« Back

PensionsEurope

Question 1. What costs will the proposed collateral requirements create for small or medium-sized entities, particular types of counterparties and particular jurisdictions? Is it possible to quantify these costs? How could the costs be reduced without compromising the objective of sound risk management and keeping the proposal aligned with international standards?
IORPs will face significant costs under the proposed Draft RTS. Please find below an explanation of the impact on IORPs by answering each of the three different parts of this question.

A. What costs will the proposed collateral requirements create for small or medium-sized entities, particular types of counterparties and particular jurisdictions?

 The requirement for IORPs to exchange Initial Margins (IM), contrary to the current practice, would increase their costs and negatively impact the retirement income of pension beneficiaries. It will also negatively affect their capacity to contribute to the long-term financing of the European economy

Requiring highly creditworthy, conservative, long term, stable, low leveraged entities such as IORPs and their asset managers to post IM, is opposed to the current market practice and will significantly increase their costs of using derivatives.

IORPs and financial institutions managing assets on their behalf currently only exchange Variation Margin (VM) with the counterparties, not IM, precisely because they are considered as having extremely low counterparty risk / being highly creditworthy (see explanation further below in this question). Mandatory IM requirements will therefore impose new and costly funding requirements for IORPs.

Given that IORPs do not have access to expedient and low-cost liquidity sources (as opposed to other institutions that have direct access to the European Central Bank’s liquidity mechanisms), the costs of exchanging IM would be very significant. This adds to the fact that the derivatives traded by IORPs and financial institutions managing assets on their behalf are typically long-dated and one directional, meaning that very little offsetting options exist in the portfolio that would reduce the overall amount.

Moreover, the one-sided exposure leaves IORPs in a disadvantaged position in management of IM in comparison to derivative dealers, which count with a larger trading flow with offsets and a broader base of counterparties to allow for lower margin requirements. Consequently, the impact of IM will be disproportionately high for IORPs and financial institutions managing assets on their behalf.

Mandatory high levels of IM will therefore oblige IORPs to put cash-reserves aside in order to meet margin rules. In practice this means that capital will need to be diverted from productive investments. The proposed draft RTS would therefore result in higher costs for IORPs and this will inevitably have a negative impact on the retirement benefits of European pensioners. And this at a time when public (pay-as-you-go) pension systems are increasingly confronted with strong economic and demographic constrains which make public pensions progressively shrink while the retirement age continues to increase.

One must bear in mind that IORPs are mostly not-for-profit institutions operating often in the legal form of foundations (or trusts). Their activity is embedded in the Social and Labour Law of EU Members States and for many of them also in collective agreements between social partners (employer and employees representatives). Their purpose is to provide old age, invalidity and survivor benefits for employees and former employees. They do not have to pay any dividends to shareholder/investors. Therefore, investment profits, investment losses and associated costs of investing will impact, directly or indirectly, on the amount of retirement income and other benefits of members of the IORP in question.

The increased costs will also force IORPs to remove capital from productive long-term investments to meet short-term margin demands. This would be against the objective expressed by the European Commission in its Communication on long term financing of the European economy published on 26 March 2014 . In this Communication, the Commission correctly identifies IORPs as long term investors and calls to mobilise private sources of long-term financing alternative to banks in order to promote long-term investments in Europe.

 IORPs use derivative instruments for risk-mitigation purposes, not for speculative reasons. Burdensome IM requirements would discourage them from using derivatives to hedge their financial risks

OTC derivatives contracts are an essential tool used by IORPs and asset managers managing assets on their behalf to mitigate the financial risks of their investments. The IORP Directive mandates IORPs to be managed on a prudential basis and specifically only permits the investment in derivative instruments insofar as they contribute to a reduction of investment risks or facilitate efficient portfolio management . Any speculative use of derivatives is therefore prohibited.

In practice, OTC derivatives are used to match the duration of pension scheme arrangements with the duration of their liabilities. IORPs need to hedge the risk arising from their long-term investments, such as interest rate, volatility or currency risks, with a view to guarantee their funding and security to provide retirement benefits.

