Both Article 2 GEN (addressing the exceptions from mandatory margining requirements or possibility to agree on thresholds) and Article 1 FP (3) (with the provisions on the phase-in and the threshold based exemption from the initial margin requirement) require a formal/written opt-out by way of contractual agreement in order to allow counterparties to benefit from exceptions (positive election requirement). This approach is unnecessary cumbersome and formalistic and would impose unreasonable burdens on market participants, in particular, on counterparties which are not subject to the clearing obligation, a great many of which will be small and medium sized corporates (which were always intended to be exempted from the requirements under Article 11 (3) EMIR). Credit institutions would be required to confront all customers engaged in derivative transactions with a demand to enter into such a formal opt-out agreement. As many customers will have relations with more than one bank, they will have to deal with different versions of such arrangements. The process to negotiate and enter into opt-out agreements with each customer will be extremely time consuming, bind resources to a very considerable degree and, of course, be very costly for both sides. In order to simplify the operational challenges for all market participants but in particular for smaller and medium sized counterparties, the exemptions should therefore be designed as directly applicable exemptions (not requiring an opt-out) paired with duty on counterparties to inform the other counterparties of their status or relevant factors (and any subsequent changes) as well as a requirement to obtain the relevant declarations/representations from counterparties.
There will be additional administrative costs due to a significant increase of agreements to be established with counterparties trading OTC derivative contracts and necessity to allocate resources to operate margin transfers on daily basis even for small FCs end-users. This may have impact on liquidity and use of capital. As a result, small end-users will have to rethink the way they use derivatives, including for hedging. Refraining from hedging risks would be an unintended consequence and contrary to the objective of financial stability. The increase in administrative and operational costs is a factor that may stop small / medium sized entities hedging their interest rate risk, thereby leaving more rather than less risk in the system. These costs are hard to quantify but the requirements will result in higher operational and funding costs. Promotion of standardised third party solutions may reduce the operational burden/IT development costs.
The need for various agreements for the purpose of not exchanging initial and variation margin will pose a significant operational burden. This will be particularly significant with respect to the relationship to small and medium sized entities.
Equity funds will have to reach for collateral transformation services (which represents costs) to meet their margining requirements, to the extent that they do not have eligible collateral assets in their portfolio or cash available.
The obligation to collect margin from both EU and third country entities may have negative effects. From both a business and regulatory sense we question why parties not governed by EMIR should effectively be required to post IM (and also be forced to disclose the information required to determine which category they fall in or whether thresholds have been breached), thereby giving EMIR an extraterritorial reach. It is difficult to see why any third country counterparty should be willing to post IM and divulge this information when it is not subject to similar regulatory obligations in their home jurisdiction. Rather, it is more likely that these counterparties will simply choose not to trade with EU entities, leading to further market fragmentation.
Additionally, small non-EU funds may be required to collect as well as post margin under the RTS as they will be deemed an FC by virtue of their manager being authorised under AIFMD. Many of these small entities with not have the legal or operational capacity to be able to do this and thus would be detrimental to their operations if this was mandated.
In the first instance it would be reasonable to only require IM to be exchanged between two parties that both are required to post IM under EMIR (namely, only FC, NFC- and NFC+ or excluding any non-EU entity). On the cross-border/extraterritorial issues we would suggest regulatory bodies work towards consistency and alignment on their respective third country rules, in terms of threshold, timing, entities covered, exemptions, etc. It should be of great concern if an EU entity requires a non-EU entity to post margin (which in all likelihood would mean that the EU entity would have to post IM in return), when there is not the equivalent regime in the country of the non-EU entity, requiring the same.
The EBF encourages the ESAs to allow for the recognition of equivalent margin rules in other jurisdictions (as contemplated by the commentary to principle 7 indicate of the BCBS-IOSCO framework). This is important for EU banks who are registered swap dealers under Title VII Dodd-Frank Act and therefore will be subject to both the EU rules and US rules implementing the BCBS-IOSCO framework.
