RegRadar - RegStream
You are here: Home / Stream / ESMA’s response to the Commission’s consultation on the BMR review

ESMA’s response to the Commission’s consultation on the BMR review

ESMA - European Securities and Markets Authority

On Feb. 14, 2020, 8:45 a.m. • In ESMA News

Source: https://www.esma.europa.eu/file/54626/download?token=HhNe0uYg

14 February 2020 ESMA70-156-1778

ESMA’s response to the Commission’s consultation on the BMR review 1 Introduction ESMA welcomes the opportunity to respond to the Commission’s Consultation on the BMR review (hereafter “the Consultation”)1. The Benchmarks Regulation (BMR)2 entered into application on 1 January 2018. During the transitional period, competent authorities have been applying the BMR and registering, authorising, endorsing or recognising EU and Third Country (TC) administrators.

In this response, ESMA wishes to share with the Commission some reflections on a number of topics mentioned in the Consultation which appear relevant to ESMA, in particular the critical benchmarks regulatory framework, the scope of the BMR and the third country regime. Further, ESMA may submit additional comments to the Commission regarding the BMR review following the analysis of the responses of the consultation paper on the MAR review3. 2 Critical benchmarks 2. 1 IBOR reform Question 1: To what extent do you think it could be useful for a competent authority to have broader powers to require the administrator to change the methodology of a critical benchmark? Very useful As the underlying market that a critical benchmark intends to measure may evolve over time, it is important that the methodology of the benchmark continues to be representative of the underlying market over time, to the extent that it can be made so in the prevailing circumstances (there may be circumstances in which a benchmark cannot be made representative but its continued publication in a more stable and sustainable form may be desirable, if only for a limited wind-down period – see comments on question 2 below). 1 https://ec. europa. eu/info/sites/info/files/business_economy_euro/banking_and_finance/documents/2019-benchmark-review- consultation-document_en. pdf 2 https://eur-lex. europa. eu/legal-content/EN/TXT/?uri=CELEX%3A32016R1011 3 https://www. esma. europa. eu/press-news/consultations/consultation-mar-review ESMA • 201-203 rue de Bercy • CS 80910 • 75589 Paris Cedex 12 • France • Tel. +33 (0) 1 58 36 43 21 • www. esma. europa. eu Article 23(6)(d) of BMR “Mandatory contribution to a critical benchmark” grants competent authorities the power to “require the administrator to change the methodology, the code of conduct referred to in Article 15 or other rules of the critical benchmark”. A simple amendment to BMR could be to include the power to require the administrator to change the methodology in Article 21 “Mandatory administration of a critical benchmark”. When the competent authority is notified of the intention of the administrator to cease the provision of a critical benchmark the competent authority should have the power to require a change to the methodology, the code of conduct or other rules of the critical benchmark, if appropriate in the period after that notification.

More broadly, whenever the authority considers that the methodology and the input data of a critical benchmark are not representative of the underlying market or economic reality that the benchmark is intended to measure, or they are anyway no longer considered BMR-compliant, the authority should have the power to require a change of a methodology on the basis of its own assessment. This situation can take place also outside the framework of Article 23. Even when the panel of supervised contributors is stable there may be circumstances in which the critical benchmark is no longer representative. So, a more comprehensive amendment to BMR could be to extend this possibility in the following cases: a) whenever the critical benchmark is no longer representative of the underlying market or economic reality that the benchmark is intended to measure; b) when the methodology or the input data are no longer considered BMR-compliant; or c) when a mandatory administration commences or is in progress. Question 2: Do you consider that such corrective powers should apply to critical benchmarks at all stages in their existence or should these powers be confined to (1) situations when a contributor notifies its intention to cease contributions or (2) situations in which mandatory administration and/or contributions of a critical benchmark are triggered? Yes / no? Please explain. Yes Please see previous answer. The extension of the provision of Article 23(6)(d) to Article 21 would provide the competent authority of a critical benchmark with an additional tool to solve the potential cessation of the provision of a critical benchmark.

In particular, for a critical benchmark based on contributors that are withdrawing, continued publication to enable an orderly cessation may result in an increasingly volatile rate, as the number of contributors reduces. Where the underlying market for a benchmark is drying up, the robustness of a benchmark may be at risk due to the absence of underlying transactions. Even though the rate cannot be made representative, it may be possible to make it more robust, more stable and more sustainable (e. g.

not dependent on the availability of contributors) by changing its methodology. Therefore, ESMA is of the opinion that competent authorities should have the possibility to mandate a change of the methodology also in the following cases: a) whenever the critical benchmark is no longer representative of the underlying market or economic reality that the benchmark is intended to measure; b) when the methodology or the input data are no longer 2 considered BMR-compliant; or c) when a mandatory administration commences or is in progress. Question 3: Are there any other changes to Article 23(6)(d) BMR relative to the change of methodology for critical benchmarks that might be desirable to improve the robustness, reliability or representativeness of the benchmark? Yes / no? Please explain. Yes Please see answers to questions 1 and 2. 2. 2 Orderly cessation of a critical benchmark Question 4: To what extent do you think that benchmark cessation plans should be approved by national competent regulators? Agree completely – not agree at all (5 categories) + explain Agree In relation to Article 28(1), authorities of critical benchmarks should assess the administrator’s procedure both at the time of authorisation and regularly once the administrator is authorised. In addition, given the importance of these plans for critical benchmarks, ESMA considers that factors to be taken into account by administrators when creating and reviewing these plans should be set out in legislation; and for critical benchmarks the relevant competent authority should verify that these factors appear to have been given due consideration. In relation to Article 28(2) and the written plans covering benchmarks, competent authorities are not in a position to approve the cessation plans of all supervised entities for all the benchmarks that supervised entities use. Instead of a regulatory approval of all written plans as per Article 28(2), competent authorities should apply their power to review the written plans following a risk-based approach, whereby written plans focusing on critical benchmarks could be considered.

