In fairness to Garry he did win the first series against IBM’s ‘Deep Blue’ in 1996 (4-2), only to lose in 1997 (3½ -2½); the first time a computer program had defeated a world champion under match tournament regulations. You can read about it here or you can watch a video about it here.
But what has a chess player losing to a supercomputer (programmed by some pretty smart people) got to do with the future of discretionary multi strat hedge funds I hear you ask? A lot, you hear me answer.
In this blog, I use the term ‘discretionary multi strat’ alternative funds to explain those products where teams of people devote their time and resources to fundamental research across asset allocation, manager/strategy selection, operational DD, and administration and governance etc, which are then packaged into convenient ‘all in one’ portfolios. In contrast, the ‘systematic multi strat’ funds create ‘all in one’ portfolios but devote their resources to exploiting statistically robust anomalies, taking a ‘systematic’ approach by using big data and massive investments in IT (e.g. people, hardware and software), removing many human biases and beliefs, and where the implementation of trades are executed with the aid of supercomputers, in real time, and with little human intervention. The humans, in this case, spend more time on data integrity, statistical analysis of various strategies, position keeping systems and in managing servers in their data centres etc.
Both approaches aim to offer significant diversification benefits by exposing investors to streams of lowly correlated returns that help to smooth the path of returns for investors – an important feature, as the advice market transitions (slowly) to objective/goals based advice.
But both approaches have vastly different return outcomes, as I highlight below in Table 1.
Disruption is everywhere you look
Not a day passes by without hearing or reading about the powerful forces of technological ‘disruption’. In financial services, most of the emphasis has been in relation to how robo advisers might disrupt traditional wealth management, how blockchain will crater investment banking and stock exchange earnings or in the banking sector, with peer-to-peer lending bypassing the need for banking channels to obtain credit etc.
Disruption is essentially about creating customer value (e.g. lower cost, safer flying, better manufacturing etc) at some point, in, or across, an existing industry value chain, or indeed, by creating a new value chain altogether.
But away from the media hype surrounding the death of the banks and financial advisers, a quiet revolution has been going on in the management of multi strat hedge funds, as the world’s best physicists apply their rigorous techniques to systematically capture lowly correlated mispriced anomalies, assisted by the ever growing power of supercomputers.
As Table 1 below highlights, the discretionary multi strat products managed by teams of people attempting to asset allocate and choose best of breed managers/strategies, has generally done a poor job of delivering against their product promises, with the exception of the Ibbotson Diversified Alternatives Trust. If the BlackRock retail fund can replicate the longer term performance of their wholesale fund, then it will also be included in the exception list – there is no reason to think that they cannot.
But when compared to groups like AQR with their Delta product and CFM with their alternative beta portfolio ISD that use supercomputers to systematically capture mispriced and lowly correlated market anomalies, the discretionary segment has been left for dead – completely disrupted into what might become future irrelevance. Like the discretionary segment, both AQR and CFM provide their IP in the form of diversified alternative solutions but here, they are offered at much lower cost, are liquid and transparent, and have far less ‘forecasting error risk.’
What is most fascinating about the discretionary multi strat segment is that there is wide spread acceptance of the disparity of returns between various hedge fund managers, within their field of expertise, supposedly making it easier to select best of breed managers or strategies – the evidence, on the surface, does not support this. (And hello survivorship bias – the Select Alternatives and van Eyk Blueprint Portfolios are no longer in the peer group). Knowing both funds well, I can safely assume that they would not have made it into the exception category.
Table 1 – Smart Human Failing to Deliver v Smart(er?) Humans Using Supercomputers that Do Deliver
* The BlackRock Fund has a wholesale vehicle that has produced a stellar 10.15% p.a. gross of fees return since inception in July 2004
** Marketed by Winston Capital (for good reason)
Obvious weakness in Table 1, include the different inception dates, no data on correlation outcomes (not provided by most managers), the loss of a number of competitors and the short time periods for some funds, however, LHP has been in Australia for over 14 years. Its 10 year return is just 1.57% p.a., which is well below its inception date return (highlighting strong early year performance, as recent periods have been average) and well below the RBA cash rate over that same period.
