21 Jun 2012, Credit Suisse , iShares , THEAM , Risk magazine
Dan Draper, Managing director, head of exchange-traded funds, Credit Suisse
Tim Lubans, Director, legal and compliance, iShares, BlackRock
Margot Pages, Head of index funds and ETF development, THEAM
ETF Risk: The European Markets and Securities Authority (Esma) is working on the final guidance for the exchange-traded fund (ETF) market. How important is this for the industry? Why is the regulator looking at this particular issue now?
Tim Lubans, iShares, BlackRock: We have seen ETFs receive an incredible amount of attention over the last 12–18 months from regulators, central banks, the press and clients. The ETF label has caused a bit of confusion in the market, so we very much welcome Esma’s efforts in drilling into the product label and creating guidelines that increase transparency and disclosure. But, within that, it’s important we create a framework that leads to sensible regulation that draws out the differences between ETFs and other exchange-traded products, and results in a classification and labelling system that is meaningful, sensible and works for investors.
Dan Draper, Credit Suisse: If we look back to 2008 and the crisis that was faced, it was really the first opportunity for many of the emerging markets – China and others – to be part of the Group of 20. Part of their conclusion was forming the Financial Stability Board (FSB), which has been a major participant in initiating a lot of the global and regional responses to these issues. ETFs and high-frequency trading were two issues that it wanted to look at initially. ETFs had grown by more than 30% each year for 11 consecutive years. They were clearly very popular tools and regulators said ‘let’s focus on a couple of areas where, as regulators, we can get our arms around and really have a look at these products as they are growing and will arguably continue to grow even faster in the future’. Within that context, Esma and other regional regulators have certainly stepped in. From the beginning, the approach has been that regulators want the industry to work together to try a form of self-regulation, but that they will hopefully provide some strong guidance along the way.
ETF Risk: There has been a lot of debate in the media about synthetic and physical replication ETFs. Can you explain the difference between the two?
Margot Pages, THEAM: Historically, fund providers have used pure physical replication. In this replication methodology, the fund managers buy each of the securities comprising the underlying tracking index. It requires rigorous rebalancing and foreign exchange management, but basically it is easy to implement and easy to understand. It does not mean there is no counterparty risk because fund managers often engage in securities lending to enhance the returns and cover some of the costs. Counterparty risk may hence arise from securities lending activity. The drawbacks of this approach are that it can incur high rebalancing costs and you may have some liquidity or access issues. One way out of this is to use optimised replication, which is still based on buying physical securities, but fund managers only buy a sample of the tracking index to match the index as closely as possible.
The other way out is to use synthetic replication, which we also call swap-based replication. Fund managers buy a basket of securities, often called the substitute basket, which may have nothing to do with the tracking index. This is a basket that is agreed between the fund manager and the counterparty, but it is always the fund manager that has the last word. The fund manager swaps the performance of this basket for that of the tracking index with the swap counterparty. As a result, the exposure is to that of the tracking index. Counterparty risk is limited to 10% under Ucits and can be mitigated through swap restrikes or collateral management, or both. This is known as the unfunded swap model.
In the funded swap model, cashflow from the client is used directly to pay the swap counterparty and the counterparty risk is mitigated through collateral management. All these methods have different characteristics, with their own advantages and disadvantages. What is important to understand is that the question is not ‘which one is the best?’, but it is how these methodologies are implemented that really matters.
ETF Risk: What is your opinion on the call for ETFs to be very clearly labelled – potentially separate labels for synthetic and physical ETFs? Would it mean anything to the majority of investors?
Tim Lubans: BlackRock and iShares have always had a preference for physically replicating ETFs – that has always been our default approach. In Europe, there have been certain markets where the underlying investments are difficult to access or there are other issues in relation to direct holdings in those securities. And so, in limited circumstances, we have developed derivatives replicating ETFs. But it’s worth flagging that we include the word ‘swap’ in our name in all cases where we have a swap-based fund – by that we mean a fund that uses derivatives to obtain and achieve the return of the benchmark index as its primary investment objective. We have also seen regulators highlight this as an issue in other jurisdictions. For example, Hong Kong has imposed a similar requirement where they have a synthetic identifier. We really believe investors should understand the way in which their products are structured and the risks associated with the investments that they are looking to make.