A need for a vast number of various agreements, for the purpose of not exchanging initial and variation margin, is required (positive election requirement). This will pose a significant operational burden, in particular with respect to the relationship to small and medium sized entities (see also response to Question 1 on the effects of this positive election requirement on small and medium sized companies).
The explanatory notes suggest that the exemptions were introduced to ease the operational burden and insure proportionate implementation of margining requirements. We welcome and agree with such considerations. However, the requirement of a positive election agreement (whether in writing or other equivalent electronic means) in order to benefit from exemptions seems in contradiction with such objective. Indeed, it will create administrative burden even for exempted counterparties, products, implementation phases. The benefit of phasing-in would be watered down as long as it does not reduce the documentation burden. The exemptions should therefore be structured a direct exemptions not requiring an opt-out (positive election). If contractual agreements are deemed to be necessary, this could be achieved by an obligation on all counterparties which are or become subject to the margin requirements to put in place the appropriate documentation (that is as of the time they become subject thereto).
It is not clear whether formal documentation is required where the transactions were entered into with exempted entities under Article 2 GEN 4 (b) and (c) i.e. NFC- of article 10 EMIR and sovereigns/supranational entities of Article 1 (4)/(5) EMIR. Indeed, the introductory section refers to writing or equivalent means only with respect to FC and NFC+, while such agreement is stated to be made with respect to “following” including (b) and (c). If so, the entities intended to be exempted would still be affected through the corresponding documentation requirement in order to benefit from the exemption.
It is not clear whether a written form is also required to be repeated each time the parties do not reach the phase-in thresholds in Article 1 FP (3) (i.e. are not subject to IM). Re-documenting all derivatives relationships with all (including exempted) counterparties seems disproportionate to the goal of fostering financial stability. In other words, we believe that where an exemption is available for practical exchange of collateral, it shall be consistently available for documentary requirements.
It is not clear what constitutes “other equivalent permanent electronic means” as alternative to writing and how this can ease the operational burden. As already mentioned above, the relevant agreements would involve negotiations and thus do not lend themselves to electronic standards (not a binary decision). In addition a significant portion of the affected counterparties will not have access to such electronic means and protocol system such as the one established for ISDA master agreements (see response to Question 1 on the limits of the use of protocol systems).
The operation burden of collateral substitution due to concentration limits of collateral must be addressed more appropriately. This requirement will result in increased settlement risks and new functionality requirements in Collateral Management systems.
The haircut on collateral for FX mismatch will result in more collateral movements and operational risks. They way exposure is calculated in CM systems will also be affected with requirements on developments as a result.
It would be helpful, if the definition of “currency mismatch” (on page 50 in the Consultation Paper) were clarified. We do not understand what is meant by settlement currency. It could, for example, mean the currency in which non-collateral payments are made in respect of a transaction, but this would not be meaningful for cross currency swaps or other transactions with payments in multiple currencies. Alternatively, it could mean any currency in which settlements may be made for the transaction, but this would then include the collateral payments themselves. Or it could have other meanings, including the base currency, but this may seldom be used for payments except for termination determinations. The haircut of 8% ought to be lower for currency pairs with low volatility, e.g. when one currency is pegged to another.
The RTS should make it clear that:
1. Netting across asset classes for the purposes of calculating VM is permitted. Article 1 GEN (3)(b) refers to collection of VM on a net basis, but this is not carried through to Article 1 VM.
2. Uncleared OTC derivatives between members of the same group should not be included in the calculation of the IM phase-in threshold. This is consistent with the application of the EUR50m IM threshold between consolidated groups. If not, there is a double-counting effect for back-to-back transactions to transfer market risk to the group member who holds the market making book and/or who contracts with external parties. This may cause a group to exceed an IM phase-in threshold when such back-to-back trading does not really represent incremental systemic risk.
The EBF considers that the setting up of the following two registers would facilitate the application of the margin rules by EU entities:
1. A public register of NFC+s. As each non-financial counterparty who exceeds the clearing threshold is required to notify ESMA and its competent authority under article 10(1) of EMIR, ESMA is in a position to maintain and publish a public register of each EU entity that is an NFC+.
