European Systemic Risk Board, Head of the ESRB Secretariat
Paul P. Andrews
International Organization of Securities Commissions, Secretary General
Benoît de Juvigny
Autorité des Marchés Financiers, Secretary General
Malta Financial Services Authority, Director General
BlackRock, EMEA Head of Government Relations and Public Policy
Federated Investors (UK) LLP, Senior Vice President, Managing Director and Chief Investment Officer, Cash
DekaBank Deutsche Girozentrale, Chief Executive Officer
Detailed SummaryThe Chair declared that there would be three themes for the discussion: the first would be about existing instruments for mitigating risks associated with asset management, and the use that can be made of them. The second would be about new approaches to preventing risks, and the third would be about money market funds and the regulation that has recently been adopted for those funds in the EU.
1. Existing practices and tools for mitigating vulnerabilities from asset management activities
1.1. Risk management practices currently used by asset managers
The Chair emphasised the importance of liquidity and leverage risks and asked industry representatives on the panel how risks are currently tackled within management companies, and what the main challenges are that they face.
An industry representative replied that although fund liquidity and leverage risks are absolutely fundamental, the issue of investment risk is broader, and multi faceted: it includes market risk, credit risk, and market liquidity risk. For managing these risks properly, it is necessary to take a holistic approach.
In the speaker’s firm, portfolio managers (PMs) are those primarily responsible for investment risk. There is no centralised investment officer (CIO) who takes asset allocation decisions or tells people how to buy or sell individual funds. For investment funds, it is essential that management should be tailored to the characteristics of the individual fund: the investment strategy, redemption terms, and the tools that act as backup in the event that issues arise, need to be appropriate to the client type, the distribution mechanism and the asset class. The PMs have to incorporate all of this into how they construct the portfolio, and will build in layers of liquidity, with minimum thresholds for the most liquid asset classes and maximum thresholds for the least liquid classes. They will then layer on derivative overlays, ETFs and other factors to add in extra liquidity.
The PMs are supported by independent specialised risk managers. Each has an area of expertise and is responsible for the risk management of certain funds within that area. They monitor and stress test the portfolios on a daily basis, provide PMs with analyses, and ensure that best practice is incorporated into each of the different funds.
The speaker’s institution also has a trading and liquidity function, which is separate from the PMs, and plays a very important role in uncovering latent liquidity and aggregating liquidity, which is important in managing investment risk. This centralised liquidity risk function examines any outliers, and the whole process aims to put in place qualitative and quantitative balances and checks and to flag issues where the risk manager and the fund manager might disagree.
There is also an internal advisory group which works with the PMs to identify certain events or market scenarios, such as the various potential outcomes of the French and German elections, and analyse how they will affect funds, so that the PMs can decide whether or not they need to adjust their portfolios. Operational readiness is also vital if an issue arises whereby it becomes necessary to implement backup plans, in the form of suspensions or gates.
The Chair asked another industry representative to give a perspective on how risks are managed and also how to avoid the propagation of risk between the asset management industry and other parts of the financial sector.
The industry speaker replied that their firm is an integrated group, including banking and asset management. This gives rise to potential conflicts of interest which are managed by having transparent fees and best execution rules in particular. The group’s dominant topic at present is the implementation of new mandatory investor protection rules, i.e. MiFID II, which involves a significant workload and high costs.
Another issue is that all market players have to invest significantly in digitalisation. A decoupling of the value chain, which occurred in the US some years ago, is happening in Europe now. Outsourcing is a priority for the speaker’s firm, which in many cases means partnering with third party institutions, and addressing the related risks is a top priority. Asset managers have to ask themselves what their unique selling proposition is, what they will continue to do in-house and what will be outsourced, who they will be partnering with and what the related operational risks are, and how they handle these operational risks. An evaluation needs to be made in particular regarding the stability of IT providers.
Partnering in terms of market making is a related issue, but is not wholly about outsourcing. Currently, European asset managers based in Europe have one key topic, which is Brexit, and it is not yet clear whether there will be a transition period and whether market making will continue to be easily available to companies on the continent.
1.2. Strengths and weaknesses of EU fund frameworks regarding the identification and mitigation of asset management vulnerabilities
The Chair reminded the audience that Europe has two major pieces of asset management legislation: UCITS and AIFMD and asked the panel for views on the positive aspects and possible shortcomings of these frameworks and whether they are sufficient for addressing asset management vulnerabilities.
