Unfortunately, it was not possible to run in-depth specific analysis regarding the cost implications for small or medium-sized entities. However, the implementation of the new regulatory framework will lead to several costs for small and medium-sized entities, as the Eur 8 bn notional threshold is most likely to force them to bear the following type of costs:
• legal costs, due to the renegotiation of existing CSAs;
• IT costs for the computation and storage of the IM for each counterparty, where no “calculation agent” be appointed by an entity, and for the implementation of “ex-ante decision” on eligible collateral;
• organizational costs, as to improve collateral management processes on dispute resolution, to set up a monitoring unit on collateral eligibility and concentration;
• segregation costs in relation to received IM (funding and liquidity costs) which cannot be reused.
Quantifying estimates for the items above was difficult as these closely depend, among other things, on decisions as outsourcing or insourcing certain activities.
In the aim of minimizing the impact in terms of costs of the new proposed framework, we consider that the introduction of specific conditions under which allowing the re-use of collateral could mitigate the impact. We suggest ESMA to consent the reuse of the received collateral exclusively with the European Central Bank (under specific and strict conditions and monitoring) so as to allow an entity to gather liquidity and for funding activity. In addition, in order to reduce the operational and the IT cost, we suggest ESMA to propose the use of a detailed “standardized (internal) model”, consistently to the feedback provided on the explanatory text on page 27-28 (please, refer to the specific section further below).
One of the major issues related to the proposed framework concerns the application of the Eur 50 mln threshold at the level of the consolidated group (to which the threshold is being extended) and is based on all non-centrally cleared derivatives between two consolidated groups, parties to a transaction. More specifically, the issue is related to the potential split of the overall threshold among the different Legal Entities of a group. As initial margins are currently calculated and settled at a local level, the obligation set in the framework would be very demanding to be managed.
A potential compromise solution – which would need to be endorsed ex ante by regulators – could be represented by the signing an overarching CSA between groups, where these agree on whether dynamic or static allocation is allowed, by appointing, at Group level, the relevant Legal Entities authorized to deal bilateral OTC derivatives and assign a percentage of the Eur 50 million threshold to each relevant Legal Entity. A renegotiation of an existing overarching/Group-Level CSA would be necessary where any new group entities (appointed by the Group as being allowed for trading) were to be added in the CSA or any change of counterparty status from NFC- to NFC+ should occur.
The definition of “Group” within EMIR (Reg. 648/2012, Art. 2 (16)) includes the case of the Institutional Protection Scheme (IPS, as defined by article 113(7) of Regulation (CRR) 575/2013). While we agree with the inclusion of IPS in the Group’s definition (as EMIR does) when we refer to the clearing exemption (Art. 4(2) EMIR), we believe that an IPS shall not be considered as Group, at least in the case an IPS does not fulfill the conditions of making use of intragroup exemptions, as it occurs in the following specific cases:
a) Exemption from initial margin (article 2 GEN, par. 3);
b) Eligible criteria to avoid wrong-way risk (article 6 LEC, par. 1, point b))
c) Concentration limits for initial and variation margins (article 7 LEC, par. 1, point [a) and] b)).
To that regard, we would like to highlight that, generally, an IPS is made up of several small banks and, at least under the relevant prudential regulation (i.e. CRD IV/CRR package), is not fully treated akin to a consolidated banking group. Therefore, including the IPS in the definition of “Group” for the purposes of the points a) to c) above would bring about unintended consequences since the corresponding requirements will likely impact disproportionately on those small banks which are not integrated in a consolidated banking group.
In addition, the ESAs consultation paper states that (art. 2 GEN, (4)(b)) financial counterparties (FC) and non-financial counterparties (NFC+, as referred to in EMIR, art. 10) may agree not to exchange initial and variation margin for transactions entered into with NFC-. It is important to note that the provision proposed by the ESAs would (1) require changes in the contractual terms under which existing contracts are entered into, which were already updated upon the entry into force of EMIR, and (2) make necessary a re-negotiation of contractual terms with clients to comply with the provision proposed by the ESAS to agree not to exchange IM and VM. This represents one of the aspects of the new framework which would demand significant efforts for the legal and commercial departments, despite that fact that EMIR does not explicitly demands the parties to formally agree in writing not to exchange IM and VM. Indeed, EMIR imposes the exchange of collateral only for FC and NFC+, therefore we deem not appropriate to impose an agreement to avoid exchanging IM and VM for transactions entered into with NFC-.
Further potential issues are highlighted below:
• as it regards the method for IM calculation, the new framework is clearly laid out in a way as to prevent disputes to arise; however, should this occur, regulators should consider to lay out a provision that would allow the parties to a transaction to agree on the identification of one of the counterparties as calculation agent. In this specific case, both parties shall also agree on the identification of a third-party calculation agent which will provide a calculation that will be binding for the parties. The latter solution proposed does represent a way to balance the pervasive powers/rights of the party that would be assigned the role of calculation agent (in the agreement) with the role of a third-party calculation agent, so that this would protect the party that is not being assigned the role of calculation agent.