Dan Draper: Let’s remember that Esma is putting together guidelines, which could potentially be binding, but you are really looking at regulative changes that need to come out of Brussels. Overall, done in a responsible way, labelling can absolutely help but it has to be done in a way that doesn’t favour certain products over others. It just needs to help in the overall disclosure process and ultimately represent the fiduciary responsibilities of the issuer in favour of the investors.
Margot Pages: It is important that providers are very transparent and disclose the methodology they are using. But putting the form of replication in the fund name can actually be quite misleading. It can stigmatise both methodologies, and associations can be made such as ‘swap-based equals counterparty risk’ or ‘physical-based equals no counterparty risk’, which are wrong, but still sometimes present in investors’ minds. We must disclose, but the name is not the right place to do this. The way Ucits addresses this type of issue is through the key investor information document (Kiid), which is a two-pager in which providers must summarise all the characteristics of the fund. It is much more relevant to put it in the Kiid than in the fund name. Remember also that categorisation is not only about physical versus swap: there are different physical methodologies – pure, optimised, with securities lending and without – and different swap methodologies – funded, unfunded, mono-counterparty and multi-counterparty. If we have to specify all of that, the name is going to be very long.
I agree that securities lending can bring lots of value to the fund and I am not saying physical replication is not as good as synthetic. Actually, we are very agnostic at BNP Paribas on the replication methodology and have developed our expertise in both. But there are different ways you can implement these strategies. It is not whether it is physical or synthetic that defines the risk: it is how it is done. So putting it in the name would be misleading.
ETF Risk: The Bank of International Settlements (BIS), the International Monetary Fund (IMF) and the FSB have drawn attention to the risk posed by the swap counterparty. Are they right to be concerned?
Dan Draper: For the most part, we are not offering products that are being promised in guaranteed form – these are not risk-free assets. We are fiduciaries and any responsible fiduciary has to make sure that, through the investment policy objectives of the fund, they are trying to meet with what the customer wants to accomplish. So, through that drilling down, yes counterparty risk is there, but it can take different forms and be done in a way in which someone can build a responsible portfolio. The types of risks that you are willing to take on need to balance out the return objectives that you have.
In Europe, we are dealing with the universal banking model, which worked really well post-WWII in the rebuilding of the European economy. Now it is different and it involves dealing with these fiduciary questions within a universal banking model versus the more segregated model that has worked in the US. I believe we should gain clarity, understand that this is an exchange-traded product and make sure we understand the risks. But, on the other hand, this is a fund and there are fiduciary responsibilities that a responsible board of directors should oversee.
Tim Lubans: Another concern is the combination between the fiduciary duty, which no doubt applies to all of us as investment managers, and the potential conflicts of interest that may arise if you have an ETF manager that is a Ucits but has a parent company that is the only swap counterparty writing a swap for that ETF provider. That ETF provider may only have one authorised participant or market-maker that is also affiliated with that counterparty. So, not withstanding the Ucits fiduciary obligations that we all have, we have concerns about the potential for conflicts of interest, which are not necessarily understood by investors. They are not explained to investors and the risks associated with those are not clearly identified, and that’s one area in which we think the Esma consultation could have gone further.
Margot Pages: The question of how the swap is executed is key. There are cases where there is a reason for the fund to deal with the parent company in a closed architecture such as innovation or operational advantages. But Esma could have looked further into the execution process of swaps.
ETF Risk: Do you think the concerns raised by the regulators last year that a bank could use the ETF sponsor as a cheap way of raising financing were overplayed? Could that happen? How does your institution address this issue?
Margot Pages: It could happen, indeed in the case where the fund manager deals exclusively with its parent investment bank. So it is important to look into it, although I don’t think there have been any big issues. At BNP Paribas, we are in a totally open architecture – we are not tied to our parent investment bank, which avoids any conflict of interests. THEAM uses both synthetic and physical replication methodologies depending on the underlying we want to track. Every time we plan to launch a fund, we assess what is the best-suited methodology, depending on the underlying. We consider the liquidity of the underlying asset, access, turnover, impact on tracking error and regulatory framework.
