Separation of smart beta strategy development from index implementation makes it unclear who is accountable from a fiduciary perspective, argues Jeremy Baskin
The past few years have seen widespread acceptance of, and investment in, smart beta strategies – often expressed as equity-risk-premia products. As an investment management firm that offers both alpha and beta solutions to the marketplace, we are wholehearted supporters of smart beta investing, believing that these strategies can be valuable tools for asset owners. But the many conversations we’ve had with asset owners, index providers, and consultants – paired with the considerable amount of press coverage on the subject – have given us reason to reflect on some of the important ambiguities of smart beta investing and highlight a particular observation related to fiduciary responsibility.
Is smart beta actually beta? Generally, these strategies are thought of as, alternatively, generic beta or ‘cheap alpha’, but there still does not seem to be a consensus as to where, precisely, they belong on the beta-to-alpha continuum. We believe that ‘alpha’ points to a focus on performance relative to a cap-weighted benchmark. This perspective is not consistent with the original goals of smart beta investing: a more mean variant-efficient view of the world that speaks to a Sharpe ratio focus, rather than information ratio.
But, the further we get down the spectrum of customisation and blending, and the more managers are asked to time risk premia, the closer we get to alpha strategies. Ultimately, our ex-ante definition may be less important than determining the expectations for the strategy in the aggregate portfolio. If the funding for the smart beta strategy is from the truly passive asset pool, and the expectations are stated in terms of total return and total risk, then the investment is being required to act as ‘beta’. Similar logic would apply if the strategy was replacing an alpha strategy with a sensitivity to tracking error, or was introduced with the expectation of such.
Even if we believe that these strategies sit firmly in the beta camp, we should not fall prey to the temptation to assume that risk premia, once defined as beta, are somehow generic, or even self-evidently transparent. Definitions can be highly subjective, heterogeneous, and often very complex, as evidenced by disparate risk-and-return outcomes for strategies that sound identical on the surface. Take, for example, the 12-month performance differential between the MSCI USA Minimum Volatility index and the S&P Low Volatility index. While both are published ‘low risk’ indices, they exhibited a nearly 5% performance differential for the one-year period ending 30 June 2015, resulting from non-trivial differences in the way each index provider defines risk and adjusts for undesired exposures.
Is smart beta investing passive or active? The industry seems to be coalescing around the idea that smart beta ideas are, indeed, active ideas. Irrespective of the approach, every smart beta strategy is purposefully weighted in a way that is meaningfully different than a purely capitalisation-weighted representation of the market. Since the investment universe that all investors start with is the cap-weighted market, decisions to deviate – regardless of method or purpose – are active decisions that lead to active strategies. But while there seems to be broad agreement on this point, we frequently hear reference to passive allocations to smart beta. In an attempt to reconcile these two points of view we find it helpful to distinguish the investment strategy from its implementation, either or both of which can be active or passive.
Using an example familiar to all, an ETF that replicates the performance of the S&P 500 index would tick the box on a passive underlying investment strategy, implemented passively. Within this framework, a smart beta index would represent an active underlying investment strategy.
In some cases, this active strategy is also implemented actively, with both functions being performed under the same roof. This pairing of active strategy with active implementation is recognisable to anyone familiar with active investment management. However, nearly all smart beta ETFs take an ‘active strategy, passive implementation’ form, with an index provider creating the active strategy and an index replicator implementing a portfolio to track a third-party smart beta index.
An index replicator’s job is to faithfully mimic the performance of an underlying index and is typically evaluated on the basis of tracking error to the index (the lower the better). They are not paid to question the investment logic behind the thing that they’ve been hired to track. Efficient implementation is what is promised, and what is expected by the asset owner. So who bears responsibility for the investment behaviour of the underlying index itself? Not the index replicator. Arguably, this responsibility falls to index providers, although most third-party providers go out of their way to emphasise that they do not take fiduciary responsibility for accuracy or investment outcomes.
So who bears the fiduciary responsibility for the effectiveness of a strategy? This question is perhaps the greatest area of ambiguity in the smart beta area, as well as the greatest area of concern: where does accountability for investment outcomes reside? It is a seductive proposition to believe that an active strategy can be labeled as ‘beta’, indexed, and implemented passively in a straight forward fashion. Acknowledging that these are active strategies, whether or not they represent alpha ideas, we argue that they should be managed and monitored in a manner commensurate with their risk. This can be difficult since many smart beta investment choices are built by one entity with another firm implementing the ideas. The separation of strategy development and implementation leads to significant ambiguity as to who, exactly, is accountable from a fiduciary perspective, which can create risk for asset owners. The extent of this risk may not be appreciated by investors who think of risk premia investing as a thoroughly passive activity.
In sum, we believe that smart beta strategies provide an attractive, cost-effective, and risk-efficient source of equity returns, often with the added benefit of imperfectly correlated outcomes. However, passive implementation of active investment ideas introduces lack of clarity as to who bears fiduciary risk for investment outcomes. As investment managers who both develop and implement our own smart beta ideas, we are admittedly biased.
It is no secret that we see risks when strategy development and implementation are separated, hence our advocacy for keeping them fully integrated. Regardless, we believe that an open discussion of the ambiguities introduced by smart beta strategies will benefit asset owners, investment managers, consultants and index providers.
Jeremy Baskin is global CEO and CIO of AXA Rosenberg