A regulator stated that the positive aspects of UCITS and AIFMD are that there is now quite a comprehensive and extensive framework for investment funds, which also takes into account the offshoots that have come out of the UCITS and AIFMD directives, such as the MMF regulations and ELTIFs and which cater for the specificities of certain funds in the market. The current regulatory landscape is completely different from the one that existed before the crisis, and has developed on the basis of the lessons learned during the crisis. This is already a good step forward, and something to build on when addressing any potentially systemic risks arising in this sector.
The directives already have a number of tools that address the vulnerabilities that have been identified as prevalent in the asset management sector. There is also a great deal of reporting being done already, which can be further used both to enable supervisors to address supervision at the local or micro level, and also for performing aggregated analysis at a higher, macro level.
The greatest weakness of these directives is their consistency at different levels. There is a need for further convergence at the legislative level, because certain areas are addressed by both directives in different ways, such as stress testing, reporting to investors and liquidity management. There is also a need for consistency at the level of the licensed entities. For example, how individual fund managers carry out stress testing is at the managers’ discretion, which implies different approaches and interpretations, and this may impact the effectiveness of systemic supervision, both at the micro and macro levels. The final area in which there is a need for consistency concerns the approaches taken at the local and supervisory levels, because different approaches may be applied by regulators to different areas in respect of these two directives, especially where asset managers may be managing more than one type of fund and therefore falling under both directives.
Another regulator agreed with previous speakers that liquidity or leverage risks are the main topics that need to be dealt with regarding asset management vulnerabilities. The UCITS and AIFMD directives already include some specifications regarding liquidity and leverage risks, which are very interesting. For the time being, the question may be more about the implementation of these existing provisions, rather than creating any new tools. There are also many liquidity management tools in place at the domestic level, notably in France, including swing pricing, redemption fees, and suspensions. Last year, the possibility of gates was partially introduced as well, which is the reason why the French law was changed to give the industry a greater possibility for using them. Each country has to do the same kind of exercise to ensure that it has the most appropriate tools.
The Chair noted that article 25 of the AIFMD directive, which allows the addressing of excessive leverage in the European Union has not been used yet, and asked why this was the case and whether there was any need to make use of this instrument.
The regulator replied that there may be a need to use this rule in the future but confirmed that for the time being, it has not been employed. It is a good example of rules that already exist in European legislation. There is already some kind of definition of leverage in EU fund regulation, with a leverage cap in the UCITS regulation and article 25 of AIFMD provides the possibility for national competent authorities, in exceptional circumstances, to put a leverage cap in place. Over the last few years, the regulator has not had any real concerns regarding the leverage of funds in his jurisdiction or the need to impose a leverage cap. Although there are a few funds with significant leverage, these are rather small and limited amounts are involved, and there is full transparency regarding their leverage; therefore, investors know what they are investing in.
However, in the future, the situation might change. This is why national authorities are working with ESMA to identify examples of situations in which they could use this kind of leverage cap. Article 25 also involves some coordination at EU level, with the possibility for ESMA to give a recommendation to other national competent authorities in terms of leverage cap. Although coordination by ESMA could be useful, the regulator would prefer national competent authorities to keep this power, because having a domestic assessment of risks is necessary. The regulator noted that in the previous month, there had been questions regarding leverage in closed-ended commercial real estate in France, but the assessment conducted by the AMF showed that leverage was not really a concern and did not require specific action.
ETFs raise different issues because they could involve an ‘illusion of liquidity’, since underlying assets may not have the same kind of liquidity as the ETF itself, and this mismatch could be a concern. This is more applicable to the new types of ETFs put on the market, the regulator noted; the numbers of ETFs are growing, and the new ones have much less equity. The AMF recently published a specific study on the French market of ETFs. A ‘circuit breaker’ which is mandatory in France was considered as a useful tool for limiting problems; secondly, the AMF identified that in many market circumstances, ETFs could have a counter cyclical effect. This ‘counter intuitive’ element in the liquidity of ETFs is interesting, but at the same time, the mismatch between ETFs and underlying assets should be worked on more at the international level.
1.3. Approach at the global level regarding vulnerabilities from asset management activities
The Chair noted that IOSCO has been given a mandate to address investment fund vulnerabilities and asked an official to comment on the further steps of this global initiative.
