ETF indexes face clean up

Regulators want to increase the level of transparency of ETF indexes and eliminate the conflicts of interest that could emerge between an index firm and affiliated ETF provider. But some participants think regulators risk going too far

The Libor rate-rigging scandal continues to have many repercussions. Regulators have looked for alternatives to the submission process, and have reviewed governance and oversight of the Libor methodology. But the after-effects have been felt beyond Libor itself – the integrity of financial benchmarks more generally have also been questioned.

Supervisors have looked to respond, and the regulatory reactions have come thick and fast. The International Organization of Securities Commissions (Iosco) published its principles for financial benchmarks for consultation in April, while the European Securities and Markets Authority (Esma) and European Banking Authority jointly released their final principles on the issue in June. They follow the UK Wheatley Review of Libor, which was published in September 2012 and specifically looked at the methodology for Libor. But supervisors are also shining a light into the world of custom ETF indexes – and they are proposing changes that some participants believe will hit the ETF industry and restrict investor choice.

The scrutiny of the ETF market is not entirely driven by the Libor scandal – work began in earnest after the Bank for International Settlements, the Financial Stability Board and the International Monetary Fund separately wrote reports in April 2011 warning about the systemic risks posed by ETFs. Esma consulted on a new set of guidelines on ETFs and other Ucits issues last year, finalising its rules in December, while Iosco concluded a lengthy investigation into the ETF market in June with a set of principles.

These principles are wide-ranging, but include a number of recommen-dations meant to increase the transparency of index constituents and calculation methodology. Market participants agree more transparency would help investors – but warn too much disclosure could harm trading strategies, and ultimately lead to less investor choice.

“We’ve always been quite supportive of transparency. Our index methodologies are published broadly and our research is readily available. The point at which I think you have to draw the line is availability of constituent-level data and when that data is made available,” says Michael Larsen, global head of affiliate relations at Research Affiliates, a Newport Beach, California-based firm focusing on asset allocation strategies.

It’s not entirely clear where regulators intend to draw that line. Among the nine Iosco principles is a recommendation that regulators require ETF providers to make certain disclosures related to the portfolio, including details on any reference index and its composition, and the operation of performance tracking. In particular, it suggests information on index calculation methodology, composition and relative weightings should be made available “in an appropriate time frame” – but doesn’t specify what that would be.

The general idea, however, is that enough information should be made available to allow the investor to replicate the index calculation – a requirement made explicitly by Esma in its ETF guidelines (see box, Esma rules on indexes for Ucits). The question is whether all investors want this information – and what they would do with it if they had it.

“The policy that you should be able to reverse-engineer every index is a good one in the context of retail investors. We have recently seen products in the non-ETF space that make this very difficult. However, even in broad indexes, corporate actions – takeovers or reverse splits – add complexity, which is why investors still rely on the index providers. They rely on index providers to process the changes so they can see the impact. If investors could truly do it themselves, then we would not need the index providers – which is clearly not the case,” says Michael John Lytle, chief development officer at European ETF provider Source.

The biggest complaint is the protection of proprietary information. A number of ETFs have been launched that are referenced to indexes based on proprietary trading strategies or algorithms. Revealing detailed methodology could mean banks are showing their hand – not only can competitors potentially replicate the trading strategies, but they may be able to profit from front-running the ETF.

Iosco recognises the concerns about proprietary information, and says regulators need to consider the appropriate balance between the level of disclosure and investor sophistication. On the other hand, it is also keen to clamp down on the possible conflicts of interest – particularly those that arise when an ETF is referenced to an index designed solely for that product and is created by a bank affiliate of the ETF provider.

In this instance, there is a risk that important non-public information may be communicated between the ETF and the affiliated index provider, Iosco says. This could lead to the index being manipulated to suit the trading position of a bank, or at the very least allow the dealer to get an early glimpse at the index and allow it to trade ahead of the ETF.

“When presented with direct conflicts of interest and an unregulated and unaudited benchmark setting, gains from manipulating these indexes may be significant due, for example, to derivatives positions, while the cost may be very low since the likelihood of detection of such behaviour in an unaudited system is low,” says Rosa Abrantes-Metz, adjunct associate professor at the Stern School of Business, New York University. “We want to have benchmarks that people can rely upon. For that, we need to have a system that has reduced incentives and opportunities to cheat, and appropriate auditing and regulation – and many benchmarks do not satisfy those requirements right now.”

