Smart beta: Blurring the lines
Remember when Compaq, Digital, Hewlett-Packard and Intel ruled the world? In the glory days for these companies, they were the champions of IT. But the pre-eminence of software over hardware handed the laurels to Microsoft and then the internet gave rise to a new generation of giants.
Apple and IBM diversified to survive the cost cutting that inevitably reduced computer hardware from a premium to a basic spend. The others named above engaged in mergers to survive, or switched models. Down both avenues, the costs in the IT hardware business were wrung out.
At an indexing conference this winter, a trustee for one of those pre-internet leviathans’ UK pension schemes told me exactly the same process had to take place in asset management. He saw it as inevitable. And he ought to know.
At the conference smart beta was the thème du jour. So the obvious question to pose is whether smart beta, factor indices and other strategies that deliberately weight portfolio positions by market capitalisation are the tools with which to cut costs.
“[With] active fees and cap-weighted benchmarks it is challenging to perform without a significant risk appetite”
They are now available ubiquitously. EDHEC-Risk Institute’s Scientific Beta is giving investors free access to the methodology and profile of tens of leading strategies, created in- house. The number on offer is planned to grow to one hundred later this year while, for a mere €10,000, ERI Scientific Beta will build you a customised index.
Even if one chooses a more familiar index name such as MSCI, S&P or FTSE, the annual licence and management fee for smart beta all-in could be in single digits (even though turnover for strategies such as minimum volatility are approximately eight times those of a broad market-cap index). Even where active managers have devised their own proprietary mechanisms, fees still find themselves closer to passive than active. AXA Investment Managers, which has gone so far as to trademark “Smart- Beta” as a single word, will charge institutional clients 15 basis points for its credit and 30 basis points for its equity pooled funds. Segregated mandates would be cheaper still.
So far so good. It sounds as if smart beta is having the same profound impact on asset management that disruptive technology such as the internet had on the IT industry. Commenta- tors such as Roger Urwin, advisory director at MSCI, have hailed the shake-up as “a revolu- tion” and early adopters of these strategies
such as De Eendragt, an insurer housing more than 20 Dutch occupational pension schemes and former European Pension Fund of the year at the IPE Awards, have reported six years of improved risk-adjusted performance.
So there is some proof of their efficacy.
|At a glance|
|• And smart beta index providers need to make the argument for purer factor exposures more forcefully.|
|• Smart beta is extending the ‘pile it high, sell it cheap’ paradigm beyond the market index.|
|• But growth is still from a low base and most asset managers have interests vested in the old active and passive models.|
|• This is leading to a re-branding of traditional-looking active styles as part of the new trend, which can result in obfuscation.|
|• New importance needs to be attached to active managers’ governance and transparency around smart beta-style products.|
But there is still a long way to go. Take MSCI itself, an organisation that has been doing quite nicely creating and licensing smart betas in the shape of factor indices. More than 90 organisa- tions worldwide now subscribe to its Factor Index module while approximately $75bn is benchmarked against its factor indices, of which Minimum Volatility is probably the most well-known. But this is hardly disruptive. Urwin himself admits that, while the conversation on smart beta has gone viral, its adoption has been “spotty”. Certain funds in Europe, such as PKA and PZFW, have drawn much attention for their profound implementation of smart beta, but they remain exceptional. MSCI does not currently earn more than $20m (€15m) a year from factor indices. But this part of the business has grown eight fold over the last two years, so that revenue – as for the actual smart beta asset managers – must surely expand.
Bill Gates once observed that technological change is overestimated over two years and underestimated over 10 years (think of the impact of 3G mobile technology). Baer Pettit, head of the index business at MSCI, says that factor indices could account for 25% of his unit’s revenues in five years’ time but he expects it to be 15-20%. He is more concerned that providers and consultants give prospective clients dispassionate and objective messages about the nature of these new products.