The official replied that IOSCO is in the early stages of doing this work. IOSCO is focused on liquidity and leverage, and will have a paper for consultation out in two to four weeks, which will define its approach and the possible need for additional guidance regarding liquidity issues. IOSCO then aims to issue a consultation paper on leverage by the end of 2017, for implementation during 2018. As of the present time, IOSCO’s work agenda covers about 18 months. In two to three years’ time, IOSCO would like the new guidance that it has issued on liquidity to have been implemented, with asset managers taking all of these standards, recommendations and guidance to heart.
On the leverage side, the issue is more complex, because IOSCO has been asked to come up with a simple, transparent and consistent measure of leverage, and that is very difficult to achieve. The official does not necessarily think that IOSCO will come up with one measure, but perhaps with a series of these, which could then be implemented a year or two after finalisation.
Regarding the difference in approach required between the banking and asset management industries in terms of ‘systemic ness’ , the official emphasised that it is important to bear in mind what asset managers do: this is effectively an agency business, different from the activities of a bank or an insurance company. There can be a debate about where the risks are, and whether they are systemic or not, but this issue needs to be looked at in the context of the activities that an asset manager performs, rather than the entity itself. Two years ago, the IOSCO board told the FSB that this was not an issue for the FSB to be involved with, but rather for IOSCO to look into, and IOSCO has been doing this ever since. To the extent that anything is systemic, it will arise from trading activities, and not from the entity itself. IOSCO has been somewhat successful in putting this argument forward.
On liquidity related issues, IOSCO is focusing on asset managers as amplifiers of risk, where the main question is that of liquidity and liquidity mismatch. Ultimately, if investors cannot get their money back when they want to, this could present problems. IOSCO is trying to address this via guidance on liquidity, to avoid ‘run concern or herding behaviour’.
The micro and macro issues are related to this; looking at what an asset manager does requires taking a ‘broad brush’ perspective, as although it is something of a micro picture, that does not take into account idiosyncratic risks arising from either a particular firm or a small segment of the industry. The micro level has to be examined, but the firm level also needs to be looked at; these are ‘two sides of the same coin’. Ultimately, it is activities that matter, and that should be the main focus of attention.
A regulator agreed that It is important to take into account that the potential systemic risks in the asset management industry are quite different from the vulnerabilities in, for example, the banking sector; from a regulatory and monetary perspective.
1.4. Possible implications of Brexit
The Chair asked whether any specific investment fund risks are related to Brexit, either pre or post Brexit. An industry representative replied that it would be too early to express an opinion on what the situation will be in two years’ time. It appears likely that the UK will adopt an approach consistent with the EU to fund governance and maintain regulations that will ensure UK products are as safe and well controlled as any others; whatever these products are in the UK, they will be at least equivalent to the EU’s UCITS and AIFMD products.
If the UK and the EU fail to reach agreement on fund distribution, the principal losers will be investors in both the UK and in the EU. Firms will be forced to duplicate efforts, which will increase costs, and in many instances, investment options in each jurisdiction will be reduced. Although it may be difficult to reach agreement on these points, this is an issue that is critical for the continued development of European capital markets.
2. Possible need for additional tools and actions to address vulnerabilities from asset management activities
2.1. Possible need for additional fund rules at the EU level
An industry representative stated that there is not that much space in the fund sector that is currently unregulated. What is needed for addressing remaining risks related to asset management activities is not additional rules, but an appropriate implementation of existing rules, which requires having an increasing amount of empirical data available to analyse the impact of the rules that already exist.
One remaining issue might be concentration risk, if there is an enormous inflow into specific UCITS structures. The investment process is key for appropriately managing risks, as mentioned by a previous speaker. PMs and salespeople have a responsibility in their own markets: they can limit inflows for example in the fintech industry; this is a question of hygiene and professional ethics. Moreover there are safeguards on the investor side: institutional investors are themselves regulated and retail investors benefit from investor protection measures.
Considering these different types of investors, attempting to create additional rules regarding how institutional investors (insurance companies, pension funds and foundations…) should invest in order to limit concentration risk would not mesh with the principal agent business of asset managers. These investors are very sophisticated; they reflect their strategic asset allocation in certain investment styles, and asset managers form part of the implementation management and innovation process. Regarding retail investors, many rules, including investor protection rules notably with MiFID II already exist. Asset managers need to adhere to these rules, translate them into an advisory process in the documentation, get approval into the prospectus, and implement available rules.
