Global Pensions gathered industry experts to discuss how the development of indices have evolved beyond market-cap weights to meet investors' changing needs
Raquel Pichardo-Allison: Andrew, what have been some of the major challenges that investors have faced in recent years, and how is that affecting their perception of investment opportunities?
Andrew Buckley: In terms of institutional investors, the biggest single challenge – or the elephant in the room – is meeting future liabilities. Ultimately, that is what it is all about. Clearly, the performance of markets in the last decade has not been particularly supportive of that so investors are currently reassessing portfolios, looking at a range of options, including passive strategies. The biggest challenge now is addressing cost, transparency and diversification opportunities, and I am sure that the investment-consultant community in particular would have a view on that.
David Thompson: Investors over the last 10 to 15 years have needed to achieve a deeper understanding of their liabilities, first and foremost, before modelling the assets in which they are able to invest. However, the characteristics of those assets, on both a risk and a return basis, have changed significantly over the last five years in particular. Added to that, as Andy said, is the low return that we have seen in many asset classes over the last 10 years.
Raquel Pichardo-Allison: How does what is going on in terms of their liabilities and low returns affect the way that people look at the types of indices they might use?
David Thompson: The first problem that investors face once they have a deep understanding of their liabilities is choosing the asset classes in which they want to invest; strategic asset allocation. Once you have decided between passive or active, the next step is to choose the appropriate benchmark or index, if you like, against which to manage – or employ a manager to manage – the assets against. It is fair to say that, if you do not choose the right benchmark, no matter which manager you choose, they will not get you out of an incorrect benchmark choice. A badly performing manager can destroy a good benchmark choice, but a good manager can never get you out of an incorrect benchmark choice.
Raquel Pichardo-Allison: Andy, we have seen a lot of the larger pension funds come under pressure to maybe use passive instead of active strategies; have you seen any movement on this?
Andy Barber: There has been some movement, but nothing as great as the press might have it. Last year, we looked at assets placed by our clients in searches, the vast majority of which were still to active; 12% of assets went into passive, which is slightly up on prior years.
Phil Tindall: I would say that we have not seen a wholesale shift from active to passive. Some of these strategy indices have changed the landscape of the options available now. One thing that our clients are realising in terms of the governance burden of diverse portfolios of active management is that there are some difficulties and complexities in running strong, active portfolios. This means that something in between traditional market-cap passive and active may be a valid middle ground.
Andrew Buckley: The range of approaches that investors can now take along that continuum has greatly increased in the last few years. It is, then, a much more complex landscape than it perhaps was five or 10 years ago.
Raquel Pichardo-Allison: Rob, when you are talking to investors in the US about fundamental indexing, do you have this conversation with them about whether or not it is active or passive, or something in between?
Rob Arnott: The very fundamental question is: what is passive and what is the core purpose of indexing? Fundamental-index strategies offer investors a passive way in which to invest in companies based on the size of their underlying businesses. From a Graham and Dodd, economy-centric world view, RAFI is studiously neutral, just as cap-weight is utterly neutral relative to the cap-weighted market. From an economy-centric world view, the cap‑weighted market is a growth-tilted, momentum-chasing, very active strategy; from the cap‑weighted-centric world view, fundamental indices are value-tilted, anti-momentum active strategies.
Critics do not yet accept the notion that both world views are correct: that there is an economy-centric world view of Graham and Dodd, and a cap‑weighted‑centric world view of modern finance theory.
Lionel Martellini: In terms of the major challenges that I have seen investors facing over the last few years, the main one is their inability to achieve smart spending on their risk budget. As painful as it is, pension funds are willing to move away from liability-hedging assets and invest in equity markets, only because they need to generate performance so as to alleviate the burden of contribution. What they hope to get in exchange for that very expensive risk budget is a fair access to some risk premium, which they do not get with cap-weighted stock indices.
Kal Ghayur: At an implementation level, our conversations with large asset owners have concentrated on one basic conflict that they face: the need to diversify across managers versus alpha erosion that manager diversification appears to cause. This basic conflict has created a situation in which returns that should have been realised from the active risk that was taken in the risk-budgeting process did not materialise. The conflict really is that large asset owners because of their size need to hire a large number of active asset managers to diversify the manager investment process risk, operational business risk and so on. On the other hand, having a large number of managers results in alpha erosion.