Without hedging against this risks, IORPs would be exposed to the volatility of the financial markets. Consequently, and due to the inherent problems of IORPs for posting IM, IORPs may become more risky entities, which is exactly the opposite to the objective of the EMIR legislation.

 Concentration limits on initial and variation margin will impose further costs for IORPs, in particular concentration limits on governments bonds

The concentration limits for initial and variation margins set out in Article 7 LEC will impose further burdens/costs to IORPs and their asset managers, particularly if the concentration limits on governments bonds are not removed. Although we understand the rationale of said limits, we believe that they go much further than internationally agreed standards and they do not reflect the collateral needs of IORPs.

In the first place it should be noted that the only asset for which concentration limits are not required is cash. However, IORPs are fully invested institutions what means that they do not have cash pools immediately available. Therefore the proposed non-limit on cash collateral is not useful for IORPs.

IORPs typically invest in asset classes that aim to reflect their long-term liabilities. In other words, an IORP will not hold a significant part of its assets in cash, as that is not consistent with investing prudently with a view to generating appropriate returns to the long-term liabilities of the IORP. Fixed income assets such as government bonds have limited returns compared to the potential returns on equity investments. But, on the other hand, they, subject to credit risk, provide a high certainty of a particular level of return.

For IORPs, which have the obligation to invest in a prudent and diversified manner , government bonds are a key investment tool . Government bonds are also used by IORPs to hedge against interest rate risk on their liabilities. Furthermore, and as opposed to cash collateral, government bonds permit IORPs to not reduce the investment returns. The economic ownership remains with the transferor; the transferee does not have to pay interest on collateral received.

Not removing concentration limits on government bonds will force IORPs to engage in asset transformation transactions, which will entail additional costs for IORPs and their asset managers. Moreover, and decisively, they will also bring undesirable valuation issues of derivative instruments, which will have an effect on the effectiveness/efficiency of the use of derivative instruments (see our response to question 5 for further details on this point).

 Small and medium-size IORPs will likely need to apply the proposed IM and VM regardless of the thresholds of Article 1 FP

The proposed draft RTS seek to reduce the burden for medium and small entities by introducing a threshold of an entities’ OTC derivative exposure above EUR 8 billion in gross notional outstanding. Although we appreciate the effort to promote proportionality, we fear that small and medium size IORPs will nevertheless be confronted with the IM and VM of this draft RTS.

Indeed, small and medium size pension funds have very little bargaining power vis-à-vis their bigger counterparties with which they engage in bilateral OTC derivative transactions (typically big banking institutions). We fear that the big banking institutions will force IORPs to apply the IM and VM requirements as described in the draft RTS if they want to be their counterparty in an OTC derivative transaction.

Indeed, derivative dealers falling above the thresholds of Article 1 FP will apply the proposed IM and VM rules when bilaterally trading OTC derivatives with large counterparties that also fall within the same thresholds. The bank’s most important bilateral OTC derivative transactions will be of this nature, and therefore we expect that they will not adapt their respective business models for small OTC derivative transactions with small and medium size IORPs and will require IORPs to apply the same IM and VM.

In view of the above, the proposed thresholds are not sufficient to reduce the burden on small and medium size IORPs. It is therefore it is of outmost importance to adequately define the margin requirements for IORPs and financial institutions managing assets on their behalf.


B. Is it possible to quantify these costs?

 The European Commission is finalising an impact assessment that will quantify the costs of central clearing OTC derivatives for IORPS and future pensioners

Article 85 (2) of EMIR establishes that by 17 August 2014 the European Commission has to publish a report were it will assess, among other things, the adverse effects of centrally clearing OTC derivatives on the retirement benefits of future pensioners.

Although the outcome of the report has not yet been published, we expect that the report will highlight the significant costs of exchanging margin collateral for fully invested entities such as IORPs, and, consequently, the costs that they entail for future European pensioners.

For example, if an IORP were to post an IM of €1 million, it might, in the current interest rate environment, expect to earn under 0.5% per annum if the IM were to be posted in cash. If the IM were posted in cash for 1 year, the return on the €1 million would be at an interest rate of 0.5% or less would be €5,000 or less.

