Response to second Joint Consultation on draft RTS on risk-mitigation techniques for OTC-derivative contracts not cleared by a CCP
Go back
In our response to the first consultation, we requested a reconsideration of the proposal to impose variation margin on physically settled foreign exchange (FX) forward contracts.
We note that the second consultation paper continues to propose that the variation margin requirement applies to FX forwards. We believe that the March 2015 BCBS-IOSCO paper on margin requirements for non-centrally cleared derivatives provides the framework for the European Supervisory Authorities (ESAs) to apply additional flexibility to the variation margining of physically-settled FX forwards, and we also note that equivalent rules proposed by the regulators in the United States do not propose to impose variation margin on FX forwards. In our view, it is important that the domestic implementations of the BCBS-IOSCO framework are consistent with one another and do not dissuade pension funds from employing FX forwards to prudently manage foreign exchange risk.
For the above reasons, and for the reasons set out in our response to the first consultation, we believe the ESAs should reconsider this requirement.
We reiterate the points made in our first consultation response in response to this issue below.
The consultation includes a proposed obligation to pay variation margin on forward FX contracts, which are typically used by investors to either generate investment returns or hedge currency exposure. In this section we focus on the potential impact on investors aiming to hedge currency exposure.
We believe the proposal would increase the investment and operational costs of using forward FX contracts, and could therefore discourage investors from prudent currency risk management. This could leave pension schemes exposed to currency risk, which would typically have a greater impact on a pension scheme’s investment returns than the credit risk arising from uncollateralised positions. We therefore believe this proposal, which aims to reduce credit risk, could leave pension schemes exposed to a far greater risk in the form of unhedged currency exposure.
We explain how the proposal could lead to increased costs and operational complexity for currency hedges below:
• Increased costs
Pension schemes implement currency hedges in order to offset the foreign currency exposure that results from holding international assets. Therefore every currency hedge is fully backed by underlying assets, but the proposed requirement to make available an additional pool of collateral to support variation margin would increase the cost of holding foreign assets with currency protection.
Given the typical level of volatility for developed market currencies, we estimate that for every 100% invested in the asset, another 10% may be required to support variation margin. This means that a pension scheme would need to make €110 principal available for every €100 of fully-hedged overseas investment. This additional cost is particularly burdensome when the underlying investments are illiquid and cannot be offered as collateral: these might include foreign property, infrastructure or private equity investments.
• Increased operational complexity
Operational difficulties in dealing with variation margin for FX hedges are likely to result in some pension schemes either ceasing to protect themselves against foreign currency exposure or holding lower than optimal levels of overseas assets. Pension funds with smaller investment teams are likely to be most affected, as they have less capacity to absorb the additional operational requirements and extensive legal documentation that arise from managing margin and collateral, and may need to incur additional costs by employing a third party to manage the process.
A good manager of a currency hedge will already take responsibility for minimising credit risk by selecting the highest quality counterparties and monitoring credit risk over time. The prudent diversification of exposure across counterparties and the use of short-dated instruments are typically used to further reduce counterparty risk. Relative to other derivatives, forward FX contracts are among the simplest, representing something as simple as a delayed cash settlement. Using them can offset the impact of currency moves, which can be large and unpredictable and lie outside the control of a pension scheme. We therefore believe regulators should reconsider the proposed requirement.
Timing around VM rules
Further clarity is required on the obligations set out in Article 1 VM(3) and (5), in particular to take account of the settlement periods for non-cash securities.
We would expect a typical counterparty:
• to calculate its exposure at the end of day T.
• to call for variation margin during the day of T+1.
• to collect variation margin by the end of T+2 for high quality government securities settled on a DvP basis (or a longer settlement period for certain asset types).