Others agree that current organisation practices open the door to potential conflicts of interest. “If you have an index that is not systematic or is calculated by the bank itself in a way that is not transparent, then it could potentially create room for the bank to act on the index while the client cannot react, and that leads to a potential conflict of interest,” says Marc Pantic, senior index structurer at Société Générale Index in Paris.

But there are other reasons for wanting greater levels of disclosure – for one thing, it makes it easier to sell the idea to investors, says Thibaud de Cherisey, European ETF development director at Invesco Asset Management in Paris. “When we select an index that is out of the mainstream indexes, we need to be sure that we can disclose the methodology – not only the components but how you arrive at the result. If not, it is a black box and more difficult in terms of educating the investor.”

Iosco makes a number of recommendations in response. When a custom index is created by an affiliate of the ETF provider, regulators could require the public disclosure of all the rules that govern the composition, inclusion and weighting of securities in each index – again, in a non-specified “appropriate” time frame. It also suggests limiting the ability to change the rules for index composition, and establishing firewalls between those who create and maintain the index and those responsible for portfolio management.

The question is how regulators will respond to these principles. The Esma ETF guidance specifies that Ucits funds need to ensure any underlying index is a benchmark for the market it references – and states an index is not an adequate benchmark if it has been created and calculated on the request of one or a limited number of market participants. This may make it difficult for Ucits funds at least to reference custom indexes in certain circumstances, some observers say.

More broadly, some industry participants believe there needs to be a clear split between the entity that creates the index and the provider of the ETF. Ultimately, banks should choose whether they want to be an index provider or a trading firm, says Alex Matturri, chief executive of S&P Dow Jones Indices in New York. “In the long run, banks should choose either to be a product issuer or a proprietary trader. When you start doing multiple functions in the business chain, potential for conflicts of interest can exist.”

Some banks claim there are real benefits for investors from combining multiple roles – for a start, there isn’t the need to pay expensive licensing fees to index providers, they point out. However, even that isn’t as simple as it first looks.

“There are certainly some proprietary indexes where you get fees deducted from the level of the index itself. You know quite often that hedging fees might be built into it, so if the index is being hedged by purchasing and then selling stocks or funds, then you might get that cost built in, which comes out of the deduction at hedge level, which then affects the index level, which then affects presumably the value of the overlying security. Moreover, you might also get an in-built management fee or some sort of fee at the index level,” says Michael Logie, a London-based partner in the finance department at law firm Ashurst.

In any case, there are real benefits from paying for an independent index provider, claims Matturri at S&P Dow Jones Indices. “Saying self-indexing saves money is a red herring. We know the cost of these products, especially structured products, and what you are paying for at the end of the day is the integrity that comes with an independent index,” he says.

But while some are advocating a clear split between index creation and trading businesses, others believe the answer could be a requirement for independent auditing of proprietary indexes. “A solution to this could be regular audits – weekly or, at least, monthly – by an independent auditor if banks want to keep administering the benchmarks. We need a system that is very robust so we can prevent manipulation, and by minimising the risk we will avoid problems in the future,” says Abrantes-Metz at the Stern School of Business.

This is a principle that to some extent can also be extended to independent index providers, some say. “Should you use an independent committee to assess whether you’re doing things right, or do you actually do it internally so you don’t have the influence of external index users on it? That’s a perfectly legitimate debate and it’s somewhere regulators might decide to weigh in,” says Gareth Parker, director for index research, design and development at Russell Indexes in London.

Ultimately, national regulators will need to decide how to interpret the Iosco principles – and a lot will depend on how they interpret terms such as “appropriate time frame” for disclosure. However, supervisors should remember that investor protection is the primary objective – to that end, focusing on eliminating potential conflicts of interest may lead to a focus on governance, to the detriment of greater transparency, some participants warn.