• The new framework should specify in which way the full disclosure of covered entity should be done to the market; potential solutions should be:
- documenting, during CSA signing, the status of the bilateral counterparties and which one allows IM collection in case entities are fall within the average notional thresholds;
- signing dedicated documentation in which a counterparty certifies its status;
- implementing a single common platform to monitor the status (average notional).
As per Article 1 FP, par. 5, we understand that the new framework does include foreign exchange forwards, swaps and currency swaps. However, we are still convinced that this position should be reconsidered and exclude physically settled FX transactions from the calculation of the average notional of non-centrally cleared OTC derivatives for the purpose of the phase-in period for the application of initial margin (IM). This would preserve consistency with the regulatory framework which excludes such transactions from the scope of the IM exchange requirement. On the same grounds, we believe that the forthcoming RTS should make it explicit that intra-group transactions which have been exempted from the margin requirements are not taken into account for the purpose of the phase-in calculation of the average notional of non-centrally cleared OTC derivatives.
In relation to Art. 1 LEC, ‘Eligibility and treatment of collateral’, we believe that cash assimilated instruments (as defined by article 4(1)(60) of CRR), should be included in point a), par. 1 of Art. 1 LEC provided that the instruments are issued by the collateral taker. Indeed, CRR recognize these instruments as credit risk mitigation instruments and the Delegated Regulation proposed by the ESAs should take this consideration into account in order to keep consistency with the CRR provisions.
Finally, the definition of practical or legal impediment proposed in the consultation paper seems to be too wide and it lays down very strict criteria for entities which are members of the same Institutional Protection Scheme (IPS). We believe the specificities of the IPS should be reflected more appropriately. Many regulatory regimes and all insolvency and resolution regimes, by nature, do contain provisions which can affect the ability of the regulated or insolvent party, or the party under resolution, to effect payments or transfer assets, and, unless such impediment is only deemed to exist upon initiation of such proceedings but not before, it would be impossible for groups and members of the same institutional protection scheme to rely on this exemption.
The proposal seems to adequately address the risks related to Covered Pool OTC derivatives dealt to close IR or CCY risk.
We consider it important and “worthwhile” to reach a deeper harmonization across existing legislations, especially with regard to minimum requirements in terms of overcollateralization requested, even if in Italy Rating Agencies apply large haircuts to bonds that are part of covered pool(s). Nonetheless, we consider that Covered Bonds rules are sufficient to grant an adequate degree of protection to the counterparty of the hedging derivative transaction without applying the IM (e.g. protection against counterparty downgrade clause, termination events clause).
With specific reference to the use of CRR instead of the UCITS definition of covered bonds, it is important to point out that article 129 of the CRR sets out specific rules for banks’ exposures in the form of covered bonds whereas the UCITS V (art. 52.4) only contains a definition (to which the CRR itself refers) for setting up investment limits of UCITS. In this respect, we highlight that the relevant Italian regulatory framework (Law no. 130/1999, as amended thereafter) identifies new eligible asset classes for covered pools. We thereby support an enlargement of the CRR definition so as to incentivize covered bond issuances without growing up risks.
Feedback on the explanatory text on page 27-28.
The Italian banking and financial industry would greatly value any market-led initiative aimed at achieving an industry-wide common approach on the margining model(s) to be used by all stakeholders. This consideration stands on the grounds that the new framework will imply a much higher (daily) frequency of calculation of the collateral/margining requirements comparable to what currently occurs. Hence, having counterparties to agree on the main features of a margining model would guarantee (a) more-frequently closer, if not identical, outcomes in terms of amount of collateral for a given trade, as well as (b) a decrease in the number of reconciliation issues and related claims.
We support the use of internal models already validated for regulatory purposes, with the following specifications:
1. Clear definition by the Regulator of the metrics (e.g. Potential Future exposure with a defined confidence interval). In particular, we suggest to use the internal model framework for counterparty risk since it is designed to model netting agreement at counterparty level;
2. Use of the internal models when at least one counterparty has an internal validated model (agreement on the calculation agent). This could allow non-internal model counterparties to have liquidity benefit when trading with internal model counterparties.
3. No need of other validations processes, because those models already passed a detailed validation process.
In addition, we consider with interest the initiative undertaken by ISDA, which has put forward a “standardized internal model”, namely an internal margining model with standardized features, and this initiative seems to move in the right direction as it appears to be easily replicable, relatively fast and easy to use. In consideration of the current market volumes of non-CCP cleared derivative instruments, the major market players operating in Italy would likely be subject to the Delegated Regulation (subject to this consultation) not starting from the first wave. When they become subject to the provisions of the DR, they will mostly have to deal with the major international players which will have applied these provisions from wave 1 and, because of this, will likely have to “comply” with the standards of the models more largely used in the market by that time. Hence and eventually, it will be generally convenient and effective to agree upon, as soon as technically possible, a market-led margining model that any counterparty will be in a position to implement and run. We believe that in the absence of such a market-led and agreed-upon margining model, implementation costs would turn to be too high, also in consideration of the potential increase of costs related to a larger number of disputes. All this is not “directly” linked to prudential requirements, rather to the liquidity profile of the counterparties which will have to segregated specific assets.