Other criteria could also be considered such as the duration of contracts referencing the benchmark and the likelihood of contract frustration and market disruption. However, the use of a critical benchmark by supervised entities should be allowed also before the assessment of the written plans by the competent authority.

The imposition of a new obligation requiring an ex-ante approval of the written plans of the supervised entities using benchmarks may result in additional administrative burden with no real regulatory benefit, as competent authorities already have the power to request the written plans maintained by EU supervised entities. In addition, the implementation of Article 28(1) and (2) by administrators and supervised entities would benefit from a general specification of what is meant by “in the event of changes” in paragraph (1) and what is meant by “material change” in paragraph (2). In particular, the distinction with the material change of the methodology as set out in Article 13 of BMR. 3 Question 5: Do you consider that supervised entities should draw up contingency plans to cover instances where a critical benchmark ceases to be representative of its underlying market? Yes It should be primarily a responsibility of the administrator to ensure and check that a critical benchmark is representative of the underlying market. Supervised entities can have different opinions in relation to representativeness of a critical benchmark, potentially leading to a fragmented use of the critical benchmarks.

The written plans should be clear and easy to implement. The definition of transparent triggers are crucial to define a viable written contingency plan. So, a public declaration by the relevant authority that a critical benchmark is not capable of being representative should be a trigger. Otherwise, the instance in which a critical benchmark is not capable of being representative of its underlying market is something difficult to be properly defined within a trigger.

The Article 28(2) plans of users of a critical benchmark should also cover the situation where the relevant authority has exercised its power to suspend the BMR authorisation of an administrator (or a benchmark – see Question 7) which provides a critical benchmark in circumstances where the critical benchmark ceases to be representative of its underlying market. Where authorities have the power to withdraw the BMR authorisation (at administrator or benchmark - see Question 7 - level) of a critical benchmark, they should do so when the critical benchmark is not representative of the underlying market and no other action can be imposed to make it representative. Once authorisation is withdrawn, as stated in Article 35(4), then the written plans adopted by the users of the critical benchmark will apply. This situation can also be reflected in the written plans with a trigger based on the withdrawal of authorisation by the relevant authority.

2. 3 Colleges Question 6: To what extent do you consider the system of supervision by colleges as currently existing appropriate for the supervision of critical benchmarks? Very appropriate – not appropriate at all (5 categories).

If not, what changes would you suggest? Very appropriate ESMA thinks that the system of supervision by colleges is appropriate as the NCAs of the supervised contributors of a benchmark should take part in any decision regarding the related benchmark. 4 3 Authorisation / registration 3. 1 Authorisation, suspension and withdrawal at benchmark level The consultation document seeks views from market participants on the applicability of Article 35 at administrator or benchmark level.

The document refers to a disruptive effect of a withdrawal or suspension at an administrator level preventing the use of all benchmarks of a particular administrator when only one of them has become non-compliant.

Question 7: Do you consider that it is currently unclear whether a competent authority has the powers to withdraw or suspend the authorisation or registration of an administrator in respect of one or more benchmarks only? Very unclear – very clear (5 categories) Very unclear ESMA agrees with the issue raised by the Commission and would like to point out that it is currently unclear whether a competent authority has the power to withdraw or suspend the authorisation or registration of an administrator in respect of one or more benchmarks. Indeed, Article 35(1) refers to the various instances where a competent authority may withdraw or suspend the authorisation or registration of an administrator.

The four instances raised in Article 35(1) can be read as applicable at an administrator level which in some circumstances could be very disruptive where most of the administrator’s benchmarks remain BMR-compliant. Therefore, ESMA believes that further clarity is needed in the BMR that it is possible to withdraw or suspend authorisation or registration at benchmark level. ESMA would like to highlight that it is unclear whether the authorisation or registration could be obtained at benchmark or administrator level. While so far different approaches have been taken, it seems relevant for NCAs to have the ability (but not the obligation) to grant registration or authorisation at benchmark level for a critical benchmark. In other instances and given the number of benchmarks at stake, the registration or authorisation should be granted at administrator level but suspension or withdrawal of that same application should be possible at benchmark level. Further, the authorisation or registration at administrator level may also concern a subset of the benchmarks that the entity provides. In addition, the other application processes such as the recognition application should follow the same path in order to ensure consistency throughout the BMR. It could also be made clearer in the BMR that competent authorities have the power to suspend or withdraw registration or authorisation in respect of a critical benchmark that is not representative of its underlying market (see comments at question 5 above). 3. 2 Continued use of non-compliant benchmarks Question 8: Do you consider that the current powers of NCAs to allow the continued provision and use in existing contracts for a benchmark for which the authorisation has been suspended are sufficient? Totally sufficient – totally insufficient (5 categories). Please explain. 5 Insufficient While the current powers of NCAs under Article 35(3) and Article 51(4) have not been used, ESMA believes that the withdrawal, and not only the suspension, of an authorisation or registration should also be referred to in Article 35(3).