“Such ideas may pave the way for a decrease in index transparency relative to the advances recently introduced by Esma, whereas the focus should be on generalising the investor protection and market competition advances introduced in the context of financial indexes used by Ucits to the maximum extent possible,” says Frédéric Ducoulombier, a Singapore-based director at research body Edhec Risk Institute–Asia.

 

Esma rules on indexes for Ucits

The European Securities and Markets Authority (Esma) guidelines on ETFs and other Ucits issues came into effect from February 18, 2013, and will be applied to already existing funds within a year. The rules establish new transparency requirements for index-tracking Ucits and update the eligibility criteria of financial indexes. With respect to transparency, Esma clarified that each index should have a clear, single objective and that the universe of the index’s components and the basis on which components are selected should be clear. Esma went further and prohibited the use of indexes that do not disclose their full calculation methodology or fail to publish their constituents together with their respective weightings.

This information is required to be accessible easily and on a complimentary basis to investors and prospective investors. In addition, Esma prohibited investment in indexes whose methodologies are not based on a set of predetermined rules and objective criteria, or that permit ‘backfilling’ of data.

One of the key objectives evident in Esma’s new requirements is the regulator’s desire to restrict Ucits’ choice of indexes to those that are built and managed in a systematic manner and for which index providers make available sufficient information to the public to allow for independent replication on a non-commercial basis. The requirements go beyond what would be needed for an understanding of the objective, methodology and historical performance of an index, which would suffice for investor orientation and a basic screening of indexes. They enable interested parties to audit the track record of indexes, observe how discretion is exercised and conduct performance and risk analyses to assess the relevance and suitability of each index against the specific goals of investors.

With respect to customised benchmarks, Esma states: “An index should not be considered as being an adequate benchmark of a market if it has been created and calculated on the request of one, or a very limited number of, market participants and according to the specifications of those market participants.”

Singapore-based Frédéric Ducoulombier, a director at research body Edhec Risk Institute–Asia, says this does not necessarily signal an opposition to innovation or strategy indexes, but shows Esma wants to assert control over asset eligibility. But he adds: “Unfortunately, the drafting of the paragraph is vague and Esma has not yet formally clarified its intention, despite industry requests to do so. There is therefore the risk that different member states will interpret this paragraph differently, causing collateral damage that was unintended by Esma, regulatory arbitrage opportunities and an uneven playing field for funds across Europe.”

 

Don't look back in anger

There are concerns over how banks create benchmarks that use back-tested historical data as an indicator of future performance. Xavier Ducros, London-based Rabobank’s head of structuring in Europe, says back-tests might look robust in theory but, when strategies start running, they can prove to be of poor quality. In addition, back-tests may have been constructed with certain rules of which the investor is not aware.

When track records rely to any material extent on back-tested data, there are risks the index methodology may have been optimised on the basis of this information, with little regard for the stability or persistence of its performance beyond this period. There are also risks that hindsight biases (for example, choosing from survivors, using restated/backfilled data, picking winners or shunning losers) entered the simulation, whether or not there was an intention to mislead. The methodology used for back-tests may also be different from that employed when an index is live.

Speaking at the Inside ETFs Europe conference in Amsterdam in June, Robeco quant researcher David Blitz voiced concerns at the lack of regulatory oversight of index providers in this regard. “One thing that strikes me as a bit odd is that, as an active manager, if you back-test a certain strategy and want to present those figures in a client presentation, then you’re basically not allowed to do it. You have to hire an independent auditor to verify the whole process. But if you’re an index provider, then you can back-test each of your strategies, put up any numbers you like and the whole world is fine with it. I don’t think there is a level playing field when it comes to showing historical performance figures.”

In the US in April, the Financial Industry Regulatory Authority issued guidance on the use of pre-inception index performance (Pip) data for certain exchange-traded products (ETPs) in communications with institutional investors. Under the guidance, Pip (also referred to as hypothetical or back-tested data) can be used to market certain passively managed ETPs, according to predefined rules. Pip data must reflect performance “during a period of time that includes multiple securities markets environments and, at a minimum, 10 years since the inception of the index”, and be current as of the most recently ended calendar quarter. It must also be “clearly labelled and presented separately from actual performance” data, with applicable dates disclosed.