Besides the above, it would also be interesting to consider the option of complementing the Credit Support Annex (CSA) to the ISDA Master Agreement with the indication of a third party accountable as Calculation Agent. This seems useful as the application of different scenarios is susceptible to bring to different results and consequent disputes.
Eventually, as it regards the list of eligible collateral on page 33 – about which no comments are asked by the ESAs, we would like to highlight that current CSAs will have to be re-negotiated with clients to comply with the new provisions. This represents one of the aspects of the new framework which imply significant efforts from the legal and commercial departments, despite the framework does not explicitly demands parties to review CSAs to include the entire range of eligible collateral referred to on page 33 and 34.
The provisions set in the draft Delegated Regulation are clear. Some of our member banks which are using IRB models, also use external rating and provide National Competent Authorities with full disclosure. However, in consideration of the changes brought to the landscape of margining and collateralization, we consider that using and referencing to external rating would allow parties to benefit of fewer disputes, relatively to the much larger number of collateral computations which are looming on the horizon.
Firstly and foremost, impacts are expected – and actually already taking place – in the IT space: for example, appropriate procedures will have to detect whether a given ISIN is acceptable as collateral and its amount. Impacts are clearly expected in the weekly and monthly post-booking monitoring activities as well as for audit and compliance processes. Other impacts can be envisaged in the collateral replacement costs in the event that a breach is detected: a Collateral Margining Unit would have to contact the counterparty and try to ask for the replacement with a different type of collateral in order to solve the breach. This can easily lead to transaction costs and, also, long negotiation with the counterparty.
As it regards small banks, especially small cooperative banks whose portfolio is almost exclusively made up of domestic sovereign securities (also due to constrains by law or Statute), these would be unduly penalized where concentration limits be introduced on that category of asset. We strongly suggest the ESAs to achieve an alignment with the CRR liquidity rules and, therefore, concentration limits should not be applied to securities issued or guaranteed by EEA Sovereigns and EEA Central Banks in (their own respective) domestic currencies. Alternatively, we believe that a proportionality threshold should be introduced and based on the amount of the collateral to be collected from an individual counterparty and issued by a single issuer with respect of the total size of the market.
As an alternative to the proposal described above, concentration limits can be “modeled” to “naturally” exempt the best issuances in terms of Rating classes, concentration of issuance, collateral types, counterparty type. Rather than being a real exemption, a huge percentage in the composition analysis assigned to AA issuers can of course avoid breaches in this kind of issuances, reduce risks and facilitate the use of best currencies and issuances as collateral.
As it regards concentration and “exemption” of specific securities, typically sovereign bonds (i.e. govies), we can highlight some Pros and Cons. Commencing from the Cons, as far as wrong way risk is concerned, we are of the opinion that Letter ‘c’ at the bottom of page 37 should be further detailed as to avoid or mitigate what it currently seems strictly implied in the provision(s) set forth, i.e. banks and financial intermediaries operating in Country XYZ would be prevented by posting collateral composed of sovereign bonds issued by such Country XYZ. In this regards, we would suggest rephrasing such Letter c) mentioned above with the following: “…they have a proven record as a reliable source of liquidity during stressed market conditions…”. Also, it would seem necessary (and appropriate) to set a correlation threshold between the counterparty and the issuer in case of country default, above the threshold that collateral can be defined as “risky” and cannot be exempted, especially in case of low rating of issuer (typically non-investment grade).
As Pros, we can identify the high liquidity of the securities in position, eligibility to central banks, lower haircut applied, smaller cost of monitoring if the bank accepts this kind of collateral only.
Finally, as it regards the principle of proportionality, and, specifically, Art. 2 LEC, par. 1, d, the principle of proportionality should be applied in the case of smaller banks, as they access markets through another institution. Namely, we deem ESAs should take into account the following EBA recommendation in its Report titles “On appropriate uniform definitions of extremely high quality liquid assets” (December 2013), where it stated “(…) based on the proportionality principle, smaller banks which access markets through another institution, will, in most cases, not have to be active in several advanced money and capital markets.”
As the conditions to be fulfilled for the reuse of initial margins appear not easily matchable, ESAs could consider the proposal to consent the reuse of at least the collateral delivered to the European Central Bank in exchange of liquidity. Otherwise, one further space to lightly increase the flexibility of the provisions set forth would be to allow the use of the collateral received by any counterparty to meet the collateral request of CCP.