ESMA believes that Article 35(3) regarding the suspension of an authorisation or registration of an administrator should also refer to the case of withdrawal of such authorisation or registration and not only to its suspension as this might lead to NCAs suspending rather than withdrawing authorisation or registration, in order to be able to permit the continued use of the benchmark in these contracts – and potentially to maintain this suspended status until all such contracts have matured.

Also, were the legislation to be amended to provide NCAs with this power in relation to withdrawn authorisation or registration, then obviously there would need to be accompanying changes to ensure that NCAs retain all necessary powers over administrators of benchmarks that are continuing to be published and used following withdrawal. Question 9: Do you consider that the powers of competent authorities to permit continued use of a benchmark when cessation of that benchmark would result in contract frustration are appropriate? Very appropriate – not appropriate at all (5 categories). Please explain. Appropriate ESMA is of the view that in order to avoid contract frustration in some specific cases, as contemplated in Article 35(3) and Article 51(4) of the BMR, the powers of the competent authorities to permit continued use of a benchmark are appropriate. However, given that some long-standing contracts contain a form of fallback provision that was intended for use in the event of short term disruption to a benchmark but which is not fit for purpose in case of a permanent replacement, and that amendment of some contracts will be difficult to achieve, ESMA thinks that it is important that the BMR gives flexibility to competent authorities to permit the continued use for all existing contracts or for some – but not all - of those contracts, or even none of them, depending upon the types and provisions of the contracts which may be affected. As this decision of continued use relates to different member states and not only the member state where the administrator is located, it is important to ensure that the usage of such benchmark in the Union is fully taken into account.

ESMA could ensure a coordination role across the different member states. ESMA further suggests that the BMR should include transparency for users of a benchmark. This permissible continued use should be communicated to market participants in order for users to know that they can continue to use the benchmark and for which contracts (if any) this permissible use is applicable. 4 Scope of BMR Question 10: Do you consider that the regulatory framework applying to non-significant benchmarks is adequately calibrated? Which adjustments would you recommend? 6 Scope should be adjusted Based on the administrators authorised or registered on ESMA register the vast majority of benchmarks (excluding interest rate and commodity benchmarks) available in the market are non-significant benchmarks.

While the BMR already includes a proportionate framework linked to different categories of benchmarks based on thresholds, it also includes a regime based on the category of the asset class underlying the benchmark (commodity or interest rate).

ESMA’s view is that the regulatory framework applying to non-significant benchmarks can be improved mainly for those benchmarks less prone to manipulation (see question 13). Question 11: Do you consider quantitative thresholds to be appropriate tools for the establishment of categories of benchmarks (non-significant, significant, critical benchmarks). If applicable, which alternative methodology or combination of methodologies would you favour? Appropriate ESMA considers that the use of thresholds for the establishment of categories of benchmarks is an appropriate tool only when the methodology is appropriately set and the input data are available (see reply to question 12). Question 12: Do you consider the calculation method used to determine the thresholds for significant and critical benchmarks remains appropriate? If applicable, please explain why and which alternatives you would consider more appropriate. ESMA considers that the calculation method used to determine the thresholds needs improvement.

ESMA believes that as set out in the Commission Delegated Regulation (EU) 2018/664 it is of paramount importance to be able to link the benchmark to the financial instrument referencing it and the way to perform this task is to use a unique identifier of benchmarks.

This unique identifier needs to be the one used in the reporting of derivatives under trade repositories and MiFID reference data.

Indeed, the calculation of the threshold is performed using those databases based on publicly available data.

These two databases use ISIN to identify benchmarks and therefore ESMA believes that the BMR should include a legal obligation for administrators to request an ISIN to identify their benchmarks (see reply to question 14). In addition, ESMA believes that the inclusion of systematic internaliser transactions renders the calculation of the thresholds complex as these transactions are not easily identifiable in the databases5.

4 https://eur-lex. europa. eu/legal-content/EN/TXT/PDF/?uri=CELEX:32018R0066&from;=en 5 https://www. esma. europa. eu/sites/default/files/library/esma70-151-370_final_report_tr_public_data_under_emir. pdf 7 Question 13: Would you consider an alternative approach appropriate for certain types of benchmarks that are less prone to manipulation. If so, please explain for which types. The BMR sets out a framework for regulated data benchmarks that are less prone to manipulation because they are already subject to a regulatory framework.

Further, those benchmarks cannot be considered as critical.