2.2. Prospects of macro-prudential tools
The Chair moved the conversation to the use of macro-prudential tools, noting that the European Commission has launched a number of consultations, including on the powers of the ESAs and the ESRB and asked the panel whether there is a need to create a stronger capacity at EU level to cope with vulnerabilities in the asset management industry and to develop macro-prudential tools in the asset management sector.
A regulator believed that this is necessary. The first reason for this is that, although the systemic risk debate for asset management was not conclusive, the debate has moved on as the result of the issuance of recommendations by the FSB. Also, structural changes have been seen in the asset management industry in past years, and the regulator anticipates more in the future.
However, an increased capacity to cope with vulnerabilities is not meant to lead to increased regulation, but more to be able to monitor potential threats, and to be able to more accurately differentiate real threats from perceived ones. This is not easy, and there is the potential for capacity building at the ESA level: the ESAs can contribute to achieving more convergence and consistency in monitoring, measurement and the application of the tools available under the directives present at the micro level. In addition, if the goal is to monitor and assess the potential prevalent systemic risks in the asset management industry, this will require an adjustment in the supervisory approach in this sector, from one based primarily on investor protection to one that is based on, or includes, prudential supervision. This prudential supervision should in the future be based more on business models and a proper identification of risks, to ensure that the supervisory approach is proportionate and takes into account the specific nature of the asset management industry.
The regulator added, however, that this cannot cover the whole extent of the capacity of prudential supervision; it also needs to be supported at the macro level. Here, it will be important for there to be cross fertilisation of the expertise of the ESAs in this sector and the expertise of the ESRB people in the area of financial stability.
An industry representative noted that they looked favourably on macroprudential measures applied to non banks, but the diversity of asset owners and investment strategies, such as investment funds, sovereign wealth funds, family offices, pension funds and insurance companies needs to be considered. This would allow regulators to have better and timelier information to intervene at the individual fund level. Fund managers should have the broadest possible toolkit, so that the boards can take the precise actions needed in the particular circumstance relevant for the particular type of fund.
The best way of addressing systemic risk is indeed to consider the diversity of investors. Although there is good information in certain areas, primarily funds, there would also need to be similar information on separate accounts, all types of asset owners, sovereign wealth funds, family offices, pension funds and insurance companies, in order to better understand how the system works before considering macroprudential measures.
What should be avoided are macroprudential measures applied only across fund sectors that may undermine that diversity; this would have the effect of re aggregating these fund categories, which could cause pro cyclical behaviour from the end investors. The speaker’s ‘worst nightmare’ would be system wide stress tests applied only across the fund sector or part of it, without due differentiation between market risk – i.e. how markets function – and systemic risk, and without good information being available for all of the different market participants. Such a scenario could create macroprudential measures, such as ‘counter cyclical’ haircutting on margins, or policies that hurt the best interests of the client such as mandatory liquidity buffers that would prevent asset managers from applying risk protocols that ultimately benefit the end investor.
It needs to be borne in mind that the end investor is not obliged to invest in markets, and certainly not obliged to invest in mutual funds. If they think that their interests are being harmed, they can retreat, and as such, these sorts of macro-prudential policies can actually cause pro cyclical behaviour. There is a need for consistent data that allows people to intervene earlier, not just in the area of liquidity risk mismatch, but also on different measures of leverage and potential loss measures. This is fundamental, because many issues arise from excessive leverage in certain funds where there are no appropriate funding terms. With data on that, it can be decided whether leverage limits are necessary or not for certain types of funds.
An official noted that it could be argued that the measures that are being discussed are ex post facto, or ‘closing the door after the horse is out of the barn’. Regulation is typically behind the curve; the official does not expect that it will ever be ahead of the curve, but the goal of regulators is to at least keep pace with the curve. As mentioned previously, a consultation paper will be issued in a number of weeks on liquidity, which will provide guidance concerning issues such as the design of funds, operational risk and dealing frequency, in the sense of trying to prevent the liquidity mismatches that market participants are fearful of. As such, all regulation is not ex post facto.
