Smart beta: Smart beta's Tower of Babel
A while back, the £5.9bn Merseyside Pension Fund decided to terminate exposure to a RAFI-benchmarked global equities strategy after five years of lacklustre returns. Performance matters in investing, arguably more than anything else. But while Peter Wallach, head of the fund, won’t be the only early adopter of smart-beta strategies without a terrific set of performance figures to show for his faith, he is very clear that the exit was not, in fact, primarily driven by returns.
The RAFI product was originally put in a portfolio of opportunistic investments built to house strategies outside the main asset categories of bonds, equities and property. When Merseyside changed the benchmark for the Opportunities portfolio from equity beta to cash-plus, the strategy simply no longer fitted. The value tilt in RAFI had far less meaning in a LIBOR context.
While outperformance would no doubt have given the strategy a greater chance of survival (Merseyside has kept Unigestion, a smart beta-style player from the active management world, for the same length of time), the moral of the story is that benchmarks and objectives matter, too.
The dilemma in the world of smart beta and factor indices is, therefore, which benchmark to choose. If market-cap portfolios have proved to be disappointing in recent decades as an investment strategy – because, as the argument goes, index-tracking is not some kind of neutral strategy but actually an active decision containing all kinds of biases – does it make sense to retain them as benchmarks?
And if smart beta portfolios do not promise regular outperformance but can indeed be prone to periods of lagging the market, does that make it more or less confusing to compare them to ‘passive’ versions of themselves, in the shape of strategy-specific indices? On the face of it, that would seem to bring some clarity to the mandate and, just as important, give managers the confidence to realign their portfolios and maintain style consistency even through the tougher times.
To put it another way, using a smart-beta benchmark means that the asset owner’s original investment convictions are, or at least should be, healthily embedded in her agent’s benchmark.
To some extent, the proponents of smart beta have only themselves to blame for both the lack of new benchmarks and the ever-rising Tower of smart-beta Babel. Certainly in the equity space, almost everyone recognises that smart beta and factor indices are active strategies: market-cap does ultimately represent the practical universe in spite of all its biases.
And yet the adoption of the term “index” for these active strategies has been widespread. From a marketing perspective, the word adds kudos. It elevates a strategy to some kind of standard. On the downside, ordinary investors and trustees don’t know what is what. The industry needs to do a better job distinguishing between active, passive and what lies between.
“We have a duty to be clear on what these strategies are and what they can do,” says Baer Pettit, head of the indexing business at MSCI. “There is a danger of the market getting flooded with thousands of strategies on backtests that may or may not have the backing of sound investment theory.”
For now, however, the complications are not causing widespread concern. Pension funds employing smart beta, be they factor tilts or risk-based formulae, seem content to use the appropriate market-cap index as a guardrail, if not their full benchmark.
De Eendragt, the Dutch pension insurance company, for example, uses Parisian manager TOBAM’s Maximum Diversification strategy for about 75% of its equity portfolio. De Eendragt CEO, Philip Menco, notes that annualised since 2008, the strategy has delivered market-cap-like returns for about two-thirds the volatility, in line with the Parisians’ initial pitch. The primary objective of harvesting the equity premium while saving some risk budget for elsewhere in the fund has been achieved.
In the case of De Eendragt, however, this metric is not enough. It uses value-at-risk, Sharpe ratios and a risk-premium approach to further assess the absolute risks it is running. As Menco puts it: “You can’t eat relative returns.”
So outperforming a market-cap index provides some helpful insights but its flaws, notably the lack of diversification exhibited during the 2007-08 financial crisis, are what brought the likes of De Eendragt away in the first place (the Dutch fund adopted this strategy, which attempts to construct the most-diversified portfolio by maximising the ratio of its portfolio constituents’ weighted average volatility to overall portfolio volatility, in January 2008, eight months before Lehmans collapsed).