ESMA believes that alternative approaches could be considered for those benchmarks that are less prone to manipulation. As stated in the consultation document, many third countries have opted for an approach whereby regulation and supervision is limited to the most critical or systemic financial benchmarks administered in their respective jurisdictions, whilst BMR covers all types of benchmarks. This raises two issues. First, such situation results in an unlevel playing field for EU benchmark administrators which administer benchmarks that are not critical nor systemic but that are obliged to comply with the BMR requirements. Second, the wide scope of BMR led to generally unexpected consequence with respect to TC benchmarks. Indeed, it becomes now clearer that TC administrators of less significant benchmarks are not necessarily incentivised to keep providing such benchmarks in the EU if they are required to comply with the relevant provisions of BMR. One way to solve these issues could therefore consist in reducing the current scope of BMR. Following a risk-based approach, ESMA believes that this could be achieved by excluding non- significant benchmarks based on regulated-data pursuant to article 3(1)(24) of BMR. Were the legislation to be amended accordingly, ESMA suggests the Commission to perform further analysis to ensure that no unintended consequences occur. 5 ESMA register of administrators and benchmarks Question 14: To what extent are you satisfied with your overall experience with the ESMA register for benchmarks and administrators? If not, how could the register be improved? Not satisfied Article 36 of the BMR requires ESMA to publish and maintain a public register including the list of administrators authorised or registered under Article 34 of the BMR, the list of benchmarks recognised under Article 32 of the BMR together with the related administrators, the list of benchmarks endorsed under Article 33 of the BMR together with the related administrators and the list of administrators and benchmarks for which the Commission has adopted an equivalence decision under Article 30 of the BMR. ESMA has set up the register as of 1 January 2018 which includes as of November 2019, 84,558 third-country benchmarks and 58 EU and TC administrators. The issue in the current set up of the register as provided in the BMR is twofold. First, it creates an unlevel playing field between EU and TC administrators as the register does not include EU benchmarks but only EU administrators (see question 15).

Second, in accordance with Article 29(1) of BMR, supervised entities can use a benchmark provided by an EU administrator included in the ESMA register or a benchmark included in the ESMA register. However, the BMR does not 8 provide for a common identifier to enable users of benchmarks to identify a benchmark in the database.

The aim of the ESMA register is to provide users with the list of administrators and benchmarks that can be used in the Union.

The number of these benchmarks or administrators can be significant and therefore the current requirement in the BMR regarding “the identities of the administrators” and “the list of benchmarks” should, in ESMA opinion, be further detailed by specifying a common identifier for administrators and for benchmarks. Indeed, it is of paramount importance that a user of a benchmark is able to identify the benchmark it is able to use in the register. Without a common identifier the only available field to identify a benchmark is the full name of the benchmark which is a free text that is prone to errors and therefore cannot be used to unambiguously identify a benchmark or an administrator. ESMA also believes that as described above, the calculation of the reference value of benchmarks is closely linked to the capacity of market participants and regulators to consistently identify benchmarks in the different data reporting systems in order to measure the use of the benchmark in the financial instruments. Finally, as described in the box below, the clear identification of the benchmarks used by the market will have other positive effects on other areas of financial supervision, as it will allow further clarity in the information reported under other sectorial regulations. Box 1. MiFID and EMIR: The negative effects of the absence of benchmark identifiers Even though data on the use of reference rates is already being collected under the applicable supervisory reporting requirements under EMIR, MiFID/R and MAR, their utility for supervisory purposes is limited due to the lack of a reliable and consistent way of identifying benchmarks and therefore the ultimate nature of the transacted financial contract or instrument. Based on September 2019 EMIR data, ESMA analysed the use of benchmark identifiers. Out of all outstanding transactions, 7. 3% had an index as underlying but within them 33% of the records (2. 4 million transactions) lack an ISIN, whichprevents financial supervisors from having a clear understanding of the financial risks. In the case of the Markets in Financial Instruments Regulation (MiFIR) and the Market Abuse Regulation (MAR) trading venues and Systematic Internalisers are required to submit identifying reference data for the relevant financial instruments to competent authorities and then to ESMA (Commission Delegated Regulation (EU) 2017/585).

This reference data on traded financial instruments, must be provided in accordance with the standards prescribed by the Regulation, and is published daily for the use of investors, market participants and financial supervisors. The reference data is key to ensure adequate surveillance of financial markets and the transparency regime prescribed by MiFIR. ESMA analysed the reference data as of September 2019 and found that for fixed-income instrument, for 15% of the records the used benchmark was reported under free text as “Other” or other variants of the word, 35% of the records referring to derivatives were affected by similar issues.

9 In the absence of reliable identifiers, i. e. ISINs, the information reported to the regulator implied additional burdens for reporting entities (as they lack a clear reference on the information to report) and for their financial supervisors, as it reduces the effectiveness of the reporting regimes. Therefore, the lack of strong identifiers results in a decreased capacity to monitor the risks for retail investors, investigate market abuse, monitor market integrity; and analyse the hazards to financial stability. Initiatives to reduce this effect, as maintaining manual mappings, are both cumbersome to implement and to maintain, given the frequent changes in free-text reported fields. ESMA has assessed the utility of the Register considering the low provision by reporting entities of international identifiers for administrators and benchmarks (i. e. LEIs and ISINs). The lack of obligation to ensure designation of adequate identifiers has additional detrimental impacts on retail investors, market participants and financial supervisors beyond the scope of the BMR. Therefore, ESMA proposes that the Benchmark Regulation shall: include ISO 6166 ISIN as the mandatory identifier for the compliant and EEA-used benchmarks; and ISO 17442 LEI as the mandatory identifier for their administrators. (see Annex for further details) Question 15: Do you consider that, for administrators authorised or registered in the EU, the register should list benchmarks instead of/in addition to administrators? Fully agreed ESMA believes that the current set of the ESMA register does not provide enough transparency to the market because (as pointed out by the Commission consultation document) the EU administrators may have benchmarks that are not BMR compliant. However, the register does not allow users of benchmarks to differentiate the BMR compliant benchmarks from those that are not BMR compliant.