2. A public register of EU FCs and NFC+s who are a member of a group whose aggregate month-end average notional amount of uncleared OTC derivatives exceeds the prevailing IM phase-in threshold. This could be combined with a requirement for such EU FCs and NFC+s to notify ESMA and their competent authorities by the 30 September following the applicable calculation period if their group exceeds or has ceased to exceed the applicable IM threshold for the next following annual period. This is consistent with paragraph 8.10 of the BCBS-IOSCO framework document.
This would help market participants to accurately apply the IM and VM requirements and reduce the reliance on representations from its counterparty to establish if and to what extent it must apply the margin rules to new transactions with that counterparty.
As far as collection of margins is concerned, the RTS foresee that this is done within one business day following the transaction date. However, standard settlement regimes applicable to securities are generally between 1 and 3 days. Hence, counterparties posting securities as collateral could be in breach of the RTS if collection of collateral is not consistent with securities settlement delays. Next to that, we would like to know to what point in time on the term “collection” itself refers to: it is not clear if this is the point of calculation, claiming or actually receiving collateral.
The EBF supports the introduction of more clarity around the definition of Minimum Transfer Amount (MTA). The MTA should be defined as per the current market practise as the minimum (threshold) amount that has to be settled between counterparties on any business day. By referring to collateral amount" it is not clear whether the definition refers to the cumulative full value or the collateral amount to be exchanged on that day.
The provision on the MTA in Article 2 GEN (4) (a) and (6) demands that the amount is calculated as the total amount of all initial margins and variation margins to be posted, that is without differentiating between variation and initial margin. The function and understanding of this MTA differs significantly from current practice. The operational introduction of this new MTA concept will be extremely challenging since it would require the implementation of new and very complex allocation and monitoring systems. In addition, the proposed new MTA concept could defeat the purpose of the MTA: Once the total amount is breached, even very minor differences (which occur regularly) would trigger margin calls needing to be processed (effective zero threshold), resulting in unnecessary and, considering the very limited risks involved, unreasonable additional operational burdens. Electronic processing can reduce these effects only to a limited extent and, is in any event, not an option in relation to those counterparties, which have no access to such electronic processing (in particular smaller and medium sized counterparties). In this context it should be taken into account that the risk exposure of credit institutions would, in any way, be addressed by existing the capital requirements under the CRR.
The described negative effects could be minimised by introducing two separate total MTAs, on for variation margins and another for initial margins, and an additional operative de-minimis threshold for any margin call (e.g. to the amount of 50,000 €).
We further propose to delete the last half sentence of Chapter 1, Article 2 GEN, paragraph 3 (p.23 of the draft RTS) which sets out the requirement to “hold capital” where no initial margin is to be exchanged (“and that they will hold capital against their exposure to their counterparties”. Such a requirement to hold capital requirements is unnecessary and may cause misunderstandings: FCs are, of course, already subject to (regulatory) capital requirements under the CRR. These, however, do not and are not intended to apply to NFC. Such requirements can also not be imposed by contractual agreement, not least because it would be impossible to determine whether the other counterparty complies with such an obligation.
We would welcome a clarification that no threshold other than Minimum Transfer Amount with respect to Variation Margin is authorised. Indeed, the second paragraph on the page 8 referring to a “minimum exchange threshold” of EUR 500,000 may lead to confusion and be interpreted as another threshold on the top of the Minimum Transfer Amount (where “margin requirement exceeds EUR 500,000”). Regardless, it is industry’s view is that the MTA should be applicable to IM and VM independently. By making it applicable to the consolidated figure of IM and VM, it would not only be operationally complex and intensive but inconsistent with current market practice."
The Proposal adequately addresses the risks and concerns of derivatives in cover pools and should not be tightened further.