The investment need that arises from all of this is for passive solutions, where the investment-process risk is almost eliminated, and yet you capture some portion of that alpha in a very high-capacity, transparent and cost-effective manner.
This then leads asset owners to more efficiently structure their portfolios in terms of their active risks and management-fee structures. It allows them to concentrate monies in truly skilled managers and to pay them a fair reward.
The motivation behind the FTSE ActiveBeta indices that we have recently launched in partnership with FTSE was exactly to solve this particular issue. Our research indicates that a significant portion of traditional active equity management returns arise from systematic sources of returns. The FTSE ActiveBeta indices have been created to provide an efficient, transparent, and cost-effective capture of these systematic sources of active equity returns. As such, these indices represent a passive alternative to traditional active management strategies.
Rob Arnott: Something that I think is quite important is breaking free from the notion that there is only one definition of beta. We wrote a piece recently entitled Beyond Cap Weight, which looked at a spectrum of some of the alternatives that have been coming out of the woodwork. One of the more interesting uses for whether it is fundamental index, Kal’s index or EDHEC’s index is as a diversification of investor beta exposure: do investors want their equity investments to capture beta tied to the market or to the broad economy, to maximise stock market diversification, or as a quest for risk-controlled equity investments?
These would point to cap-weight, fundamental-index, equal-weight, and EDHEC’s new diversity‑weighted programmes respectively – and of course there are others. Why on earth should an investor be singularly exposed to the first of these – to a popularity-weighted, growth-focused manifestation of stock-market beta?
Phil Tindall: In terms of where active fits in the spectrum, something that our clients struggle with is governance.
One issue that we try to get clients to think about is how much governance they have and how much time they should spend on different activities. Typically, we find that clients spend a lot of time on active management – seeing managers on a quarterly basis etc. There is nothing wrong with that, necessarily, but clients with a limited governance budget could do more of other things that they do not do, such as inclusion of alternative assets and a more dynamic approach to allocation. If these strategies take away the burden on governance, they are worth their weight in gold.
Raquel Pichardo-Allison: I would be interested to hear from some of the other consultants like David and Andy in terms of what they are recommending to clients around alternative types of beta.
Andy Barber: If we are just talking about equities, we are in danger of overcomplicating the thing as far as clients are concerned. If a client has decided to spend some of their risk budget investing in equities, they are doing it because they think that there is a premium to be had. The question then is: what are the ways in which you can access that equity risk premium?
Historically, the starting point was always just cap-weighted benchmarks; now, we have far more tools to work with and it is a lot easier to say, ‘Here are the characteristics of a market-cap benchmark. This is why it became a starting point. Here are some new approaches, which also capture the equity risk premium. Here are their characteristics’.
Looked at in a fairly simple way like that, clients are very willing to look at alternatives to a cap‑weighted index and maybe to have a combination of ways of accessing the equity risk premium. You then get into the debate about active versus passive.
Andrew Buckley: The important point to note about these approaches is that it is not about choosing one or the other – market cap weighted or strategy indices. It is about combining them to maximise diversification benefits, and by doing so investors will be able to better manage risk and return across differing markets. What we have to be careful with, is not to completely turn our back on market capitalisation; it is still the best proxy that we have for liquidity.
Andy Barber: We are not in danger of doing that. Whatever strategy you adopt, you are never going to get away from the fact that, even if you go to another form of indexation because it is lower-risk, it is inevitably going to be compared to a cap-weighted benchmark. You could have a period in which a cap-weighted benchmark does very well, and then you have to remind people about the risk and why you went away from it, but it is always going to be the comparison point.
Lionel Martellini: It will be the benchmark of the benchmarks.
Rob Arnott: Not only that, but the market is cap-weighted. Non-cap-weighted portfolios have to be offset by other active strategies that are diametrically taking on the opposite portfolio bets.