Dan Draper: We have seen changes since January and this emphasis on self-regulation has been taken to heart. This debate has led to a number of providers – particularly synthetic – not only disclosing counterparties but also collateral, really looking at different models. At Credit Suisse, we are the second-largest physical provider in the European market after iShares, but we do have a few swap ETFs as well. We have converted 10 of our 16 original swap ETFs into physical, not solely because of the regulation environment but because we had changes in demand that came about before April 2011. We have adjusted our business model accordingly. We will keep a swap platform where it is demonstrably better to do swap ETFs, but we built in the flexibility to potentially switch into physical. When the regulatory papers came out in April 2011, it increased the demand, so we worked with our fund management team in the months leading up to those regulatory announcements. We felt very comfortable getting proper tracking error versus meeting that client demand.
ETF Risk: Has there been a change in investor perception since the regulatory papers emerged last year? Are they becoming more familiar with the nuances between synthetic and physical?
Tim Lubans: We had a similar experience to that of Credit Suisse in terms of increased scrutiny and due diligence from investors, Clients have wanted to know more about how products are structured, how we manage those products, as well as the risks and how they would play out in a default scenario. Fortunately for us, the increasing demand for physical funds has worked in our favour because iShares is predominantly a physical provider and we continue to build on that physical fund range. We still see the benefit of synthetic funds in certain limited circumstances and there will be cases where it is appropriate to launch a derivative-replicating fund rather than a physical fund, but our default approach and preference would always be physical. It is actually quite pleasing to see, in terms of industry alignment, that the regulatory attention has engaged us and encouraged us to work as an industry to enhance our own standards and build the ETF product category itself, which is obviously in our collective interests. So we have been encouraging clients, and clients have been coming to us seeking that due-diligence information in terms of structuring, product detail and risk.
Margot Pages: We are being asked some very detailed questions about how we implement both methodologies that require lots of disclosure. We realise through these questions that investors do have a better understanding of these methodologies, which is very good. Because we use both replication methodologies, we are well positioned to explain to clients the differences between the two and why we have chosen one rather than the other for each particular fund. This is very reassuring for investors. Nevertheless, we observe that there is still a client misperception that physical replication is always less risky than synthetic. We notice that they underestimate the tracking error risk of the optimised replication methodology, for example. With this methodology, you rely on the proprietary model of the manager, which is rarely transparent and incurs model risk. Investors and regulators clearly overlook this point.
Dan Draper: There is a stereotype that all swap ETFs are equal. We can talk about unfunded versus funded but, even on the return characteristics, they can be very different because you have this universal banking model where you have the swap provider inside, maybe alongside the issuer as well. If there is a positive swap spread, some of that will be subsidised through that combined model. At other times, the swap spread is a cost that is passed on to the investor. Even the performance, the basic idea that you are going to get this total return less the total expense ratio, may not always be the case. It depends on the issuer and sometimes depends on the ETF. There has been such a proliferation of swap ETFs within the last three to four years, you just haven’t had a long enough track record and performance history to really start seeing these differences.
ETF Risk:Esma has suggested the same rules be in place for total return swaps on synthetic and all Ucits and securities lending on physical replicating ETFs. It has also called for the collateral to be significantly diversified. Do you think this is along the right lines?
Margot Pages: It seems the credit crisis has pushed investors and regulators to focus so much on counterparty risk that they may sometimes disregard market risk. The confusion is such that Esma has suggested the application of a rule initially designed to mitigate investment risk to reduce counterparty risk. It is proposing that the Ucits diversification rule, which requires concentration limits on the index, also be applied to the combination of collateral and portfolio assets, which are actually not of the same nature. This will be very counterproductive because it will introduce higher operational risk, resulting from the constant back and forth between the fund and the counterparties to respect those new rules. You have to think of all the side-effects of that. It means it will be very complicated to deal with multiple counterparties, and providers will most probably switch to a single counterparty model. Another impact is that the operational burden will most probably lead providers to make fewer adjustments. This type of rule would hence be counterproductive and introduce operational risk rather than mitigating counterparty risks.