On a plain reading, Article 1 VM(5) requires the collection of variation margin by T+1. While we believe it is possible to call for variation margin by T+1, the operational process required to source and transfer collateral means that it is not practicable to have transferred and/or have collected the variation margin by T+1; we note that this concern is expressed in the text for consultation. This requirement would impose risks and give rise to the concerns that the ESAs have already noted. It is our view that Article 1 VM(5) should be deleted and that the general rule in Article 1 VM(3) should apply to all transactions (subject to our comment on collection below).
We note that both Article 1 VM(3) and (5) impose an obligation to “collect”. It is not clear what ramifications there are for a collecting counterparty if (for example) an operational failure causes a settlement failure and the relevant timelines are breached; we would welcome further clarity on this aspect of the rule.
We set out below the unintended consequences these rules would present, firstly to all market participants, and secondly to pension schemes.
Issues with the current proposal relating to all market participants:
• We believe that the proposed issuer concentration limit, when applied to sovereign bonds issued by EU governments outside the eurozone, would have significant adverse implications specific to users of non-euro-denominated OTC derivatives and would introduce material new risks. This would include, for example, derivatives denominated in British sterling, the Polish zloty, the Hungarian forint and the Swedish krona.
• We also believe the proposed concentration limits, when combined with the proposed FX haircut set out in paragraphs 6 and 7 of Annex 2, will disproportionately affect entities based in the European Union but outside the eurozone.
• Derivative valuations are directly linked to the format of the underlying collateral that is used for margining the derivatives. It is not clear how derivative valuations could work when there is a collateral mix of 50% in one jurisdiction and 50% in another jurisdiction or cash. It is likely that end-users would receive valuations based off the worse valuation curve in this scenario.
• We are concerned the concentration rules are likely to result in rules that may not be possible to implement for transactions with buy-side clients with more than one asset manager and/or mandates. At best, if it can be implemented, we feel that it is likely to lead to significant delays in the settlement of margins, going against the overall objective of prudent risk management.
We currently post daily variation margin on a T+1 basis for most derivatives products. It is not clear how these complex concentration rules could be applied in situations when clients have more than one asset manager or more than one mandate. First, the calculations would need to be done across all asset managers, all mandates, and all groups of the buy-side client to determine if the initial margin collected goes above EUR 1bn to determine if the entity is in-scope for the concentration rules (Article 7 LEC 3(c)). Then, either simultaneously or immediately afterwards, a calculation must be done to calculate the total VM and IM across all asset managers, all mandates and across all groups, to determine how much of the total VM and IM goes above EUR 1bn (Article 7 LEC 2). Then some form of further communication needs to happen across all managers and/or the client to ensure that the 50% rule is met. It is difficult to envisage how this could be implemented at all in practice. If it can indeed be implemented somehow, it is likely to lead to significant delays to the settlement of margins, which we feel would go against the objective of prudent risk management.
Note that given that banks are typically on the other side of the transaction with buy-side clients, this issue would impact not just the buy-side clients but also banks, who would be required to post collateral in the relevant format so that the buy-side client could meet the 50% requirement rules on collateral that it collects.
• We are concerned the practical and timing implications of the complex concentration rules need further thought to ensure that it does not create any unintended consequences: Any end-user subject to the initial margin rules is likely to only become in-scope for the initial margin rules from 2020. This is because it is unlikely that an end user would have EUR 750 billion of non-cleared derivatives, but it is possible that an end-user could have EUR 8 billion of non-cleared derivatives. Based on this, an end-user could only become in-scope for any concentration rules from 2020 (based on Article 7 LEC 3(c)).
Once an end-user is in-scope for the concentration rules, it would apply to not only IM but also VM. However, it is not clear if this would apply to previously posted VM or just VM posted from the date when the entity becomes in-scope for the concentration rules. Either way we are concerned that there are significant unintended consequences and hurdles to overcome.