The exemption should be extended to securitisation swaps for securitisations based on the same asset classes on the basis that such swaps effectively work in the same way as covered bond swaps. The requirement for securitisation vehicles to post initial and/or variation margin on swaps that they transact to hedge the cash flows on the underlying assets they hold against the principal and interest on the notes they issue will have a material adverse effect on the effectiveness of such transactions or structures as a funding tool. If a securitisation is required to post initial and/or variation margin, this will also have adverse implications for the asset encumbrance ratio of the originator bank as a result of asset pools having to be larger to generate adequate cashflows and/or the use of issuance proceeds, to fund initial and/or variation margin. This is detrimental to the position of unsecured bank creditors. We believe that adequate protection against counterparty credit risk is provided to swap counterparties through security on the assets that the vehicle holds and the over-collateralisation features embedded within such transactions and structures.
For this reason, we encourage the ESAs to consider exempting securitisation vehicles from the scope of the EMIR margin rules in order to ensure that this important funding tool for banks to support real economy remains effective.
Regarding Article 3 GEN (1)(a), it would be preferable that this requirement was removed in its entirety. It is not clear why it has been included and effectively precludes the covered bond issuer being the basis swap provider to the covered bond vehicle (to the extent that the application of the RTS is not precluded because the hedging transactions are entered into within the same legal entity). Should this requirement be retained, it is important that it is limited to insolvency related defaults only. As presently drafted no event of default (e.g. payment default) relating to the issuer would be permitted. Such a requirement would be incompatible with market practice and reach beyond the requirements applied by the rating agencies for AAA compliant covered bond related derivatives. The purpose of this restriction should be to avoid that the derivative is terminated as a result of the issuer’s insolvency, not to prevent the counterparty from terminating upon other limited non-insolvency related defaults. Therefore, we propose a change to add the words “insolvency related” before “default” in paragraph (a) of Article 3 GEN.
In addition, GEN 3 (1)(b) should be qualified with “except in the case of the default or insolvency of the derivative counterparty” to allow for flip clauses required by rating agencies as a condition to their ratings of covered bonds.
The regulation of covered bond programs and hedging related to them is quite strict and limiting at the moment. In our opinion, this level is adequate and even tighter regulation wouldn’t benefit the investors or the hedging counterparties. The underlying cover pool assets already cover the interest of the hedging counterparties to the same extent as the cover pool assets cover the interest of the covered bond holders and tighter regulation would only add administrative burden without notable benefits. A requirement to post collateral in respect of the derivatives which are also covered by the cover pool assets would result in a situation where the hedging counterparties would have a stronger protection than the investors. This would not be consistent with the investor’s protection point of view.
Regarding Article 3 GEN (1)(f), we believe it should be sufficient to have a de facto 2% over collateralisation and not a necessity to have a legal requirement in each jurisdiction.
At this stage it is not clear whether the proposed rules on the use of IRB models are suitable for practical purposes, in particular if it realistic for the parties to share sufficient information about the IRB model to be used.
The counterparty providing the model will of course need to provide appropriate information on the model to the counterparties to be confident in accessing sufficient information.
Ultimately, we believe that the best approach will be the application of a unified margin model as this will greatly reduce the complexities and operational problems since all counterparties relying on this unified model will have a common understanding of the information required for this model and the manner in which it is to be implemented. We support the use of internal models for determining collateral haircuts. However, these haircut estimates should not be run separately from the IM model themselves. To not take into account any correlations between the unsecured exposure, collateral or exchange rates, is likely to lead to more disputes than if they were otherwise taken into account.
Also, we believe that the standard schedule proposed would be overly penal in a number of cases, e.g. Danish Flex bonds are given a minimum 12% haircut for >5y issues, which could be harmful to the both issuers of and investors in these bonds.
We understand that Article 7 LEC, concentration limits for initial and variation margins, imposes an upper limit on how much collateral (received) in relation to a single issuer, or entities that are part of the same group or those with close links, shall not exceed 50 % of the collateral collected from any single counterparty.