2.3. Optimizing the use of stress tests
The Chair noted that there are now some requirements for investment funds to have stress tests in the EU, and asked the panel whether they consider these to be sufficient or not.
A regulator confirmed that stress test obligations are included in UCITS and AIFMD regulations, but these are not an easy subject. Traditionally, asset management regulators were more involved in the stress testing of assets in the market and in the related questions of asset liability and liquidity of funds. There are several thousand funds and 600 asset management companies in France for example, and all of these are obliged to carry out stress tests individually, but the framework of these tests is not regulated.
Enquiries conducted over the course of the last year in France showed many interesting practices regarding stress tests, but the objective is not to make precise recommendations. The speaker stated that they understand the industry’s perspective that putting a precise regulatory tool in place for stress testing all or part of the asset management sector would be impossible, because each market is different and each fund is different. They are against stress tests being completely centralised; some discretion needs to be offered to asset managers, although not too much, and it will be interesting to see how this kind of question will evolve over the coming years at the European or IOSCO level.
An official considered that stress testing should be both “from the bottom”, which is already occurring at the fund level, and “from the top”, although these latter tests need to be conducted in a meaningful and smart way.
The way in which top-down stress testing can be performed is to focus on similar funds. There are many different types of funds; having one global stress test that would apply to all types of funds would be meaningless, and would not produce results that would be useful to anyone. However, a stress test for funds that deal in oil and gas futures for example, or something similar, could be more meaningful. There needs to be a balance between the regulatory and the industry sides, and the industry will have to accept that there will probably be some sort of industry wide stress testing, done in a smart and meaningful way. The solution has to be both bottom up and top down.
An industry representative suggested that industry and regulators both need to appreciate all of the implications of portfolio composition, and should look at a wide range of different stresses on portfolios. The asset management industry cannot be looked at in isolation because it only represents a small percentage of the whole asset universe. Bottom up stress testing of each individual investment fund is critical, and for the first time with the MMF regulation, not only will local regulators have information on money market fund holdings, but so will ESMA.
There will be complete transparency on portfolio holdings of all European money market funds, which will permit regulators to monitor market activity in real time and take the necessary action on any risks that they perceive. This transparency should, and could, be used to ensure that funds are holding sufficient liquidity; that there is sufficient liquidity in the marketplace; and that firms are not taking inappropriate risks to obtain higher yields which would be contrary to the objective of a MMF, or that of any other UCITS type fund.
Another industry representative suggested that a process such as the Supervisory Review and Evaluation Process (SREP) used for banks could be useful in the asset management sector. Indeed, when the ECB looks into stress tests, they focus very much on internal stress testing, whereas the top down EBA stress test for example is about consistent data. Focusing on internal stress testing rather than only on a top down approach seems the best approach since it allows one to assess from the outside whether the internal stress testing performed by the investment fund is appropriate and fits in with the business model, asset allocation and investment style. Indeed the stress testing of the risk appetite and processes of investment funds is a ‘very individual’ topic. Running a commercial real estate debt fund portfolio is completely different from running a highly liquid equities portfolio for example. Any solution will need to be tailored in terms of internal stress test methods, bearing in mind that controlling risk and asset management is essential for the industry’s success.
The Chair suggested that one possible solution for macro-prudential assessments would be to analyse the basis of information that already exists and notably the information about EMIR transactions, distinguishing them by typology. EMIR data could be used for identifying risks and possibly for stress tests.
3. Prospects of the new European Money Market Fund regulation (MMFR)
3.1. New MMF products proposed
The Chair invited the panel to provide views on the recently adopted MMFR, the new products proposed by the regulation, and how this may help with risk mitigation.
An industry representative noted that the European Parliament has finalised its work on Money Market Fund (MMF) reform at the plenary stage. European regulators, after appreciating the massive disruptions that a regulation similar to the one that was adopted in the US might cause in Europe, have sought to find a safer alternative that is able to maintain a stable price in normal market conditions. In the US, MMF reform has had the undesired effect of reducing real economy funding by about $1.3 trillion, which is roughly the same size as the European MMF industry. As prime CNAV (Constant Net Asset Value) MMFs are no longer available to the majority of investors in the US, they simply do not want to use VNAV (Variable NAV) funds, which might also impose fees and gates.