In contrast, if the IM posted were €1 million of government bonds, with a yield of, for example, 3% a year posted for 1 year, the return on the IM for the IORP posted on the IM would be €30,000 (or a return which could be in the region of 6 times the return on cash) .

In PensionsEurope we believe that the European Supervisory Authorities should carefully review the report of the European Commission once it is finalised before adopting a decision on the risk-mitigation techniques for not cleared OTC derivatives applicable for IORPs and the financial institutions managing assets on their behalf.

In the past the Investment Management Association (IMA) had estimated that central clearing under current arrangements would reduce investment returns for a fully immunised Liability Driven Investment (LDI) portfolio by 1.1-1.9 percentage points . In other words, on a portfolio of €1 billion, this would reduce the annual rate of return on the portfolio by between €1.1 million and €1.9 million. If the maturity of the particular portfolio was 20 years (and ignoring any discount), the nominal amount of this additional cost would be between €22 million and €38 million.

C. How could the costs be reduced without compromising the objective of sound risk management and keeping the proposal aligned with international standards?

 The costs for IORPs and their beneficiaries would be significantly reduced by not requiring IORPs to post IM and by eliminating the proposed concentration limits on government bonds

As explained above in the response to this question, the requirement to exchange IM with their counterparties would not only be against the current practice but would also impose significant costs for IORPs (regardless of their size) and their beneficiaries. It will also discourage them from using derivative instruments to hedge their financial risks.

Article 2 GEN of the consultation paper refers to “risk management procedures in specific cases”. It is PensionsEurope’s view that the specific features and purpose of IORPs, together with the reason / way they use derivative instruments, as well as the spirit of the EMIR Level I text as regards the treatment of pension scheme arrangements, represent sufficient reasons that justify including a specific reference in Article 2 GEN granting IORPs and their asset managers the possibility to not exchange IM.

Moreover, also as explained above, the proposed concentration limits on initial and variation margin, especially on government bonds, will also result in increased costs for IORPs. We therefore call for the removal of these concentration limits, or at least grant an exemption for IORPs in this regard on the basis of its special features, namely due to the composition of its investment portfolio.

 The extremely low counterparty risk of IORPs ensures that the objective of sound risk management is not compromised.

EMIR and the proposed rules on risk-mitigation techniques on non-cleared OTC derivative transactions have for primary objective to the reduction of risk in the European financial markets. In this regard, IORPs and financial institutions managing assets on their behalf already represent very low counterparty risk:

- IORPs are already required to be prudently managed and in a diversified basis (see Article 18 of the IORP Directive) . IORPs and their asset managers therefore need to make extensive use of derivative instruments to hedge their investment risk such as volatility, currency or interest rate risks. The use of derivative instruments for speculative purposes is not allowed by the IORP Directive, which specifically only allows IORPs to use derivative instruments for risk-mitigation purposes or to increase the portfolio efficiency . The use of derivative instruments by IORPs therefore seeks to protect the savings of future pensioner by making IORPs secure counterparties protected against risks that may arise.

- IORPs are not leveraged entities. Contrary to other highly leveraged financial institutions, IORPs are not allowed to be leveraged. Article 18 (2) of the IORP Directive (2003/41/EC) forbids IORPS “from borrowing or acting as a guarantor on behalf of third parties”. Member State national competent authorities may occasionally authorise institutions to carry out some borrowing only for liquidity purposes and on a temporary basis.

- European pension funds are also subject to an extensive set of rules regarding their solvency and liability coverage ratio, which are set either at EU or at national level. The regulatory framework ensures that pension funds’ coverage ratios do not fall below certain minimum levels. Additionally, IORPs are able to fulfil their obligation as they count with the funding and/or backing from one or several employer companies / plan sponsors of the IORP. They also benefit from other protection mechanism such as national pension protection funds . Eventually, they could also make use of benefit reduction mechanisms such as increasing pension contributions, or not deciding to index pensions or cut pension benefits.

- As long-term investors, IORPs invest in appropriate investments reflecting their long-term liabilities. Consequently, the risk of an IORP not to fulfil its obligations as counterparty in a derivative contract in the short term is very remote, as a shortage in the short-term can be recuperated in the long-term.