The proposed concentration limits will require fundamental changes to existing collateral management systems and procedures as they differ significantly from current practice. This will result in substantial operational costs. This is further exacerbated by the fact that the proposed requirements differ from and conflict with similar requirements under the CRR. We see no real benefit in imposing restrictions on small amounts of exchanged collateral (as proposed in Article 7 LEC) as this will impose significant (and potentially harmful) operational burden for many counterparties. As an example this this could be especially problematic for certain entities e.g. SPVs that only hold certain types of assets and thereby are unable to meet these rules. The same applies to insurers, pension funds and other types of investment funds, which have to comply with strict investment rules and as such may not have a portfolio that allows sufficient diversification in securities offered as collateral. Moreover, we think that the risks inherent to the concentration of positions (directionality, volatility, liquidity) cannot be adequately captured by the basic diversification guidelines as suggested in the RTS. These risks should be measured dynamically by internal haircut models. As a consequence, collateral posted that is already subject to a haircut model should not be subject to concentration rules.
In particular the requirement to apply the concentration limits in relation to each individual counterparty increases the complexity and thus the operational challenges unnecessarily. Such counterparty-based concentration limits will also be challenging for the relevant counterparties, in particular small and medium-sized counterparties.
The purpose of concentration limits is served just as effectively but with significantly less complexity by permitting the collateralised counterparty to apply these in relation to all counterparties, that is in relation to its total exposure. This would not only reduce the operational complexity for the collateralised party. It would also prevent that collateralising counterparties are effectively denied access to the market simply because they are unable to diversify the collateral they have at their disposal.
Moreover, in some cases the concentration limits may actually be counterproductive as they might force counterparties to replace collateral of a very high grade by collateral of a lesser grade simply because the concentration limits have been breached. This may apply in particular to debt securities issued by sovereigns. In the context of equity derivatives customarily secured by collateral in the form of the underlying equity, forcing market participants to diversify collateral and to substitute it with uncorrelated collateral would result in a discrepancy between exposure and collateral which would increase the risks of collateral takers. This would also prove inefficient for collateral providers holding the underlying equity who will be deprived of other sources of liquidity in order to meet the diversification requirement. For the same reasons, collateral in the form of underlying equity should also be exempted from the requirement of the inclusion in main indexes.
If a decision is made to introduce concentration limits then they should be defined in such a way that it only defines limits to ensure that the value of and ability to liquidate the collateral is secured in the event of a counterparty default. We think that all concentration limits should be set in both absolute and relative terms, i.e. first when the absolute value of the collateral exceed a threshold, the relative limit will be applied so that only the largest of collateral positions (where concentration to a single name may in a material way impact the value of and ability to liquidate the collateral in event of a default) are required to be diversified. The absolute limits could be calibrated to the estimated credit quality of the collateral. In practise most of the collateral sent is in form of cash, therefore we do not think that it is appropriate to introduce concentration limits at the current time. It is as well not proportionate to introduce complex concentration limit mechanisms to address the small proportion of securities used as collateral today.
In addition, we also suggest that the concentration limits in relation to highly rated government bonds and covered bonds should at the least be significantly increased because the operational costs will out weight the liquidation issues under stressed scenarios. Counterparties should also be able to bilaterally agree concentration limits in relation to some minimum requirements.
To reiterate the above, in practise most collateral is sent in the form of cash. It would therefore seem inappropriate to introduce more concentration limits.
In European law-making, one should always ensure that the capability of European companies to create growth is not harmed with more stringent requirements. A level playing field between different markets is crucial especially in derivatives markets which are global in nature. Therefore we urge ESAs to follow the flexibility provided in global rules and ensure a level playing field for European companies. In that respect, we would like to point out that the scope of instruments covered under EMIR is wider than under the Dodd-Frank Act. Financial instruments such as equity options or derivatives on equity indices are neither considered “swaps” nor “security based swaps” in the US and consequently are not subject to margin requirements, contrary to the rules in Europe. This creates a major competitive disadvantage for European banks.
The requirements set out in BCBS-IOSCO are well suited to ensure high level protection of the original collateral. Some requirements for rehypothecation of collateral are also provided for in other European regulations.
We suggest that cash may be reinvested (instead of not re-used at all) in a very restricted range of products, which would be in line with the approach retained in the ESMA guidelines for UCITS.