There has been much discussion about the wider range of funds proposed in the EU regulation and how these may be appropriate for clients, with the right asset class and relevant risk controls. One product that will be innovative, and is expected to replace the CNAV MMF, is the Low Volatility NAV MMF (LVNAV). This product has been devised to address the risks that are perceived to emanate from CNAV MMFs and it achieves this aim. At the same time, investors will retain the ability to invest in prime MMFs whose price should remain stable in normal market conditions. Introducing a Europe wide set of criteria that must be adhered to by all fund promoters has meant that all liquidity funds that want to be considered for short term investments must be extremely robust and obey these criteria: these include daily and weekly liquidity, as well as having a mark to market price extremely close to the quoted market value. This is further enhanced by a greater supervisory oversight of stress testing, credit methodology and all aspects of portfolio transparency. This product has been designed to dramatically reduce risks.
The new LVNAV product is currently being presented to CNAV investors. LVNAVs should be the product of choice for most investors who need easy access to their cash with minimal credit and capital risk in the coming years. Undoubtedly, the number of providers of this product will be fewer than those currently offering CNAV, but the MMF industry will flourish, the speaker believed, as investors continue to appreciate the reduction of risk by having their cash managed by credit experts, who diversify both by interest as well as credit risk.
Another industry representative concurred that the end MMF agreement is good for the end clients, because there is much optionality: i.e. government CNAV, pure VNAV and LVNAV, which are close to VNAV, but with stable prices unless something happens. The speaker’s firm is planning to offer all of these products, even government VNAV to those clients who do not want to have fees and gates, which is a good development.
A regulator agreed that the new MMF regulation is an interesting and acceptable compromise, and that it will usefully complete the UCITS and AIFMD directives, even if it is difficult to anticipate what will happen in the coming years. Some countries such as France have always opposed CNAV because of financial stability concerns. The new MMFR is, however, an interesting compromise, because it takes into account the situation in other European countries and proposes a limited CNAV and the LVNAV. This latter new tool is interesting, with its enhanced liquidity buffers, the restricted use of amortised accounting, and the fact that an entity is obliged to always monitor the difference between the market net asset value and the stable asset value for unit holders.
3.2. New liquidity management requirements for MMFs
A regulator emphasized that the new requirements imposed by the MMFR are useful but their implementation process will be delicate. Over the next 18 months, there will be the possibility that every national competent authority will have access to all the portfolios of all the MMFs domiciled in their jurisdiction, and putting this in place in time will not be easy to organise. There will also be a completely new obligation by which the asset manager has to know who its unit holders are. This does not exist in other European asset management regulations, and will be similarly difficult to put in place. However, knowing unit holders better will be interesting for assessing the liquidity mismatch between assets and liabilities.
An industry representative added that knowing who the asset owners are is not just relevant to MMFs, but also for liquidity management overall. When asset managers forecast redemptions, they are looking at the underlying client makeup and trying to see how those individual client segments work. However, there is one area in which this is not possible, which is omnibus accounts, for which there is no visibility on the underlying clients. It is not necessary to have a look through to the individual client level, but having visibility on the client segments would help the speaker’s institution to improve the quality of their predictive analysis and their forecasting of client redemptions, which helps in terms of the stress testing and liquidity of those funds.
Finally, the liquidity provisions for government CNAV and LVNAV are tight. Although having liquidity provisions is positive, these are very prescriptive in terms of how entities have to meet them. Government CNAV funds, in particular, are heavily reliant on repo, and the repo market in Europe is very stressed. That could end up causing issues, and as such, there is a need to try to find solutions broadly across the industry. However, the money market providers should try to find solutions in their area, as it is an issue for the functioning of the funds for the future.
3.3. International consistency of MMF rules
An official agreed that the EU MMF regulation is a good development. There has been much divergence and disputes during the development of MMF proposals, but the amount of consistency and convergence that has been achieved between the EU and the US – the other prime jurisdiction – is encouraging.
There is consistency in at least four areas; the first is diversification ratios. Both jurisdictions have stricter diversification ratios, which is positive. There is also a requirement on both sides with respect to stress testing to maintain a stable NAV; with liquidity ratios, there are now requirements for daily and weekly ratios; and there are also new restrictions, gates, and new fees, both in the US and in Europe, all of which are positive developments. The provisions are not identical, but the path of travel is certainly in a more positive than negative direction.