- IORPs and their asset managers posting adequate and well-controlled VM provide sufficient protection (collateralisation) against counterparty failure.

For the above reasons the risk of an eventual default of an IORP is negligible and they are highly creditworthy counterparties. Financial markets participants indeed perceive it as such and do not require IORPs and their asset managers to exchange IM in bilateral OTC derivative contracts.

We would like to recall that the Commission itself acknowledged that IORPs did not experience the same problems as other financial institutions during the crisis : pension funds did not require any support in terms of funding from public finances. On the contrary, pension funds paid the costs of the crisis, in terms of lower investment returns. Furthermore, they contributed to water down the crisis, by keeping their long-term liabilities in the financial markets.

In PensionsEurope we consider that the proposed “risk-based” mitigation techniques for non-cleared OTC derivatives are not adequately “risk based”, because they only (try to) differentiate in terms of size of the entities, but not in terms of the risk each of them effectively pose.

 Recognising the specific features of IORPs will be in line with the declared objective of implementing international standards while taking into account the specificities of the European financial markets

The consultation paper states that the objective of the Draft RTS is to implement internationally agreed minimum standards while taking into account the specific aspects of the European financial markets . One of these specificities of the European financial markets is the special treatment given to IORPs and their dedicated asset managers in the EMIR Level I legislation.

Indeed, EMIR establishes a temporary exemption from central clearing obligation for pension scheme arrangements in recognition of their specific features and the special role of IORPs . This special treatment is unique in the international comparative law, and should also be reflected in the margin requirements of the proposed Draft RTS.

EMIR level I text mandates IORPS to be subject to bilateral collateralisation, but it does specify that this should mean to exchange both IM and VM. Well-controlled (applying due diligence) VM posted by IORPs and their asset managers provide sufficient protection (collateralisation) against counterparty failure. Requiring highly creditworthy, long term, unleveraged institutions such as IORPs to post IM risk making superfluous the temporary exemption for IORPs from the clearing obligation. It will entail significant adverse effects for IORPs and their beneficiaries and would be against the rationale behind the treatment of pension scheme arrangements in EMIR.
Question 2. Are there particular aspects, for instance of an operational nature, that are not addressed in an appropriate manner? If yes, please provide the rationale for the concerns and potential solutions.
1. Rationale behind EMIR temporary exemption from the clearing obligation for IORPs

EMIR establishes a temporary exemption from central clearing for pension scheme arrangements , due to the difficulty of IORPs to post cash variation margin in CCPs, and also in recognition of the “diversity of pension systems across the Union” and of the “special role of IORPs and the materially adverse effects that the new legislation would have on them and on pension beneficiaries” .

It is PensionsEurope’s firm view that the introduction of the obligation to exchange IM for IORPs would be against the rationale behind the transitional exemption from central clearing set for pension scheme arrangements under EMIR. Indeed, EMIR establishes that pension scheme arrangements should be subject to bilateral collateralisation , but it does not establish that this bilateral collateralisation should consist in the need to post both IM and VM. As explained in the previous question, posting IM would have significant “material adverse effects” on IORPs (regardless of their size) and on their beneficiaries and would compromise their “special role”.

We are aware that the IOSCO and BCBS report on Margin requirements for non-centrally cleared derivatives of September 2013 was agreed by international regulators. Nevertheless, EMIR was adopted by European democratically-elected institutions and therefore the spirit and rational of EMIR should prevail, in particular as regards European specificities. Moreover, the IOSCO/BCBS report would allow room for such treatment when it states that “the precise definition of financial firms, non-financial firms and systemically important nonfinancial firms will be determined by appropriate national regulation” .

The European Parliament and the Council will eventually be granted a scrutiny period to consider whether to endorse or not the proposed RTS in this consultation paper. It remains to be seen if they would endorse the RTS if they understand that they are against what they had originally approved in 2012 and could have a negative impact on the different pension systems of the European Member States.