Investors, reporting entities and supervisors need to have clarity on eligible benchmarks and on their main metadata.

ESMA therefore proposes that the register should include information at benchmark level for both EU and TC benchmarks to enhance transparency to and clarity for benchmark users on the benchmarks that they can lawfully use.

ESMA should, in a central location, publish all benchmarks and their key metadata (e. g.

Name, ISIN, CFI, FISN, date of authorisation or withdrawal of a benchmark) as well as the information on their administrators.

The access should be machine-to-machine readable, so market participants can execute due diligence tasks at low cost through so-called RegTech, as described in the reply to the question 14. The central location will reduce the costs for administrators, retail investors, market participants and financial supervisors for both searching for the relevant data in a comprehensive publication and using the data, as it will benefit from a common structure that in its current configuration allows for both human lookup and machine-readable access. ESMA acknowledges that setting up such a central register, containing all compliant benchmarks with frequent updates in its content, constitutes a substantial change to the current functioning of the register as it requires IT systems to ensure the timely transmission of data from administrators to their supervisor (NCA or ESMA) and from the NCAs to ESMA, 10 as well as a publication system by ESMA.

This will imply also relevant operating costs to ensure the prompt publication of high-quality and frequently-updated data. Given that publication was already foreseen under the current BMR for non-EU administrators, some of those IT systems are already in place (i. e. transmission of data from NCAs to ESMA and ESMA publishing the data) but additional costs might be driven by increased volumes and some additional improvements in the content of the current register.

While EU Administrators are currently not required to build the IT systems allowing the timely transmission of benchmark information to NCAs, the relevant development costs can be foreseen were the legislation to be amended. However, the operational costs to produce the information relevant for the register, should be negligible as administrators should maintain this data for their on-going operations.

In addition, ESMA would further consider from a practical perspective the various means to minimise such costs. A decentralised publication sometimes is considered as an alternative.

In such alternative, administrators publish the benchmarks in their website and NCAs provide to ESMA an updated list of hyperlinks to the location of these publications to be published. This alternative, although attractive in terms of low development costs, is impractical for market participants and investors when trying to validate an offered benchmark as compliant.

Beyond the need to maintain updated hyperlinks, the publication by administrators through different formats, supportive files, different denominations of variables or even in different locations within the same page, will make any automated process ineffective or very costly to maintain.

Therefore, this alternative is not proposed by ESMA. Regarding the proposed centralised register, a relevant example when considering feasibility is the current publication of reference data on instruments collected under Article 4 of the Market Abuse Regulation and Article 27 MiFIR and published by ESMA (FIRDS system). This system publishes information for more than 16 million references with data provided by NCAs to ESMA or, under a delegation agreement, directly by trading venues to ESMA.

The operations executed by FIRDS are much more complex in nature than the ones foreseen for benchmarks.

Further, the provision by ANNA of its ISIN lookup service described in the annex to the reply to the question 14 is not a substitute of the proposed increased scope of the benchmark information in the ESMA register. The reason is that ANNA’s service might contain information for other benchmarks that are not compliant according to the BMR, a relevant status for the users of ESMA register.

Also, were the legislation to be amended to list all benchmarks on ESMA’s register, then obviously there would need to be accompanying changes to ensure consistency of the BMR. For example, Article 29 on the use of benchmarks would need to be amended to reflect those changes. 6 Benchmark statement Question 16: In your experience, how useful do you find the benchmark statement? 11 Useful Pursuant to Article 27 of the BMR, the benchmark statement shall include different elements relating to the calculation of the benchmark, the procedures governing the determination of the benchmark and the procedures dealing with errors in input data. The purpose of the benchmark statement in the BMR is to summarise the methodology document.

Noting this overlap between the two documents, a link could be provided in the Benchmark Statement to the publicly available methodology document/s (or the page / site that hosts this information). Further, the Benchmark Statement is currently not included in the authorisation requirements, rather administrators are required to publish the benchmark statement within two weeks of inclusion in the Register. Given the importance of the Benchmark Statement, consideration could be given to an explicit power for authorities to request it as part of the authorisation process. Question 17: How could the format and the content of the benchmark statement be further improved? In the context of the new low carbon benchmarks regulation, the Technical Expert group6 has provided the commission with an advice including a template for the benchmark statement to be applicable for all benchmarks except interest rates and currency benchmarks that includes a number of quantitative information at the level of each benchmark and provides more valuable Environment Social Governance (ESG) information. ESMA is of the view that in order to enhance the usability of the benchmark statement by investors and to achieve its objective to enhance transparency and comparability of benchmarks, the information included may be improved by adding quantitative information regarding the benchmark, for example the historical performance of the index, the top 10 constituents, and the key information (such as ISIN and currency). Question 18: Do you consider that the option to publish the benchmark statement at benchmark level and at family level should be maintained? While so far different approaches have been taken, it seems reasonable for administrators to group some benchmarks in one family and provide one benchmark statement at the level of the family of benchmarks.