Phil Tindall: In a way, it is unfortunate that we have grown up with market cap and the capital asset pricing model (CAPM). To me, however, it is ironic because a lot of these strategies have the same risk as market cap, or even slightly less, and that is the important thing certainly relative to liabilities: absolute risk, not tracking error. Unfortunately, however, because we start with CAPM, we get tracking error and, in some cases, quite high tracking error. It is difficult to move clients away from that. If you have the CAPM mindset or start point, you are going to see quite long periods of underperformance, so clients have to go into this with a very long-term mindset. Ideally though, they should throw tracking error relative to market-cap away, and just think about this as an absolute‑risk‑relative-to-liability problem.
David Thompson: We have spoken quite a lot about equities, but there are other asset classes in which market‑weighted indices certainly are not the best starting point. We have done a lot of work in credit and, if you look at credit indices and their market-weighted indices, would you want to invest a greater percentage of your portfolio in a group of issuers or a sector that has seen a significant increase in issuance over the last two years? The answer is that you certainly would not. It is quite clear, in the credit area, market-weighted indices do not work.
Lionel Martellini: There are places where it is even more obvious that market-cap-weighted does not necessarily make sense, and the corporate bond market is a perfect example.
Equity and bond markets are the two places where efforts in terms of providing better benchmarks should be focusing.
David Thompson: The benchmarks are easily ‘sliced and diced’. We have found it quite easy to create bespoke benchmarks for our clients, something we have done a good deal of over the last couple of years.
Rob Arnott: It is an important hallmark of most of these new alternative index approaches, whether they are called passive or formulaic active or whatever, that they are valuation-indifferent; in other words, the index weight is no longer tied to price. We have tested valuation-indifferent indices all over the world, and they work nearly everywhere we test them. Variants of valuation-indifferent indices even work in commodities and currencies.
Raquel Pichardo-Allison: What do strategy-based indices offer investors that are true to the purpose of indexing?
Rob Arnott: When we look at the non-cap indices, they are true to the purpose of indexing if we recognise that the purpose of indexing varies from one investor to another. If an investor’s goal is to match the market – i.e. to make no active bets relative to the market – cap-weighting is the correct choice. If their goal is to match the look and composition of the broad economy, fundamental index is utterly passive relative to the broad economy. If it is to create beta exposure that is risk-minimising, minimum-variance strategies come into play.
It is, then, true to the purpose of indexing if we broaden the purpose of indexing to represent broadly the development of core portfolios
Andrew Buckley: You could argue that cap-weighting itself is a strategic choice. Fundamentally, indices have changed over the years in line with investor sophistication. These new indices may use an evolved strategy that goes beyond market representation, but are still true to indexing – using rules based methodologies to provide both beta and active returns. The work we have done with both EDHEC-Risk and Research Affiliates is bridging the gap between passive and active management.
Lionel Martellini: Not only do they have active choices in terms of their constituency etc, but in some cases you do not necessarily have transparency. It is not necessarily the case for FTSE but, as we all know, in some cases the decision process around which stocks are to be included in the index is very fuzzy.
David Thompson: They do enable investors to determine how much of their assets they want managed in a particular style. If that is the purpose of indexation, the strategy benchmarks are true to indexation, I would suggest. It also allows managers to play to their particular strengths: you can divide your assets into different styles, and then allocate them accordingly to managers who are good in each particular style.
Raquel Pichardo-Allison: Is it something that you are actively recommending to your clients to go alongside market cap?
David Thompson: It is something that we are always looking at. As I mentioned earlier, the third decision that we make is: what is the appropriate benchmark? That is absolutely key.
Raquel Pichardo-Allison: Phil, can you tell us what you are recommending to clients?
Phil Tindall: It is very client-specific, but we feel strongly enough about these strategies that clients should move some of their passive money out of traditional market cap. Almost regardless of whether you feel that one is better than another, there is a diversification argument: who knows which one will outperform over the short or medium-term? Why would we start with market cap? Separately from that, we do think that there are some return positives from these non-market-cap strategies, partly to do with non-price-weighting in the index or use of optimisation techniques, for example.
We have had this idea of ‘beta primes’ for a number of years. This came from a desire to find better benchmarks than market cap, rather than necessarily replacing active management.