With the occasion of the endorsement procedure of previous EMIR Delegated legislations, the ECON Committee of the European Parliament debated a motion for a resolution to reject the regulatory technical standards for similar reason . The resolution was finally not voted in the Plenary of the European Parliament so as to avoid further delays in the implementation of the Regulation, but warned against similar situations in the future. More recently, Ms Sharon Bowels, former Chair of the ECON Committee, expressed the need to respect the position of the European Parliament on EMIR on the Level II legislation as far as the treatment of European pension funds is concerned .

2. The use of external credit ratings by IORPs

a) The credit quality requirements for internal assessments and the use of external credit ratings should be the same

PensionsEurope understands the ESAs’ intentions to reduce the systematic reliance on external credit ratings. We support encouraging stakeholders to perform due diligence when using ratings of external credit assessment institutions (ECAIs) and to combine them with their own credit risk assessments.

However, we firmly consider that the reduction of the reliance in credit ratings must not result in entities such as IORPs having to replace the work undertaken by the credit ratings agencies. In the absence of viable and more reliable alternatives, the ESA’s may promote the use of internal credit ratings through broad principles, but in any case IORPs must be able to continue using credit ratings. Notwithstanding their drawbacks and limitations, credit ratings are a valuable tool to evaluate the credit worthiness of many businesses, governments and financial instruments.

In this sense, IORPs must not be penalised for using ECAIs instead of performing their own credit risk assessments. This is currently the case in the draft RTS. Indeed, Article 3 (5) LEC penalises the use of external credit ratings vis-à-vis the performance of own risk assessments (Article 3 (4) LEC).

PensionsEurope strongly believes that this is not an acceptable approach, since it would be applicable to the all entities, without taking into account the nature, scale and complexity of their activities which is required by the Credit Rating Agencies Regulation (CRA Regulation) . In this regard, on 12 May 2014 the Financial Stability Board (FSB) published its thematic peer review report on reducing reliance on credit rating agency ratings. In this report, the FSB states that “The FSB, in collaboration with the standard setting bodies, should provide clearer guidance for smaller financial entities, in particular pension funds, on how to more effectively implement the FSB Principles to address the issue of proportionality for smaller entities”. It also recognises that “the use of CRA ratings provides economies of scale in analysing credit on behalf of smaller and less sophisticated investors, and that implementation should differentiate according to size and sophistication of the firm”. Furthermore, it should also be noted that the implementation of the CRA Regulation as regards the reduction of overreliance in credit ratings it is still at a very early stage. The European Supervisory Authorities still need to develop the Guidelines that will inform stakeholders and national supervisory authorities how to proceed in this topic.

In view of the above, in PensionsEurope we strongly believe that the credit quality requirements for internal assessments and for the use ECAIs should be equal.

b) In those cases were an asset itself is not rated, it should be possible to assess the credit quality of the issuer.

An important part of the investment portfolio of IORPs and financial entities managing assets on their behalf is composed of government bonds, which are also exchanged as variation margin in their bilateral derivative transactions. In order to ensure the liquidity of the assets posted, often minimum ratings apply. Typically, these ratings are applied on the credit quality of the “issuer” of the security, since frequently government bonds themselves are not rated.

Therefore, in order to reduce the burden on IORPs, instead of Article 3 LEC only requiring the credit quality of “assets”, it should say also say that in case the asset itself is not rated, a valid approach would be the credit quality assessment of the “issuer”.

c) Minimum rating requirements should apply to all assets posted, irrespective of the issuer.

Contrary to what is established in Article 1 (3) LEC, we see no reason to exclude the assessment of the credit quality of government bonds issued by Member States governments, central banks, regional governments and public sector entities denominated in domestic currency. In PensionsEurope we believe that minimum rating requirements should apply to all assets posted, regardless of who is the issuer.

d) IORPs are likely to not be able to use the Internal Rating Based (IRB) model of their counterparty to assess the credit quality of collateral

When IORPs engage in OTC derivate transactions with banking institutions, traditionally the latter are not willing to share with IORPs information on the models and pricing techniques that they use to determine the credit quality of the collateral posted. Unless adequate measures are introduced to force/oblige banking institutions to share with their counterparties this information, chances are that they will continue with their current practice of not sharing such information. In case this circumstance was to persist, IORPs will not be in a position to adequately understand the IRB model of their counterparty, and therefore they will not be able to use it to assess the credit quality of collateral. This will leave IORPs in a disadvantageous position vis-à-vis their counterparties to assess the credit quality of collateral, especially if we take into account that many of the small and medium size IORPs are not sufficiently equipped to produce their own internal risk assessment models.