If the rules are intended to be applied to all VM, including previously posted VM, this would create a “big bang” date and it would likely create significant disruption. It’s not clear how market participants could readjust the collateral mix that has been already posted. Furthermore, readjusting the collateral mix for derivatives contracts would impact derivative valuations for previously traded derivative contracts. This is because derivative valuations are directly linked to the underlying collateral that is posted. Counterparty banks would likely try to pass on the cost of the change in derivatives valuation to end-users. End users could therefore incur additional costs and this is not something that they would be able to quantify at the time of entry into the derivative contract.
If the rules are intended to capture only the VM posted from the date when the entity becomes subject to the concentration rules, it is not clear how this could work in practice. It is difficult to envisage how we could implement two different VM calculations taking into account different time periods.
Note that a transaction between a bank and an end-user would become subject to the concentration rules at the point when the end-user becomes an in-scope entity, as the bank is already likely to be a systemically important institution. Therefore the timing issue described above in relation to when the concentration rules would come into effect for a transaction would be an issue not just for the end-users but also for the bank on the other side of the transaction.
Issues with the current proposal relating to pension schemes (or pension scheme arrangements as defined in EMIR level 1):
• Pension schemes are not systemically important institutions: While the current ESAs’ proposal tries to bring into scope only systemically important institutions, Article 7 LEC 3(c) brings into scope entities that are not systemically important institutions. This Article brings into scope any counterparty that collects initial margin in excess of EUR 1 billion which would include large pension schemes with over EUR 8 billion gross notional of non-cleared derivatives positions which would be subject to the initial margin rules.
Pension schemes are subject to a requirement to prudently manage risk. As part of this, pension schemes use derivatives to prudently manage financial solvency and reduce the risk that the pension schemes may become financially insolvent. Pension schemes are not systemically important institutions and should not be treated as such. Pension schemes are asset-rich, do not take large amounts of leverage and are focused on managing their financial solvency to pay current and future pensioners’ retirement incomes.
• Concentration rules increase risk when applied to pension schemes in the European Union but not within the eurozone: The impact of these concentration rules would be significant for pension schemes within the EU but not within the eurozone, such as UK pension schemes. UK pension schemes hold a significant amount of sterling UK government bonds (gilts). These gilts provide certain interest rate and inflation risk protection against the liabilities of UK pension schemes and therefore play an important role for UK pension schemes managing their financial solvency. Other EU government bonds would not provide this protection against these liability risks as EU government bonds are not denominated in sterling and are not linked to UK interest or inflation rates. A UK pension scheme required to comply with the concentration limits would likely have to convert some of its gilts into non-UK bonds, introducing not only currency risk but also increasing the financial solvency risk to the pension scheme. The net effect of the concentration rules on large pension schemes would therefore be to increase risk, rather than reduce risk, for pension beneficiaries, the sponsor corporate responsible for the employee pension schemes, and the financial system overall.
• Concentration rules would force pension schemes to either take more risk or hold more cash, going against EMIR level 1 policymakers’ objective: In order to comply with the concentration limit requirement without introducing currency risks, a UK pension scheme would need to hold cash instead of gilts. Holding cash would introduce significant costs for pensioners. EMIR level 1 text already recognises that pension schemes do not hold cash and must not be forced to hold cash. The temporary relief for pension schemes from the central clearing obligation is a reflection of pension schemes’ inability to post cash for variation margin required in central clearing. We feel that these concentration rules would effectively force large UK pension schemes to either a) take unwarranted risk and go against their principle of managing risk prudently, or b) hold cash incurring significant costs for pensioners, going against the EMIR level 1 policymakers’ objective.
We therefore believe that the concentration limits, even in their modified form, continue to be of concern and we believe that they should be removed. We would be happy to engage on ways to make this work if it cannot be removed entirely (e.g. applying concentration rules only for initial margin and only between banks).
Respondents are invited to comment on the proposal in this section concerning the timing of calculation, call and delivery of initial and variation margins.
Treatment for FXIn our response to the first consultation, we requested a reconsideration of the proposal to impose variation margin on physically settled foreign exchange (FX) forward contracts.