That said, we do believe that many active managers charge a lot of money relative to the value they add, and if we can find something that we think does better than market cap and is cheap, that is also pretty good for the industry, then we should advise our clients accordingly. It raises the bar in terms of what active managers need to achieve. We are not ruling out active management at all; we are simply saying that there is a new bar.
Andy Barber: Dissatisfaction with cap weight goes back a long way, to the example of Japan in the late 1980s. Five years ago, there was not much in the way of viable alternatives; now, the debate has opened up and quite a lot is going on and being talked about.
Phil Tindall: Alternative approaches to market cap are not new. There was a form of fundamental index, GDP weighting, which was used to weight regional equity indices. This followed criticism of the Japan market cap allocation in the late 1980s, which reached 50%. As far as I know, it did not really take off, but it is the same principle as the stock level now.
Raquel Pichardo-Allison: Kal, are there certain asset classes that are better suited than others to strategy indices?
Kal Ghayur: Yes. Before I answer that question, can I go back to an earlier comment?
Raquel Pichardo-Allison: Absolutely.
Kal Ghayur: This whole group has been overly critical of market-cap weighting, and I wanted to put this in the right perspective. We should not forget the fact that market-cap-weighting, to a certain extent, is a macro-consistent, market-clearing weighting scheme. I completely agree with Phil that, from an industry-wide adoption point of view, the contribution that strategy indices bring is that they offer investors tools and vehicles to diversify whatever problems they perceive a market-cap-weighting scheme has. That is the way in which these so-called strategy indices – a term that I do not particularly like–
Raquel Pichardo-Allison: What would you prefer?
Kal Ghayur: I would like to talk in terms of sources of returns. Some sources of returns are systematic in nature and are independent of whether an index is market cap weighted or not; others arise from weighting schemes that deviate from market cap weighting. I would like to make that distinction between the sources of returns to differentiate between the various types of strategy indices, but we seem to be lumping everything together. We need to be cognisant of the fact that it is a diversification argument. We are providing options to investors. No one is talking about getting rid of market cap altogether in the next five years.
Returning to your question in terms of what asset classes are better suited to strategy indices than others, I do not have anything insightful to say on this, other than to make the comment that the development of strategy indices requires a tremendous amount of analysis and research. As such, since having a high-quality data set is a requirement, it is a lot easier to create strategy indices for some asset classes, such as equities, than it is for others asset classes where data is a major constraint.
David Thompson: Is that not the same for creating any index?
Kal Ghayur: For certain asset classes, you have very good quality data, stretching back almost 100 years; for others, you do not. In those asset classes, it becomes a lot more difficult to develop more granular strategy indices.
Rob Arnott: There is more to it, Kal, than what you are describing. It is not just a data-availability issue; it is an issue relating to how markets behave. What we are mostly talking about is a whole array of valuation-indifferent strategies, where price is not part of the weighting scheme. The success of valuation-indifferent strategies is based on market inefficiency, based in turn on a pricing error. If the price is wrong, and if one of the market’s main functions is to identify fair value and, therefore, to correct pricing errors, then you are going to have a pricing error and a market that seeks to correct these errors, leading to mean-reverting error in price. The bigger the pricing error and the more quickly the market can detect and correct these errors, the greater the value added by the valuation‑indifferent indices.
Kal Ghayur: Rob, I agree with you, and that is one specific example of strategy indexing. Clearly, you have tested it and it worked, and that is great. At the other extreme, however, momentum strategies also have been demonstrated to work across many different asset classes. What I was referring to was strategy indices as a group, where you have a whole suite of strategy indices, like you do in equities. It is very difficult to develop something like that for private equity, real estate or hedge funds. That is what I was referring to.
Lionel Martellini: We have seen in the past many attempts at launching active and passive hedge fund indices. We have done a lot of research on these questions and it is fair to say that those attempts have not been overly successful, for a number of years.
Raquel Pichardo-Allison: What are some of the risks associated with these types of indices?
Lionel Martellini: Again, it is a question of how we define risk. What is important is to try to understand the market conditions under which non-cap-weighted strategies might disappoint or deliver a performance that is maybe not as high as cap-weighted indices. The bottom line is that we probably do not expect any of those strategies to outperform market-cap-weighted indices in all market conditions.