A packaged alternative solution makes a great deal of sense
Conceptually, the idea of partnering with a fund manager to select the best fund managers/strategies makes a great deal of sense, given the global nature of the alternative fund market and the sheer complexity of the asset class. Having people scour the world looking for alternative fund manager talent, given these factors, seems to make sense, but at what point will fund managers who offer these discretionary portfolios ask themselves “how much longer should we offer these underperforming portfolios to the public?”
And I should know. As the previous CEO of Select Asset Management, we used to market a discretionary ‘all in one’ alternative portfolio managed with a small team here in Australia that was well rated by research houses and listed on numerous dealer APLs and model portfolios, but we worked out that to offer better value to our clients, we needed a global team to manage it, and so, given that expense, we transferred the management of the fund to Neuberger Berman, but as Table 1 shows, this has not started off well.
It is no surprise that Alternatives are under represented in retail portfolios
Based on our estimates, less than 4% of retail FUM is allocated to the alternatives asset class, with most of that invested in managed futures – strategies that are great for providing tail risk hedging (hello supercomputers!) but not so great at providing returns when in sideways markets, characterised by intra period volatility or at critical market turning points.
But is it any wonder, given the results of the discretionary multi start funds to date that advisers have avoided these funds – the FUM in some of the funds is below breakeven. For advisers that have used them, they have been on the back foot ever since with their clients. For those that have exited altogether, many now use cash as their ‘alternative’ asset – the path of least resistance. Cash doesn’t blow up, is easy to explain, is liquid, preserves capital etc – explainability plays a big role in product choice but I’m arguing to not throw the baby out with the bathwater (who would do that BTW?) and take a fresh look at the disrupters like AQR and CFM, as a compliment to their managed futures positions (slight overlap). Or indeed, as a new way to re-enter the asset class but with more confidence.
Many advisers are scared off by what they consider to be ‘black box’ investments
What’s scarier? A plane that takes off, flies and lands by itself carrying 100s of passengers with the aid of sophisticated software programs, or programs that systematically capture lowly correlated mispricing opportunities that actually work using the same computational power?
Neither are scary; they are both simple examples of how technology is used to reduce human error and do a better job – both still need smart people to make them work though (for now…).
And far from being ‘black boxes’, these new strategies are ‘white’ in colour – we know what goes in, and we know what comes out – it is not a mystery, as full disclosure is offered to investors. The ‘trick’ as it were, is in trade execution and data management i.e. connecting the PhDs, with the supercomputers, to the market exchanges – this is where massive amounts of capex is spent by the likes of CFM and AQR, and the managed futures providers in Winston, Aspect and Man AHL.
For example, CFM collect 2.5TB of data each day (the equivalent of a typical academic research library), have more than 1,500 servers, monitor over 1,000,000 instruments on a daily basis, nearly half of their employees are IT engineers, and in total, they have invested over €100m in sophisticated IT systems over the past decade. AQR would have made similar investments in delivering their expertise to market, if not more, given their size.
Needless to say, the barriers to entry in the systematic multi strat segment are as high as the discretionary multi strat funds – perhaps higher (disruption does not come cheap here).
Where to from here?
It is my hope that advisers take a fresh look at the diversification benefits offered by the systematic multi strat funds offered by the likes of AQR and CFM, where supercomputers, programmed by smart PhD, consistently deliver on their product promise of generating returns from lowly correlated mispricing anomalies, thus, providing important diversification benefits to client portfolios who are looking for smoother pathways in the management of their wealth.
In a world where fixed income returns are so low, providing little real income and with the future potential of rising inflation and hence capital losses on fixed rate bonds, advisers need to consider how best to achieve portfolio diversification at low cost, and in a liquid format. Fixed income may not deliver like it has done in the past.
And as M. Turnbull might say, “there has never been a better time to invest in diversified alternative strategies that actually work for investors.”
Andrew Fairweather, Founding Partner