“The fundamental laws of portfolio theory have not been suspended,” he says. “These factors are tilts which might well lag the markets for long periods of time.”
And so he concludes that the future holds plenty of room for passive, active and smart beta in between. “Factor indices sit in this grey area and still have to justify themselves but there is strong economic logic and sound investment theory underpinning them. They have a lot to show.”
Fifty shades of grey
From Pettit’s comments it starts to become clear why smart beta may not be a clear-cut conduit for cost-cutting in asset management at all. There are disruptive players in this market such as EDHEC and Research Affiliates, whose success is built on smart beta strategies and which do very little else. There are houses such as AXA IM and Robeco genuinely prepared to be disruptive, as well as the big indexers such as BlackRock and SSGA.
“The fundamental laws of portfolio theory have not been suspended. These factors are tilts which might well lag the markets for long periods of time”
But most participants, just like MSCI, are well-established businesses wary of cannibalising existing revenues from either plain-vanilla passive or active management businesses or both. There are few Mark Zuckerbergs in the world of asset management: goofy upstarts with nothing to lose by selling only the new.
There are, on the other hand, a lot of suppliers keen to retain their market share, energetically developing their own versions of smart beta to rapidly populate this grey area. A lot of active houses with any quant capability are busily dusting down those systematic strategies and deciding on a new name to suit the zeitgeist. Blue Sky Group, a Dutch fiduciary manager, reckons it holds 50 strategies in the low-volatility family alone on its database.
It is not just Vanguard that sees little substance in the trend. Frank Umlauf, co-founder and partner of Frankfurt-based investment advisory firm, Tadjo Consulting, points out there are always fads in investing. “We have seen them in the past: 130/30 or portable alpha,” he says.
He adds that investors are only tilting their allocations more towards value and size factors with many smart-beta products, and that this was already achievable with existing style indices.
James Hughes, group CIO of HSBC Insurance, with worldwide financial assets of $90bn (€67bn), does not agree. “My memory of combining managers stylistically is that in a growth market you’d be worrying about your value manager and in a value market you’d be fretting about your growth manager,” he says. “Add in active fees and cap-weighted benchmarks and it becomes increasingly challenging to perform without a significant risk appetite.”
HSBC Insurance has taken advantage of one of the new equity strategies on offer, handily supplied by an affiliate, HSBC Global Asset Management. It is a major commitment, not a toe in the water, applied to a significant amount of the group’s equity portfolio. The strategy’s long-term aim is to deliver two-thirds of the volatility of developed global equity markets, measured by standard deviation, and in the 20 months since its inception has done that with aplomb.
Much of the framework to the strategy seems like smart beta. It begins from a belief in the low-volatility anomaly and its long-term measure of success turns away from short-term market benchmarks: the market-cap index is not the prime benchmark but a useful, secondary comparison alongside the MSCI Minimum Volatility index.
But when Hughes explains how the portfolio is put together – by ranking the universe according to criteria such as return on capital, enterprise value and EBITDA, followed by further refinements by HSBC Global Asset Management’s analysts – it sounds like a standard active management style; quant at heart but with a proprietorial overlay. Vis Nayar, global head of investment research at HSBC GAM, created the strategy while Hughes addressed the specific requirements of the insurance company regarding its equity allocation.
The issues with such strategies – and it applies to numerous others including AXA SmartBeta – is how much the proprietary processes enhance or detract from the smart beta. One could even be more precise and ask how much they complicate clients’ ability to interrogate and comprehend how the strategy is likely to perform, both relative to other strategies and by absolute risk metrics, once a mandate has been awarded.
These are inevitable lines of tension as smart beta transfers from academic theory to real products. The academics on smart beta do not tend to introduce proprietorial twists; if so, they tend fully to disclose the workings. In general, they are also reliant on commonly available sources of data, which suit their efforts of justifying their theories in open territory.
Commercial providers, on the other hand, tend to keep the recipe for their secret sauce hidden while at the same time promoting that sauce as a means of differentiating themselves.