This point is further developed in our response to question 4 of the consultation paper.

3. Importance of government bonds for IORPs

Government bonds represent a key investment tool for IORPs and their asset managers. The imposition of concentration limits on government bonds will cause important operational disruptions to IORPs and their asset managers, not only in terms of costs but also, importantly, in terms of efficiency in the use of derivative instruments.

A very important part of the asset mix of European IORPs is composed by government bonds (in particular from the country of residence of the IORP) . Subject to credit risk, government bonds provide a certainty of return on the investment for the IORP and can be used by IORPs to hedge the interest rate risk inherent in calculating the present capital value of the obligations of the IORP to pay pensions in the future.

Historically, IORPs have been using government bonds as VM on their bilateral derivative transactions. Establishing limits to the use of government bonds in IM and VM will force IORPs to disinvest in government bonds and have in their portfolio bigger amounts of riskier assets. Decisively, concentration limits on government bonds will also bring to IORPS valuation issues of derivatives instruments, which would impact their efficiency in using derivatives; they valuation of high quality government bonds based on EONIA discounting will not be always possible due to diversification requirements.

PensionsEurope considers that concentration limits on government bonds should not apply to IORPs. Alternatively, we suggest less harmful measures for IORPs such as setting a maximum on the amount of collateral posted as a percentage of the total amount of government bonds issued by one single issuer.

In our response to question 5 the impact of concentration limits on government bonds on IORPs is further developed.
Question 3. Does the proposal adequately address the risks and concerns of counterparties to derivatives in cover pools or should the requirements be further tightened? Are the requirements, such as the use of the CRR instead of a UCITS definition of covered bonds, necessary ones to address the risks adequately? Is the market-based solution as outlined in the cost-benefit analysis section, e.g. where a third party would post the collateral on behalf of the covered bond issuer/cover pool, an adequate and feasible alternative for covered bonds which do not meet the conditions mentioned in the proposed technical standards?
IORPs and financial institution managing assets on their behalf are important investors in covered bonds. In this sense, in PensionsEurope we are concerned about the security of these vehicles.

We would prefer that the covered bond issuer / cover pool does not exchange any collateral at all, even in the case of hedging derivatives. This would enhance the security of the pool behind the bonds.
Question 4. In respect of the use of a counterparty IRB model, are the counterparties confident that they will be able to access sufficient information to ensure appropriate transparency and to allow them to demonstrate an adequate understanding to their supervisory authority?
Traditionally, banking institutions are not open to share proprietary information on the models and pricing techniques that they use. Consequently, if an IORP has to demonstrate an adequate understanding to their supervisory authority
of the Internal Rating Based (IRB) model of its counterparty, then it is likely that this option will not be able to be used; IORPs will not have sufficient information, and therefore IORPs will not be able to adequately understand the IRB model of their counterparty. Exactly the same applies to the IM model of the counterparty.

If IORPs and their asset managers would have sufficient information on the IRB model, they would be able to use the IRB model of their counterparty banking institutions and therefore both counterparties would be in equal conditions. If they lack this information, IORPs will not be able to use one or two of the three options provided in Article 3 (1) LEC, as opposed to their counterparties of bilateral derivative transactions (banking institutions) which would be able to use the three options.

In the case of small and medium size IORPs, they will not be able to use 2 of the 3 options of Article 3 (1) LEC, since it is likely that they will not have the necessary resources to produce their own internal models. Indeed, small and medium size IORPs, (which regardless of the thresholds will likely need to apply the draft RTS since banks will not be willing to adapt their business model for smaller counterparties) are frequently not equipped with sufficient resources to set up internal models and therefore it is more cost-efficient for them to apply due-diligence when using external credit ratings.