There is clearly a wrong perception of the purpose of collateral. Collateral is not meant to be kept as an asset in the portfolio, but to be quickly liquidated if the counterparty defaults.
Tim Lubans: At a European level, the industry is unanimous that collateral is intended to address a completely different issue in risk than the actual investment exposure. And that aggregating diversification across the two doesn’t really make any sense. It presents many more issues than it would perhaps resolve. We would be open to diversification standards or parameters being applicable to collateral, but they shouldn’t be aggregated with the diversification and collateral requirements in relation to the actual holdings. The key factors in relation to collateral are liquidity and the quality of that collateral.
ETF Risk: Should Esma draw up a comprehensive list of eligible collateral that firms could use?
Tim Lubans: We wouldn’t want to find ourselves in a position where Esma or any other regulatory body is prescribing specific collateral, which may or may not be taken and doesn’t have the ability to move with markets. For example, with what is happening in Europe at the moment, Greek bonds may have been acceptable a few years ago but we are clearly now in a different position. So it is much more important that the investment manager, who is best placed to manage that risk on an ongoing basis, has the flexibility to really assess what is appropriate as collateral within those qualitative criteria.
Margot Pages: It is important to leave some flexibility to the fund manager. Markets evolve and they need to adapt to the changing situations. I regret that this can be seen as a sign of diminishing confidence in the fund manager’s duty. If the regulator imposes such specific guidelines and firm lists, then what does the role of the fund manager become? When investors pay a fund manager to get investment exposure, it is not to do something that they could replicate themselves following deterministic guidelines and rules. Investors are buying expertise and the fund managers need to have sufficient flexibility to add value and adapt the portfolio to changing situations.
ETF Risk: Should limits be imposed on the proportion of an ETF portfolio that can be lent out as part of a securities lending programme? And what is your opinion on whether cash collateral that comes back from securities lending should be invested in risk-free assets?
Tim Lubans: We don’t believe such a cap is necessary. Provided a securities lending programme is appropriately managed with a sufficiently robust risk management framework with multiple borrowers and the risks and benefits outlined to investors, a considerable proportion of the fund can still be out on loan and still be perfectly acceptable from a risk perspective. It is probably worth flagging that our iShares ETFs have less than 20% out on loan on average, so we are not talking about the kinds of figures that perhaps some people have in mind.
On the second point, we want to make sure cash remains a viable option in terms of collateral from the borrower’s perspective. We’re concerned that limiting it to risk-free re-investment may diminish the attractiveness of providing cash collateral. It is not necessary to require risk-free investments and, in fact, the two key considerations from our perspective are the preservation of capital and maintaining liquidity in relation to that collateral.
We have separate boards of directors that are responsible for assessing and undertaking a performance analysis of BlackRock in its capacity as a securities lending agent. That performance review happens on a regular basis and we are benchmarked against other providers in terms of performance, risk management and returns. It is not just BlackRock making a decision to engage in securities lending, it is endorsed by the fund boards.
Dan Draper: Securities lending has traditionally been part and parcel of trying to provide an enhanced return. We are clearly in a very risk-averse environment right now and preservation of capital is key, but securities lending done responsibly does provide enhanced returns. I imagine most of the audience, and we on the panel, have money deposited in a bank. What does a bank do? It lends it out in order to give us a better return because we don’t need that money until the future. This idea of lending, which is inherent and has been part of indexing since the beginning, needs to be done in a very responsible manner. That is why, out of Credit Suisse’s 43 physically replicated Ucits ETFs, we have done the due diligence and only engage in securities lending on seven of those because we have only identified that the extra benefit to the end-investor of engaging in securities lending justifies the potential risk within those seven ETFs. Done in a responsible way, securities lending enhances long-term returns in select areas.
ETF Risk: Should the same rules be in place for physical ETFs as well as the synthetic, particularly with regard to collateral?