Here is a simple example of what differentiation can do to similar smart beta strategies even when they are extremely transparent and originating from the same methodology, but one key ingredient apart. Compare the returns from FTSE RAFI Europe and Russell Fundamental Europe. Over 10 years to the end of December, FTSE RAFI had achieved annualised returns of 8.53% for 22.79% annualised volatility while Russell Fundamental had notched up 9.96% with 21.21% volatility.
As spelt out on the Research Affiliates website, FTSE RAFI’s four company-fundamental portfolio weighting criteria are book value, cash flow, sales and dividends. Russell’s are: retained cash flow, adjusted sales and dividends plus buybacks.
So the explanation of at least some of the differential is easily defined (there are also differences in the universes of stocks). Nevertheless, definition and beliefs are not the same thing. Fiduciaries who recognise the varying outcomes of these indices have by consequence to put in plenty of thought before selecting between them. ‘Fundamental indexing ’ is not comfort enough.
In the brave newish world of smart beta, even superficially similar offerings require greater examination and analysis. This is not a question of asset manager cost but almost cer- tainly more spend by asset owners on internal governance, strategy research and subsequent monitoring. Such subtleties as whether the inclusion of book value is essential in assessing a company’s wealth creation are not evident in most of the introductory literature on smart beta, even though the likes of Urwin have long emphasised its attendant governance demands.
Such subtleties are, however, common differ- entials among active houses. They are evident in the proprietorial adaptations for smart betas as exemplified by HSBC GAM’s Low-Vol strategy. Which brings us back to doubts that smart beta might not be so disruptive after all, given the overlaps with traditional investment strategies.
“[Volatility] is one factor a lot of sophisticated investors already perceive as an unwanted risk”
There are plenty of active equity managers in business today who are very aware of factors and attempt to reduce their exposure to them as part of their process. The GAM Euro- pean equity strategies run by Niall Gallagher in London would be a case in point. Gallagher is a bottom-up stockpicker who wants as much company-idiosyncratic risk in his portfolio as possible. So he and his colleagues use standard regression analysis to strip out, or at least reduce, unwanted factors. The techniques are exactly the same as for building factor indices. A clean divide between shiny new factor-based and traditional equity management does not exist.
So where does all this leave clients? Russell Investments still believes there are plenty more ways that factor awareness can sharpen up portfolios. Crispin Lace, a director in the pen- sions solutions group at Russell, believes that whatever factor management occurs within any one strategy, holistically most clients could still benefit from further work.
As an example he points to the overwhelming evidence that active managers as a group bring an overexposure to volatility to clients’ portfolios. “This brings down your Sharpe ratio and it’s one factor that a lot of sophisticated investors already perceive as an unwanted risk,” he observes.
Russell has found plenty of ways to assist clients to remove this risk – a simple method is to buy a low-volatility index as a complement to the existing roster of active managers. There is further analysis to be done about whether one reduces passive or active management to do so. Most of Russell’s advisory clients are large and sophisticated. In the US they use both active and passive, so what to exchange for that lower volatility is a moot point. But Lace is optimistic that the understanding and analysis around styles and factors has grown phenomenally over the last 20 years; so the right tools are more readily available.
In conclusion, there is no doubt that smart beta will apply pressure to the cost of pension asset management. This has not occurred in a vacuum; many retirement schemes fret over fees paid to third-party agents not least because they are being watched more by regulators and participants. Costs, unlike investment perfor- mance, are certain.
However, not only is this trend confused by the variety of products on offer; comparison and attribution becomes difficult as each invest- ment house serves up its own adaptation, often combining factors and sometimes overlaid with qualitative checks. There is absolutely nothing wrong with this per se – there may indeed be a lot right with it. The flexibility introduced helps avoid, for example, the front-running of an index and brings factor diversification. It also leaves asset owners and their advisers with a whole load more questions to ponder and deci- sions to make.