We note that the second consultation paper continues to propose that the variation margin requirement applies to FX forwards. We believe that the March 2015 BCBS-IOSCO paper on margin requirements for non-centrally cleared derivatives provides the framework for the European Supervisory Authorities (ESAs) to apply additional flexibility to the variation margining of physically-settled FX forwards, and we also note that equivalent rules proposed by the regulators in the United States do not propose to impose variation margin on FX forwards. In our view, it is important that the domestic implementations of the BCBS-IOSCO framework are consistent with one another and do not dissuade pension funds from employing FX forwards to prudently manage foreign exchange risk.
For the above reasons, and for the reasons set out in our response to the first consultation, we believe the ESAs should reconsider this requirement.
We reiterate the points made in our first consultation response in response to this issue below.
The consultation includes a proposed obligation to pay variation margin on forward FX contracts, which are typically used by investors to either generate investment returns or hedge currency exposure. In this section we focus on the potential impact on investors aiming to hedge currency exposure.
We believe the proposal would increase the investment and operational costs of using forward FX contracts, and could therefore discourage investors from prudent currency risk management. This could leave pension schemes exposed to currency risk, which would typically have a greater impact on a pension scheme’s investment returns than the credit risk arising from uncollateralised positions. We therefore believe this proposal, which aims to reduce credit risk, could leave pension schemes exposed to a far greater risk in the form of unhedged currency exposure.
We explain how the proposal could lead to increased costs and operational complexity for currency hedges below:
• Increased costs
Pension schemes implement currency hedges in order to offset the foreign currency exposure that results from holding international assets. Therefore every currency hedge is fully backed by underlying assets, but the proposed requirement to make available an additional pool of collateral to support variation margin would increase the cost of holding foreign assets with currency protection.
Given the typical level of volatility for developed market currencies, we estimate that for every 100% invested in the asset, another 10% may be required to support variation margin. This means that a pension scheme would need to make €110 principal available for every €100 of fully-hedged overseas investment. This additional cost is particularly burdensome when the underlying investments are illiquid and cannot be offered as collateral: these might include foreign property, infrastructure or private equity investments.
• Increased operational complexity
Operational difficulties in dealing with variation margin for FX hedges are likely to result in some pension schemes either ceasing to protect themselves against foreign currency exposure or holding lower than optimal levels of overseas assets. Pension funds with smaller investment teams are likely to be most affected, as they have less capacity to absorb the additional operational requirements and extensive legal documentation that arise from managing margin and collateral, and may need to incur additional costs by employing a third party to manage the process.
A good manager of a currency hedge will already take responsibility for minimising credit risk by selecting the highest quality counterparties and monitoring credit risk over time. The prudent diversification of exposure across counterparties and the use of short-dated instruments are typically used to further reduce counterparty risk. Relative to other derivatives, forward FX contracts are among the simplest, representing something as simple as a delayed cash settlement. Using them can offset the impact of currency moves, which can be large and unpredictable and lie outside the control of a pension scheme. We therefore believe regulators should reconsider the proposed requirement.
Timing around VM rules
Further clarity is required on the obligations set out in Article 1 VM(3) and (5), in particular to take account of the settlement periods for non-cash securities.
We would expect a typical counterparty:
• to calculate its exposure at the end of day T.
• to call for variation margin during the day of T+1.
• to collect variation margin by the end of T+2 for high quality government securities settled on a DvP basis (or a longer settlement period for certain asset types).
On a plain reading, Article 1 VM(5) requires the collection of variation margin by T+1. While we believe it is possible to call for variation margin by T+1, the operational process required to source and transfer collateral means that it is not practicable to have transferred and/or have collected the variation margin by T+1; we note that this concern is expressed in the text for consultation. This requirement would impose risks and give rise to the concerns that the ESAs have already noted. It is our view that Article 1 VM(5) should be deleted and that the general rule in Article 1 VM(3) should apply to all transactions (subject to our comment on collection below).
