective of encouraging the securitisation market, and would significantly exceed the EBA’s mandate under Article 395(2) of the CRR.
A QIS would have to be performed before such an aggregate limit is set, and the alternative options listed in the CRR (a lower large exposure limit for individual exposures or an aggregate limit for large exposures to shadow banks) would have to be examined.
If, additionally, this requirement were to come into force without a suitable grandfathering arrangement, the institutions would be forced to terminate some of their current exposures before the agreed terms, with unforeseeable consequences for the markets.
Because institutions have already implemented comprehensive analysis and control processes for their significant investments in funds and securitisations, Option 2 of the fallback approach would have considerably less restrictive consequences for the markets.
As a result, Option 2 not only rewards institutions with a potentially higher exposure limit for knowing their counterparties, but in fact more adequately reflects the reality of their risk profile. Hence, Option 2 could potentially provide more incentives to gather information about exposures than Option 1. However, Option 2 will only be attractive for those institutions (1) whose exposures to shadow banking entities currently amount to more than 25% of their eligible capital, and (2) if an exposure assessment according to the requirements of the principal approach results in higher individual and aggregate limits than the proposed 25% in Option 1 of the fallback approach. At first sight, these potentially higher limits if Option 2 is applied are apparently less conservative than the defined 25% limit in Option 1. However, as pointed out above, they reflect the real risk profile of exposures to shadow banking entities better than a fixed percentage, and therefore represent a more prudent approach that should be in the EBA’s interest.
In light of the high level of requirements associated with the principal approach, institutions (particularly smaller and medium-sized institutions) should be free to choose the fallback approach. We do not share the concern that giving institutions such a choice would lead to regulatory arbitrage (see paragraph 29).
In this context, we again call for an EU-wide QIS in particular for the assessment of questions 5 and 6.
6. Taking into account, in particular, the fact that the 25% limit is consistent with the currentlimit in the large exposures framework, do you agree it is an adequate limit for the fallback approach? If not, why? What would the impact of such a limit be in the case of Option 1? And in the case of Option 2?
We wish to refer to our remarks above on question 5. In our opinion, a purely static aggregate internal imit of 25% for exposures to shadow banking entities in relation to an institution’s eligible capital does not make allowance for any institution-specific characteristics regarding its risk management of exposures to shadow banking entities. Additionally, we wish to comment that the assumption of interconnectedness (correlation of 1) for all shadow banks implied by the application of a 25% limit to all shadow banks (Option 1) is unrealistic in our opinion and is evidently not shared by the EBA, which also expects interconnectedness between shadow banks to be examined when individual limits are set.In our view, the proposed 25% aggregate internal limit is therefore not consistent with the large exposures regime, because the limit of 25% of eligible capital in the large exposure regime always only refers to a limit on the risk concentration in respect of an individual borrower or a group of connected clients. However, shadow banks as a whole do not represent a group of connected clients because of control or interconnectedness. Consequently, 25% is far too low for an aggregate limit. We wish to recall at this point that CRD II specified a limit of 800% of own funds for all of an institution’s large exposures. In this case too, the limit referred only to those exposures that exceeded the 10% large exposure limit.
Moreover, the fallback approach would probably have to be used in particular by small and medium-sized regional institutions because it is more likely than not that the substantial process requirements of the principal approach cannot be met by banks of this size. From the viewpoint of these regional institutions, investments in funds, including for example in real estate funds with a national scope, help diversifying their portfolio, which is a positive factor from a risk perspective. Limiting these investments, together with the investments in other shadow banks, to an aggregate of 25% of eligible capital, would unreasonably disadvantage small and medium-sized institutions, including in terms of risk diversification.