However, ESMA believes that a benchmark statement that encompasses hundreds of thousands of benchmarks does not provide users of such a benchmark with the granular information needed to choose between two benchmarks most appropriate for their investments.

Therefore, ESMA believes that the BMR should further specify the concept of family of benchmarks.

A family of benchmarks could be a group of benchmarks that have different variants linked to the maturity or tenor or currency but that measure the same underlying 6 https://ec. europa. eu/info/sites/info/files/business_economy_euro/banking_and_finance/documents/190930-sustainable- finance-teg-final-report-climate-benchmarks-and-disclosures_en. pdf 12 market or economic reality.

While currently the calculation of the threshold is based on the range of maturities or tenors of the benchmark (this is mainly applicable to interest rate benchmarks), a number of administrators provide different “variants” of the same benchmark. These variants could be based on a methodology distinction, e. g. price return vs net return. Some of the variants could also be a different currency of the same benchmark, i. e.

the benchmarks measure the same underlying market however have a different currency which consist in the application of an FX rate to the base currency.

Therefore, ESMA considers that the ability to publish the benchmark statement at the family level should be maintained but proposes to specify further, and reduce the scope of, the current definition of family of benchmarks. 7 Supervision of climate-related benchmarks Question 19: Do you consider that competent authorities should have explicit powers to verify (1) whether the chosen climate-related benchmark complies with the requirement of the Regulation and (2) whether the investment strategy referencing this index aligns with the chosen benchmark? The BMR includes powers for competent authorities to verify that the chosen climate-related benchmark, like any other benchmark, complies with the requirement of the regulation.

The BMR requires competent authorities to register or authorise an administrator located in the EU. However, the BMR does not provide competent authorities with a process for the on-going supervision of the benchmarks already included on the ESMA register. While Article 41 of the BMR refers to the powers of competent authorities and specifies the means by which authorities can fulfil their duties, it does not include any provision related to the ongoing supervision.

The same process would be applicable to climate-related benchmarks.

The competent authorities will assess the benchmarks included in the authorisation or registration application and in this context will be able to verify that the chosen benchmarks, including potentially climate-related benchmarks, comply with the requirements of the BMR (i. e.

minimum requirements of the methodology and disclosure requirements).

ESMA generally believes that the ongoing supervision of benchmarks merits to be made more explicit in the BMR. Indeed, the BMR is silent on the process when administrators authorised or registered launch a new benchmark in a category of benchmarks that did not exist at the time of the authorisation, e. g. climate-related benchmarks and commodity benchmarks. ESMA recommends that the BMR includes specific requirements for administrators to notify their relevant competent authority when a new benchmark category is provided.

ESMA agrees that it is a best practice for competent authorities to verify that the investment strategy of an investment product referencing a benchmark aligns with the chosen benchmark. However, ESMA does not support the inclusion of a specific requirement in the BMR for competent authorities to check the alignment of the investment strategy of the investment product to its referenced benchmark.

13 Question 20: Do you consider that competent authorities should have explicit powers to prevent supervised entities from referencing a climate-related benchmark, if such benchmark does not respect the rules applicable to climate-related benchmarks or of the investment strategy referencing the climate-related benchmark is not aligned with the reference benchmark? ESMA highlights that all benchmarks following ESG objectives fall currently under the scope of the BMR and in particular the non-significant benchmarks category and therefore are governed by the relevant requirements under Article 26 of the BMR. Further, ESMA is of the view that the BMR should include a proposal to provide competent authorities with explicit powers to prevent an administrator from launching EU climate-related benchmarks identified as the two new types of ‘Climate-related Benchmarks’ (the EU Paris Aligned Benchmark and the EU Climate Transition Benchmark) while not respecting the rules as stipulated in the delegated acts to be adopted by the Commission. However, ESMA does not believe that these powers should be applied to supervised entities but rather to administrators that the BMR regulate. 8 Commodity benchmarks Question 21: Do you consider the current conditions under which a commodity benchmark is subject to the requirements in Title II of the BMR are appropriate? No Price assessments rely to a great extent on voluntary contributions. The requirements in Title II of the BMR do not seem to be appropriate for these prices. The condition in Article 19(1) which imposes the requirements in Title II of the BMR on commodity benchmarks based on submissions by contributors the majority of which are supervised entities may not have had the intended effect.

Since Title II of the BMR cannot be applied to most price assessment, the condition in Article 19(1) merely imposes a restriction on the number of (potential) contributions and hence negatively affects the representativeness (and potentially the continuity) of the assessments. What is more, with the implementation of MiFID II, many commodity trading houses have become investment firms thereby reinforcing the impact of this requirement. Therefore, ESMA suggests to the Commission to review the application and appropriateness of the requirement in Article 19(1). Question 22: Do you consider that the compound de minimis threshold for commodity benchmarks is appropriately set? Yes ESMA considers that as commodity benchmarks are impacted by seasonal effects, it is important that those are taken into account in the calculation of the threshold.

Commodity benchmarks should be in scope of BMR only if the average value of total notional amount of the listed instruments over a relevant period of time is above the threshold.

14 9 Non-EEA benchmarks Question 23: To what extent would the potential issues in relation to FX forwards affect you? If so, how would you propose to address these potential issues? ESMA agrees with the issue raised by the Commission on the FX forwards and believes that the Commission’s suggestion that these benchmarks may be suited to an exemption might be useful, as this would preserve the integrity of the BMR’s existing third country regimes and would allow businesses to continue to hedge their currency risks by using these benchmarks.