TOBAM’s charismatic co-founder, Yves Choueifaty, indefatigably reminds us that this leaves the strategy without an obvious benchmark – indeed, he likes to refer to his maximum diversification concept as “the anti-benchmark”. In one sense, Menco agrees, in that his team does not study the Maximum Diversification strategy for factor or sector biases.
The acknowledgement, however, that one has changed the very definition of risk for the majority of one’s equity portfolio brings us to the question of governance. Whatever the extra, detailed metrics and modelling used by pension fund executives, investment advisers and third-party managers, there is no doubt that one reason market-cap benchmarks still shadow smart beta is that trustees are less adventurous in their thinking than the likes of Choueifaty. Menco and his peers across Europe still hold comparisons between anti-benchmark strategies and MSCI or FTSE in mind for the sake of explaining the role of the new to fiduciaries and scheme members.
This brings us back to the role of pension funds in incorporating smart beta. If the commercial players are responsible for some of the cluttered confusion thus far, asset owners could untangle much of the marketing jargon by leading benchmark reform. This harks back to the point that third-party managers would follow the endorsement of a smart beta index by a client. When such an index and any benchmarked strategy underperformed the market cap, not just the manager but the asset owner would have to face the critics for their decision. And of course, better-governed pension schemes would have thought ahead to such scenarios.
With the exception of Denmark’s PKA and a few other big schemes, however, asset owners have not been prepared to recalibrate their risk-and-return metrics so profoundly. For those funds, such as Merseyside, which have hired a smart-beta strategy alongside traditional managers (in European equity, Unigestion is employed alongside JP Morgan Asset Management and an in-house team), the preponderance remains diversifying against a traditional index and between active styles, rather than inventing a separate paradigm.
A similar message comes from Irish Life Investment Managers in Dublin, where Shane Cahill heads indexed fund management.
“We did a lot of research into smart beta in 2011 and ended up selecting MSCI Minimum Volatility and RAFI as two strategies to track,” he says. “Some of the preparatory work we did was around how the strategies fared in down markets. So, for example, we analysed relative performance and participation ratios in bull and bear markets, as well as the size and duration of drawdowns.”
Irish Life IM was not looking at the two strategies in combination. Like many other providers, it wanted to offer something credible and new to clients who’d been burned by the fallout from 2008. Thus far, Cahill sees two different types of buyer. First, are very large DB plans with the resources to try something new, albeit not necessarily in large amounts of their total risk budget. The second group are DC plans in which smart beta accounts for up to 25% of assets. But in both cases he sees little demand for either strategy-specific benchmarks or any other alternative to market-cap.
The importance of all of this has become very clear over the past 12 months in particular, and the past three years in general, as both the most popular strategies, low-volatility and fundamental indexing, have had mixed fortunes. A superficial comparison on performance numbers alone with the global-cap index, for example, doesn’t make the MSCI Minimum Volatility index look great.
“Minimum volatility returns since Irish Life Investment Managers launched its product are in line with historical expectations,” says Cahill, “but the distinctions between the alternative indices and market-cap need to be explained to clients.”
Evidently it is not the case that smart beta is a simple solution in a volatile asset class such as equity. Some asset owners would like to benchmark them against traditional market-cap indices but in periods of dull-to-disappointing performance it is necessary to accept that alternative indices require alternative forms of measurement, such as participation and Sortino ratios. Just like smart betas themselves, these measures are not new. They have been used on a daily basis by risk management teams in the middle office for decades. Now they must have wider appeal, most especially as the ‘grace period’ for smart beta comes to an end.
Ten years ago, hedge funds could promise the world: they were an alternative to the long-only equity exposures that had gone so wrong in the tech bubble. Likewise today, smart beta can be said to be attractive by contrast: it is not associated with any of the instruments, strategies or organisations that fell in 2008.
That is hardly a perfect or sustainable endorsement. Hedge funds have enjoyed mixed fortunes since their arrival in institutional investing. Will the same be true of smart beta? To know one way or the other, the first thing institutions need to do is arrive at a suitable benchmarking and monitoring framework.