Margot Pages: Yes, there should be harmonisation of rules between securities lending and derivatives, as any source of counterparty risk should be treated the same. However, as I said earlier, setting additional rules than the ones that are currently in place would be counter-productive. If we do implement these diversification rules on collateral, it will benefit synthetic replication over physical replication. This is because synthetic replication managers may circumvent the operational burden of such rules by using a single counterparty model. For securities lending activity, on the other hand, this is not possible – you cannot find just one borrower that would be willing to take all of your securities, or at least not at the best price – so this means the rule considerably limits the securities lending possibilities of physical funds. Regulators might be favouring synthetic replication over physical without actually realising that they are going in this direction. It is the same thing with the French tax on financial transactions where physical funds will be taxed and synthetic funds will not.
Tim Lubans: That is not to say that, in the context of synthetic ETFs, there isn’t lending going on. The information isn’t necessarily known to investors and it is not always clear what an investment bank is doing. But, generally speaking, it is possible to conceive of a situation where you have an investment banking parent writing a swap with its ETF provider, hedging its underlying by investing in those securities directly and then lending them out and earning a fee on that. None of that information is really transparent to investors although it may be of interest to them and regulators.
ETF Risk: How can physically backed ETFs versus synthetically backed ETFs contribute to systemic risk in a time of financial crisis?
Margot Pages: Regulators have been very concerned about the systemic risk impact of ETFs. I have never actually understood why they have focused so much on ETFs. It is most probably because ETFs have grown very quickly and this has attracted attention, but there is no reason to distinguish ETFs from other funds. All of the things we have discussed in this panel – swaps, securities lending, etc. – are common practice and concern all types of Ucits funds. There is nothing actually specific to ETFs.
Tim Lubans: It is important to note that Esma’s focus is really on the investor protection angle. Systemic risk is obviously very relevant from the perspective of other bodies such as the FSB and central banks around the world. There may be areas in which additional analysis is necessary – and there are certain questions that are being asked within the context of shadow banking that will need to be addressed – but derivatives investment or securities lending are not unique to ETFs. That is a management-wide global issue.
Dan Draper: Whether it is in Washington, Paris, Basel or Hong Kong, ETFs have established they are going to increase and become mainstream products. But there is this macro-prudential approach whereby regulators are obliged and expected to ask questions in advance. They do not want to be on the back foot. The thought process is ‘we may not ask all the right questions, we may not get all the answers that we need right now, but we need to ask these questions and get our arms around it from the beginning.’ From that very general philosophical perspective, it has been a healthy process.
ETF Risk: In its January paper, Esma highlighted the problems with the current assumption that all Ucits funds are non-complex. It suggested that that will be addressed in an upcoming revision to the Markets in Financial Instruments Directive (Mifid). Is this something you think is important?
Margot Pages: It’s a complicated decision because there isn’t just one set of products that is simple and another one that is complex. Where do you draw the line? And what defines complexity – is it the investment strategy or the way you implement it? It is important to understand that complexity does not equate to risk. You can have very sophisticated investment strategies that aim to maximise returns while minimising risk. We have to be careful not to stigmatise the products and avoid wrong associations. It is clearly best addressed within Mifid rather than in Esma because all Ucits are non-complex by definition.
ETF Risk: What do you think will be the biggest challenges within the Esma guidance?
Tim Lubans: We’ve already made some improvements through industry alignment and, through the Esma guidelines, we should get to a place where ETFs are even better regulated, better understood and ultimately a better-labelled product. The biggest issue is it doesn’t address the entire universe of non-ETF exchange-traded products and it is still not clear to investors where that distinction lies. It will be our job and the job of the industry to help educate regulators and the rest of the industry to make a distinction between these regulated Ucits fund vehicles and the variety of other products that are exchange-traded and made available to investors in similar ways.
Dan Draper: On a consumer level, there are going to be big changes in the advisory community over the next few years, with things like the Retail Distribution Review in the UK. So the way people will want to use products like ETFs is going to change. Mifid is going to potentially have a big impact on all traded products. We are lucky, in some ways, to be at the beginning of this review, getting our house in order now. If you look at a proper fiduciary-managed ETF, it is going to be more attractive in this post-crisis world. So it is a really good story not just for ETFs, but in the broader context.
Margot Pages: Concerning the Esma guidelines, most of the transparency requirements already correspond to the industry’s best practice, so there is not much that we can expect there. But it is much more the collateral side that is worrying to us.