Andrew Buckley: We don’t see why strategy indices should just be limited to equities. If currency is considered an asset class as some argue, then we have already provided a strategy index for it – the FTSE Forward Rate Bias Index Series, which is designed to capture the carry trade.
Raquel Pichardo-Allison: Andy, what are the biggest obstacles that pension funds face if they are looking to incorporate some of these types of indices into their portfolios?
Andy Barber: It is probably governance. If people are prepared to devote the time to considering the pros and cons of various alternatives to cap-weight, there are no serious obstacles.
Phil Tindall: There is an interesting point here, and it is predicated on cap-weighting being the starting point. The difference between picking an active manager and picking a strategy index is that there is a big, one-off decision that this thing will be better than market cap, whereas, in the old way of doing things, your consultant advises on an active manager for you and you do the monitoring etc. An implementation decision becomes a strategy decision.
Rob Arnott: There is a middle ground: that the consultant and the sponsor can take the view that the cap-weighted markets represent only one way to achieve beta, and that there are others. The cap‑weighted market can be the wrong choice at times. Therefore, diversifying their exposure offers upside, in terms of the opportunity to potentially outperform, and comparatively little risk from a governance or second-guessing vantage point, because these are broadly diversified, core‑like portfolios. If you put a third of your core into one or an array of these ideas, you have the opportunity to potentially improve returns noticeably, and not a whole lot of exposure to a risk of enough downside to encourage second-guessing.
Andy Barber: It leads to the interesting question of how people will look to benchmark active managers, because I do not suppose that many active managers would be terribly keen on being benchmarked against one of these alternative indices.
Andrew Buckley: That is true, particularly where those indices are designed to capture systematic alpha returns. What strategy indices show and the message that is key to understand here, is that the variety of index investment options available today is substantial, whether that is simple geographic exposure through for example the FTSE All-World or a robust alternative to active management. Active managers may not like it, but the divide between active and passive investment is blurring rapidly.
Rob Arnott: That raises a really interesting point: if people hate the idea of being benchmarked against these, because they fear that they are going to underperform or that the easy bit is being taken away, why on earth would this not be seen as a useful addition to the core investment toolkit?
Raquel Pichardo-Allison: Are managers willing to be benchmarked against more alternative-type strategy indices?
Rob Arnott: I am aware of one jumbo sponsor who has put his managers on a dual benchmark – cap weight and fundamental weight – from the vantage point that, if they underperform both, the sponsor is going to have short patience. And, if they cannot beat both indices and they are earning an active management fee, it is not clear that they are earning the fee. Needless to say, that is not well received by the managers.
Raquel Pichardo-Allison: David, have you had conversations with managers about this?
David Thompson: Not in respect of the equity indices, but we have designed bespoke benchmarks in other asset classes, such as bonds. Of course, the managers would like to measure themselves against a simple easy to beat benchmark, but our responsibility is not to keep the managers happy but to do what is best for our clients.
Kal Ghayur: Rob, is the manager doing well as long as they outperform one of the benchmarks?
Rob Arnott: If they outperform one of the benchmarks, they are watched; if they outperform, they are feted as heroes; and, if they underperform both, they are given a relatively short leash.
Kal Ghayur: To which index is their portfolio construction and tracking error relative to?
Rob Arnott: Cap weight.
Andy Barber: Have the managers adapted their process?
Rob Arnott: I do not know; I have not been part of those conversations.
Andrew Buckley: Is it a dual fee structure whereby the manager is paid one level of fees.
Rob Arnott: I doubt it; the manager collects their fees as long as they are employed.
Andy Barber: I do not like to feel sorry for managers, but in that case...
Raquel Pichardo-Allison: Looking ahead, where do you think that there is still a need for more product development?
David Thompson: Certainly, as we have already discussed, on the credit side we are doing a lot of work on developing bespoke benchmarks. That will evolve, rather than blindly following weightings and ratings on the credit side, which has clearly been shown to be unworkable over the last three years in particular. On the commodities side, there was a huge rally in 2007 and a huge collapse in 2008 in the indices, and there is probably some further work to be done there, although some excellent work is being done by a group of people. Rob and his colleagues are doing a lot of work on the equity side, and I am sure that that will continue, because there is so much money allocated to equities.