The BMR already exempts certain public policy benchmarks, for example those administered by central banks and those administered by public authorities. An exemption that builds upon this existing rationale to allow for certain types of public policy benchmarks, including those linked to not fully convertible currencies and relevant FX rates may be useful. However, it may be worth considering a limit on this exemption – for example, where such a benchmark meets the critical benchmark threshold, it should not be exempt from the BMR.

Question 24: What improvements in the above procedures do you recommend? This question is also linked to the questions above on the scope of the BMR, as the EU legislators need to ensure a level playing field between EU and non-EU administrators.

In relation to the general scope of BMR and the 3rd country regimes, the BMR should clearly state the scope of the exemptions in Article 2 of the BMR.

Also, the Commission should consider the instances in which a 3rd country benchmark becomes “systemically important” in the Union because of its extensive use by EU supervised entities. Such benchmark cannot be classified as a critical benchmark under the current BMR. In the hypothetical case in which this benchmark is covered by an equivalence regime, the risk of financial stability stemming from its role in the EU financial system are limited. But if this is not the case, it may be appropriate to include in the Regulation additional requirements applicable to this “TC critical benchmark” and additional powers for the relevant authority.

In relation to the recognition regime, the role of the legal representative should be defined in the L1 text. Its responsibilities and legal liabilities should be clearly stated in BMR (see relevant Q&A; published by ESMA), and BMR recitals can provide examples of which types of entities are fit to take the role of legal representative under the BMR recognition regime. Recognition could be incentivised by a clear legal framework around the legal representative that minimises administrative burdens and additional costs while ensuring oversight of the TC benchmarks and accountability vis-à-vis the relevant authorities.

The need of cooperation arrangements with the TC authority under Article 32(5)(a) should be clarified as needed only in cases where the administrator is subject to supervision “in relation to the provisions of benchmarks”. The recognition regime also currently relies on the concept of the Member State of reference, which has been difficult to implement by TC administrators. The ESAs review text solves this issue as ESMA will become the competent authority for the recognition regime and there will be no more need of the Member State of reference. So, no need for additional actions on this issue.

15 16 Annex – Additional information regarding question 14 As described in the EC Fitness Check of EU Supervisory Reporting Requirements 7 the insufficient use of standards, common formats and identifiers is one of causes of an inefficient use of data reported under the applicable regime8. In addition, the same document identified that the lack of mandatory reporting of LEI represents a key data gap to be addressed by the Union legislators9.

ESMA argues that these two generic findings should be addressed accordingly in the BMR review, as the introduction of an obligation to use harmonized and standardized identifiers for both the legal entities and the benchmarks they administer would considerably enhance the effective use and interpretation of the information by retail investors, market participants and financial authorities. In the case of administrators, a robust identifier of the legal entity is needed to execute adequate supervision of administrators, considering the conflicts of interests that might arise given their ownership structure.

This becomes more relevant for the cases of third-country administrators when the information for the supervisor endorsing the administrator might be more limited. In the case of the benchmarks, both from EEA and third-country administrators, the mandatory use of strong international identifiers will have the following benefits regarding the objectives of the BMR: • The unambiguous identification of the benchmark reinforced by its publication in a public register will ensure clarity of the contracts referring to such benchmark and thus contribute to investor protection. In the absence of an identifier subject to strict formats, standards and governance processes, the administrator-generated full name of the benchmark (i. e. free text) prone to errors and lacking a robust process for its generation and review.

identification would rely on the Moreover, it is a common practice among benchmark administrators to provide groups of benchmarks based on a similar basket of instrument with similar text strings in its description.

But each benchmark in the group may drastically differ in key characteristics (e. g.

currency, hour of calculation, methodology) making an unambiguous identification important for its user10.

• The quality of the register will improve as LEI and ISIN fields will need to be reported: currently, although benchmarks administrators are supposed to provide both LEI and ISINs of their benchmarks when available, the absence of a duty to obtain them 7 https://ec. europa. eu/info/publications/191107-fitness-check-supervisory-reporting_en 8 Ibid. , p. 9. 9 Ibid. , pp. 68 and 113.

10 For example, as 15 November 2019, in ESMA Benchmark Register, among the benchmarks reported by STOXX Ltd. there are 486 benchmarks whose name contain the string “EURO STOXX 50” and whose behaviour compared to the ordinary “EURO STOXX 50 ESTX 50 EUR (Price)” with ISIN EU0009658145 may be completely the opposite as in the case of “EURO STOXX 50 Daily Short” with ISIN CH0029194971. 17 prevents ESMA and NCAs to make mandatory their provision to the ESMA Register. Consequently, there is no practical possibility to reject any transmitted information lacking those identifiers 11

12 .

Making LEI and ISINs mandatory would allow retail investors and market participants further clarity on the scope of compliant benchmarks. • The existence of a complete Register including strong identifiers would allow further use of so-called RegTech to allow market participants to follow regulatory and compliance requirements more effectively and efficiently. Finally, it should be noted that the explicit requirement to use international standards, including in particular LEI and ISIN, is already contained in the L1 text of other regulations such as SFTR and EMIR REFIT. Inclusion of an equivalent requirement in the BMR would further support harmonisation of the regulatory requirements in line with the findings of the Fitness Check. Furthermore, the L1 text of MiFIR requires use of the LEI by the investment firms in the transaction reports to identify their clients13.