PensionsEurope believes that this situation unfair and therefore we demand that measures are adopted to force banking institutions to provide sufficient information to their IORP counterparties on their IRB model. PensionsEurope supports requiring credit institutions to disclose to their counterparties sufficient information on the structure of the models that they use as well as the data they take into account to produce these models.
The latter would allow IORPs to reduce the use of credit ratings provided by ECAIs, which would be in line with the objective of the CRA Regulation of reducing the overreliance on external credit ratings without requiring further costs for IORPs. In order to ensure that this requirement is fulfilled by the banking institutions, IORPs and financial institutions managing assets on their behalf should be able to notify such behaviour to competent authorities. In exchange, IORPs should apply due diligence aiming to understand the functioning of the model and its potential limitations when using the IRB models of their counterparties.

Moreover, and bearing in mind the limitations of using the IRB model of their counterparties, it is PensionsEurope’s strong opinion that IORPs should not be penalised for using external credit ratings from ECAIs instead of internal credit risk assessments. The credit quality requirements for internal assessments and for the use ECAIs should be equal.

On 12 May 2014 the Financial Stability Board (FSB) published a report on this matter where it recognised the particularities of pension funds in this regards and established that “the use of CRA ratings provides economies of scale in analysing credit on behalf of smaller and less sophisticated investors, and that implementation should differentiate according to size and sophistication of the firm”. Article 5 (a) (2) of Regulation (EU) No 462/2013 (CRA Regulation) also establishes that in this issue competent authorities must take into account the nature, scale and complexity of the activities of the different undertakings, which not only is not currently the case, but on the contrary, the current draft RTS would leave the weaker counterparty (the IORP) of the bilateral derivate transaction in a disadvantageous position.
Question 5. How would the introduction of concentration limits impact the management of collateral (please provide if possible quantitative information)? Are there arguments for exempting specific securities from concentration limits and how could negative effects be mitigated? What are the pros and cons of exempting securities issued by the governments or central banks of the same jurisdiction? Should proportionality requirements be introduced, if yes, how should these be calibrated to prevent liquidation issues under stressed market conditions?
PensionsEurope welcomes the large list of eligible collateral included on Article 1 LEC of the consultation paper. We believe that the types of permitted collateral should be broad enough to ensure that there is sufficient eligible collateral available to all market participants.

We also understand the concerns of the ESA’s when imposing concentration limits on initial and variation margins and their objective to promote collateral diversification. Indeed the aggregate collateral demand imposed by EMIR together with (punitive) margin policies of CCPs and clearing members may lead to collateral shortage for IORPs in times of markets stress, which is contrary to what EMIR intended and which will undermine the countercyclical role that IORPs traditionally play

However, we feel that the proposed concentration limits, as they currently stand, will cause a significant burden for IORPs and their asset managers, mainly for two reasons:

1. IORPs should not be subject to concentration limits on government bonds

Government bonds are a key investment tool for IORPs and their asset managers, which have a long tradition of using government bonds as VM in their bilateral derivative transactions.

Indeed, IORPs are obliged to be invested in a prudent and secure basis in order to be able to match their long-term liabilities . Although government bonds offer lower investment returns than other types of securities, they have the advantage that, subject to credit risk, they offer more predictable returns and allow IORPs to hold them for long periods of time with a view to matching all or part of their long-term liabilities. Government bonds (especially from the country (issuer) of residence of the IORP) are also used by IORPs as a tool to hedge against interest rate risk. Government bonds also have the advantage (as opposed to cash collateral) to not reduce the investment returns. Indeed, the economic ownership remains with the transferor, and also the transferee does not have to pay interest on collateral received.

As a result, if concentration limits are not removed, IORPs will have to engage in asset transformation (sell part of their portfolio of government bonds) in order to meet their collateral needs. This would entail lower returns / additional costs for IORPs and their beneficiaries.

Importantly, concentration limits on government bonds will also introduce unwanted valuation issues for derivatives. Indeed, when high grade government bonds are posted as collateral, the valuation of the accompanying derivatives is straight forward (based on EONIA discounting). The requirement of diversification in government bonds as collateral will lead to difficulties in valuation of derivatives and therefore disturb secondary markets. This is much to the disadvantage of the IORPs that rely on efficient markets in derivative instruments.