We note that both Article 1 VM(3) and (5) impose an obligation to “collect”. It is not clear what ramifications there are for a collecting counterparty if (for example) an operational failure causes a settlement failure and the relevant timelines are breached; we would welcome further clarity on this aspect of the rule.
Respondents are invited to comment on whether the requirements of this section concerning the concentration limits address the concerns expressed on the previous proposal.
While we appreciate the ESAs’ efforts to reduce the scope of the concentration rules, we do not support the concentration rules and believe a reconsideration is needed. We feel that the concentration rules go over and above the international standards set by BCBS-IOSCO causing potential for inconsistency with rules set by other international regulators. Separately, we believe the current proposal would introduce significant unintended consequences and practical issues.We set out below the unintended consequences these rules would present, firstly to all market participants, and secondly to pension schemes.
Issues with the current proposal relating to all market participants:
• We believe that the proposed issuer concentration limit, when applied to sovereign bonds issued by EU governments outside the eurozone, would have significant adverse implications specific to users of non-euro-denominated OTC derivatives and would introduce material new risks. This would include, for example, derivatives denominated in British sterling, the Polish zloty, the Hungarian forint and the Swedish krona.
• We also believe the proposed concentration limits, when combined with the proposed FX haircut set out in paragraphs 6 and 7 of Annex 2, will disproportionately affect entities based in the European Union but outside the eurozone.
• Derivative valuations are directly linked to the format of the underlying collateral that is used for margining the derivatives. It is not clear how derivative valuations could work when there is a collateral mix of 50% in one jurisdiction and 50% in another jurisdiction or cash. It is likely that end-users would receive valuations based off the worse valuation curve in this scenario.
• We are concerned the concentration rules are likely to result in rules that may not be possible to implement for transactions with buy-side clients with more than one asset manager and/or mandates. At best, if it can be implemented, we feel that it is likely to lead to significant delays in the settlement of margins, going against the overall objective of prudent risk management.
We currently post daily variation margin on a T+1 basis for most derivatives products. It is not clear how these complex concentration rules could be applied in situations when clients have more than one asset manager or more than one mandate. First, the calculations would need to be done across all asset managers, all mandates, and all groups of the buy-side client to determine if the initial margin collected goes above EUR 1bn to determine if the entity is in-scope for the concentration rules (Article 7 LEC 3(c)). Then, either simultaneously or immediately afterwards, a calculation must be done to calculate the total VM and IM across all asset managers, all mandates and across all groups, to determine how much of the total VM and IM goes above EUR 1bn (Article 7 LEC 2). Then some form of further communication needs to happen across all managers and/or the client to ensure that the 50% rule is met. It is difficult to envisage how this could be implemented at all in practice. If it can indeed be implemented somehow, it is likely to lead to significant delays to the settlement of margins, which we feel would go against the objective of prudent risk management.
Note that given that banks are typically on the other side of the transaction with buy-side clients, this issue would impact not just the buy-side clients but also banks, who would be required to post collateral in the relevant format so that the buy-side client could meet the 50% requirement rules on collateral that it collects.
• We are concerned the practical and timing implications of the complex concentration rules need further thought to ensure that it does not create any unintended consequences: Any end-user subject to the initial margin rules is likely to only become in-scope for the initial margin rules from 2020. This is because it is unlikely that an end user would have EUR 750 billion of non-cleared derivatives, but it is possible that an end-user could have EUR 8 billion of non-cleared derivatives. Based on this, an end-user could only become in-scope for any concentration rules from 2020 (based on Article 7 LEC 3(c)).
Once an end-user is in-scope for the concentration rules, it would apply to not only IM but also VM. However, it is not clear if this would apply to previously posted VM or just VM posted from the date when the entity becomes in-scope for the concentration rules. Either way we are concerned that there are significant unintended consequences and hurdles to overcome.