As stated in BMR recital number 3 “[b]enchmarks are vital in pricing cross-border transactions, thereby facilitating the effective functioning of the internal market in a wide variety of financial instruments and services”. Therefore, the existence of clear identifiers on the administrators and on the benchmarks has positive effects beyond the specific scope of the BMR, as it will be described below. This had also already been raised by ESMA in its Technical Advice under the Benchmark Regulation14, in which ESMA had recommended to the EC to request the assignment of ISINs to benchmarks from administrators.

ESMA had, on that occasion, highlighted that the adoption of such identifiers would have improved the measurement of the reference value of the benchmark and would have made more reliable and easier the consequential evaluation of critical benchmarks.

Hence, ESMA proposes that the BMR should set the mandatory obligation for each administrator to obtain a LEI, and in addition ISINs for their compliant benchmarks that are used in the EU as defined in Article 3(1)(7) of the BMR. This section describes the merits of these international standards. Both identifiers are governed international standards: the ISO 17442 in the case of the Legal Entity Identifier for unique identification of legal entities relevant for financial transactions; and ISO 6166 for the unique identification of financial instruments. They are very well known and extensively used by international investors, market participants and supervisors, so third- country benchmark administrators can easily obtain those if required by EU Regulation. 11 This might result in non-provision of available identifiers difficult to identify clearly by financial supervisors and therefore a decreased quality in the Register.

12 Furthermore, the quality of the published registers shall improve as those standards will reduce the possibility of inconsistencies, duplications and other quality issues. 13 Following to the implementation of transaction reporting under MiFIR in January 2018 resulted in a sharp increase of EEA LEIs from 0. 3 million in September 2017 to 0. 7mn in April 2018. 14 https://www. esma. europa. eu/sites/default/files/library/2016- 1560_final_report_on_technical_advice_on_benchmarks_regulation. pdf, p. 27. 18 In both cases, appropriate governance frameworks and infrastructure for supporting their generation and use is available as described below: - LEI: The Global Legal Entity Foundation (GLEIF15) is the supranational not-for-profit organisation established by the Financial Stability Board, tasked to support the implementation and use of LEI. GLEIF accredits the 33 Local Operating Units (LOUs) that issue LEIs.

- ISIN: The Association of National Numbering Agencies (ANNA 16 ) is the global association of 117 national numbering agencies (NNAs) with the mission to provide reliable, accurate and trusted means to identify and describe securities and other reference information like indices that can be used by all nations and their markets. Each NNA maintains the relevant codes for each instrument including the ISIN (ISO 6166), CFI (ISO 10962) and FISN (ISO 17442) as well as links the information with the LEI of the issuer of the instrument.

In accordance with ANNA’s ISO 6166 “ISIN Registration Authority obligations”, when fees are charged by the NNA for ISIN allocation, this is done on a cost-recovery basis only. In both cases, the frameworks have well established quality rules for ensuring the adequate and unique allocation of identifiers to legal entities and instruments according to the abovementioned international standards.

The public access for retail investors, market participants and regulators is also ensured in both cases. GLEIF provides access to all available LEIs both with lookup access and machine- readable format. ANNA also offers an ISIN lookup service which enable users to search for ISINs issued by NNAs and includes in the search results the CFIs and FISNs of the instruments. Furthermore, for the ISINs for derivatives issued by ANNA DSB, the users may find the most relevant instrument characteristics in the respective database.

Finally, the process of assignation of both LEI and ISIN generates valuable additional contextual information described below: - LEI: The so-called Level 2 data provides information on the ownership structure of the legal entities, although at this point still incomplete in its scope. This information could be relevant to determine possible conflict of interests arising from the ownership or control of the administrator. - ISIN: As described above, the allocation of an ISIN includes the assignation of both a FISN and a CFI for each instrument or in this case a benchmark. Both provide relevant contextual information that shall reinforce the transparency for the users of benchmarks as described below: o FISN: The Financial Instrument Short Name standard provide a consistent and uniform approach to standardize short descriptions for financial instruments. It 15 https://www. gleif. org/en/about-lei/get-an-lei-find-lei-issuing-organizations 16 https://www. anna-web. org/anna/about-anna. 19 aims to harmonize existing market practices which are in use on a national and individual entity level17. The existence of an independently standardized short name can be also in the future considered in other regulations for providing standardised descriptions of the related benchmarks to retail investors. o CFI: The Classification of Financial Instruments code provides a uniform set of codes to classify financial instruments and in addition it contains a category specific for reference instruments (e. g. currencies, commodities, baskets and indices). In the case of indices, the CFI provides a clear classification regarding the asset classes (e. g. equities, debt, real estate, commodities), weighing types and index return type.

Regarding the possible costs for benchmark administrators to obtain LEI and ISINs for their benchmarks, it shall be considered that both frameworks are operated and managed by international non-profit organizations, with low cost for each individual assignation of identifiers.

17 https://www. iso. org/obp/ui/#iso:std:iso:18774:ed-1:v1:en 20