Rob Arnott: We have seen the early rounds or waves of innovation in the indexing community, and of a blurring of the distinction between active and passive.
What has not kept pace is the means for benchmarking results. We believe that there are more efficient forms of beta and that cap-weighted beta is not necessarily the only appropriate measure for beta. If you can achieve higher returns with valuation-indifferent indices, maybe active management should be judged against some sort of blend that includes fundamental-weight, equal‑weight, minimum-weight and cap-weight, or something like that. However, platform measurement and benchmarking have not made any serious strides commensurate with what we have seen in the index-innovation world.
The other area where there is ample room for further development is in finance theory. We have an entire industry – the indexing community – and an entire academic community built on what is a weak foundation – the foundation of efficient markets – at a time when nobody, even in the academic community, really believes in pure market efficiency.
Kal Ghayur: Finance theory is clearly a starting point, especially with respect to issues surrounding asset allocation. We need to go back to that. Other than that, however, in the index world, for a very long time asset owners have always complained that their asset managers are delivering betas and charging alpha fees for doing so. Therefore, one area is to have a much closer look at, and a much better understanding of the nature of, active management returns. We have proposed one solution, which is the FTSE ActiveBeta Indices and associated methodology, which provides useful insights into what gives rise to systematic sources of returns and how a very large portion of traditional active-management returns come, essentially, from these systematic sources of returns that we have misclassified as alpha in the past.
More importantly, index providers, along with some of us who are doing the research to create new types of indices, should really concentrate on customisation. Asset owners are really looking for the ability to start with something but then to be able to customise it to their own portfolio needs.
Andrew Buckley: We have talked a lot about evolution of strategy indices, and that is certainly one of the themes that we are seeing in the evolution of the indexing world. The other theme that we are seeing at FTSE is exactly the one that Kal has put his finger on: customisation. It is not quite as sexy as strategy indices, and sometimes it is difficult, but it is about creating precision benchmarks that are tailored to the specific needs of an investor.
Lionel Martellini: One of the things that we have been working on quite a lot is inflation hedging, which is an issue that goes back to the liability-hedging side of the question. The current tools, like inflation-linked bonds, have a number of problems: firstly, there is a huge capacity problem in most places, with some pension funds willing to buy 10 times the outstanding amount of inflation-linked bonds that can be found; there is also a very serious lack of performance in some sense.
There is also a significant interest in volatility indices. We have reason to believe that existing volatility indices have a number of shortcomings and that the theme of volatility as a factor is an important question. We are going to see more in this direction.
Phil Tindall: One area of interest for us is around things that hedge funds do; we do not necessarily want hedge‑fund replication, but some of the strategies that they use. We are quite close in terms of doing that – momentum and value applied across markets, for example, are the kinds of things commodity trading advisors or global macro-funds do. In terms of almost every strategy, you could create something for which there is an index; for example, convertible arbitrage etc. There are fund managers doing these strategies in a more beta-like way, but I like the model of starting with an index. It is a much cleaner route to go down.
Raquel Pichardo-Allison: Thank you very much, everyone, for joining. I hope that you found it to be an interesting and engaging conversation.
Raquel Pichardo-Allison, editor, Global Pensions
Rob Arnott, chairman, Research Affiliates
Andy Barber, investment leader of the global manager research team, MercerAndrew Buckley, executive director of strategy, FTSE Group
Khalid (Kal) Ghayur, CEO and CIO, Westpeak Global AdvisorsLionel Martellini, scientific director, EDHEC Risk Institute
David Thompson, head of manager research, Redington
Phil Tindall, senior investment consultant, Towers Watson
Canada Life has signed a £351m bulk annuity contract insuring the pensioner liabilities of 2,510 members and dependents in the AA UK Pension Scheme.
In this week's Pensions Buzz, we want to know if you believe there is ever a case for combining retirement savings products with other savings products, and if the PPF levy for sponsorless schemes is appropriate for DB consolidators.
The Insolvency Service has disqualified four directors of trustee firms from running companies for a total of 34 years following an investigation.