If the rules are intended to be applied to all VM, including previously posted VM, this would create a “big bang” date and it would likely create significant disruption. It’s not clear how market participants could readjust the collateral mix that has been already posted. Furthermore, readjusting the collateral mix for derivatives contracts would impact derivative valuations for previously traded derivative contracts. This is because derivative valuations are directly linked to the underlying collateral that is posted. Counterparty banks would likely try to pass on the cost of the change in derivatives valuation to end-users. End users could therefore incur additional costs and this is not something that they would be able to quantify at the time of entry into the derivative contract.
If the rules are intended to capture only the VM posted from the date when the entity becomes subject to the concentration rules, it is not clear how this could work in practice. It is difficult to envisage how we could implement two different VM calculations taking into account different time periods.
Note that a transaction between a bank and an end-user would become subject to the concentration rules at the point when the end-user becomes an in-scope entity, as the bank is already likely to be a systemically important institution. Therefore the timing issue described above in relation to when the concentration rules would come into effect for a transaction would be an issue not just for the end-users but also for the bank on the other side of the transaction.
Issues with the current proposal relating to pension schemes (or pension scheme arrangements as defined in EMIR level 1):
• Pension schemes are not systemically important institutions: While the current ESAs’ proposal tries to bring into scope only systemically important institutions, Article 7 LEC 3(c) brings into scope entities that are not systemically important institutions. This Article brings into scope any counterparty that collects initial margin in excess of EUR 1 billion which would include large pension schemes with over EUR 8 billion gross notional of non-cleared derivatives positions which would be subject to the initial margin rules.
Pension schemes are subject to a requirement to prudently manage risk. As part of this, pension schemes use derivatives to prudently manage financial solvency and reduce the risk that the pension schemes may become financially insolvent. Pension schemes are not systemically important institutions and should not be treated as such. Pension schemes are asset-rich, do not take large amounts of leverage and are focused on managing their financial solvency to pay current and future pensioners’ retirement incomes.
• Concentration rules increase risk when applied to pension schemes in the European Union but not within the eurozone: The impact of these concentration rules would be significant for pension schemes within the EU but not within the eurozone, such as UK pension schemes. UK pension schemes hold a significant amount of sterling UK government bonds (gilts). These gilts provide certain interest rate and inflation risk protection against the liabilities of UK pension schemes and therefore play an important role for UK pension schemes managing their financial solvency. Other EU government bonds would not provide this protection against these liability risks as EU government bonds are not denominated in sterling and are not linked to UK interest or inflation rates. A UK pension scheme required to comply with the concentration limits would likely have to convert some of its gilts into non-UK bonds, introducing not only currency risk but also increasing the financial solvency risk to the pension scheme. The net effect of the concentration rules on large pension schemes would therefore be to increase risk, rather than reduce risk, for pension beneficiaries, the sponsor corporate responsible for the employee pension schemes, and the financial system overall.
• Concentration rules would force pension schemes to either take more risk or hold more cash, going against EMIR level 1 policymakers’ objective: In order to comply with the concentration limit requirement without introducing currency risks, a UK pension scheme would need to hold cash instead of gilts. Holding cash would introduce significant costs for pensioners. EMIR level 1 text already recognises that pension schemes do not hold cash and must not be forced to hold cash. The temporary relief for pension schemes from the central clearing obligation is a reflection of pension schemes’ inability to post cash for variation margin required in central clearing. We feel that these concentration rules would effectively force large UK pension schemes to either a) take unwarranted risk and go against their principle of managing risk prudently, or b) hold cash incurring significant costs for pensioners, going against the EMIR level 1 policymakers’ objective.
We therefore believe that the concentration limits, even in their modified form, continue to be of concern and we believe that they should be removed. We would be happy to engage on ways to make this work if it cannot be removed entirely (e.g. applying concentration rules only for initial margin and only between banks).