Furthermore, government bonds have also proven to be most liquid in terms of distressed market conditions. Indeed, in adverse market conditions market participants resort to government bonds seeking security. It is PensionsEurope view that concentration limits should only apply to those types of securities for which liquidity might suffer in times of distressed market conditions. This is certainly not the case of government bonds.

Additionally, concentration limits on government bonds could also have an impact on Member States’ financial needs. Indeed, not only IORPs (which already are large investors on government bonds) but also other market participants would be discouraged from purchasing debt securities. They will reduce their exposure to government bonds, which would inevitable have an impact on Member States’ public finances.

In view of the foregoing, PensionsEurope intensely supports completely removing the concentration limits on government bonds. However, bearing in mind this is a political issue, we propose and alternative approach: we suggest setting the maximum on the amount of collateral posted as a percentage of the total amount of government bonds issued by one single issuer: “The total amount of collateral collected from one counterparty shall not consist of more than 25% of the total amount issued and outstanding from one single issuer.” Or explained in another way: Bank A may only post 25% of the amount issued by the Spanish government to IORP B in Spanish government bonds. The 25% figure is small enough to ensure diversification of collateral and big enough for IORPs and their asset managers with big positions to not to be too much constrained. Also, IORPs that have spread their exposures amongst multiple counterparties will not be hindered by this rule.

2. In addition to removing the concentration limits on government bonds, concentration limits on all the other types of securities should not apply below a threshold of EUR 100 million in collateral

In addition to the concerns on the concentration limits on government bonds outlined above, in PensionsEurope we fear that dealing with the concentration limits in general will lead to increased operational burdens for IORPs and financial institutions managing assets on their behalf.

The proposed threshold of EUR 100 million would ensure proportionality between the objective of the proposed measures and the costs/burdens for IORPs of implementing it. Indeed, concentration limits will force IORPs to manage their collateral flows of IM and VM on a daily basis to ensure that the limits are respected. Additionally, they will need to engage in asset transformations to make sure that they have available a wide range of different types of assets to be exchanged as collateral.

For this reason, we consider reasonable to set a threshold of EUR 100 million in collateral as a minimum requirement to apply concentration limits. This will reduce the operational burden and complexity as well as risks and costs linked to managing collateral.

It should be noted that establishing a threshold of EUR 100 million will significantly reduce the burden for small and medium size IORPs (as stated on question 1 we expect them to be subject to IM and VM of this Draft RTS regardless of the thresholds of Article 1 FP) but it would not be enough for bigger IORPs and financial institutions managing assets on their behalf.
Question 6. How will market participants be able to ensure the fulfilment of all the conditions for the reuse of initial margins as required in the BCBS-IOSCO framework? Can the respondents identify which companies in the EU would require reuse or re-hypothecation of collateral as an essential component of their business models?
There are different views among our members about the re-use of collateral. Some are in favour of the practice and see liquidity squeeze and systemic crisis as a consequence of its interdiction; they believe that a better approach to this matter would be to determine the prohibition of re-use of collateral on a case-by-case basis. They prefer the more flexible approach adopted on this matter by IOSCO and the BCBS. They believe that it should be up to the counterparties involved in the derivative transactions to freely decide whether to allow the re-use of the collateral that they exchange. They fear that banning the re-use of collateral could make it more difficult to access financing and the latter would also be more expensive.

On the other hand, others see re-use of VM as incompatible with the segregation of their assets. In fact, VM is the value of an outstanding contract and in case of a default can be offset against outstanding positions. As the (liquidity) risk of re-use of VM lies completely with the VM receiver, VM re-use shall be allowed. It is already a current market practice for some IORPs and entities managing assets on their behalf to re-use VM collateral. VM needs to be re-used in cases when collateral needs to yield a certain return to be paid to the VM provider. Maintaining such a practice would not increase systemic risk, while restricting it would seriously impede on the pool of available collateral for IORPs which re-use VM.
Contact name
Bram VAN MALDEREN