Alpha Male has attended more hedge fund conferences than he cares to remember. Many of them have begun with several empty seats and ended with far more. But apparently the good citizens of Geneva know a hot financial topic when they see one. You know all those seats along the back wall for late comers? All packed. You know the aisle – where you walk – to get to your seat? Also packed (with extra chairs that had to be brought in). The main conference hall of the Hotel “President Wilson” in Geneva was overflowing yesterday morning as Professor Bill Fung of the London Business School kicked off this two-day gabfest on hedge fund replication. Thankfully, it appears the Geneva fire marshal must have been off having a chocolate eclair at some swanky cafe by the lake.
Why all the interest? Hedge fund replication – that esoteric and highly quantitative discipline that had struggled for attention only a year ago – has suddenly hit the mainstream.
But rather than freaking out about it, it seems that many hedge fund operators have embraced the old enemy and have positioned hedge fund clones as a complement, not a substitute, to traditional hedge funds. For example, Fung himself told the audience:
“Hedge fund replication is here to supplement existing hedge fund strategies by providing greater liquidity and transparency.”
As a result, many (actually most) of the speakers yesterday came armed with their own offerings. We found nothing inherently wrong with this. But it did make me long for the days when a small minority of speakers pitched their wares while the majority just wanted to share some cool research about the behavior of hedge funds. But I guess that’s what they mean by the “institutionalization” of the hedge fund business.
There was a lot of talk about hedge fund indexes, even though the event was not actually about that topic. In fact, speakers seemed to jump seamlessly between hedge fund replication models and hedge fund indexes – as if the two were synonymous.
And I guess they are synonymous to those who are in the replication business to produce alternatives to existing hedge fund indexes (e.g. JP Morgan). But to many (e.g. Fulcrum Asset Management), hedge fund replication is more than a tool to match the average hedge fund return – it’s a tool to produce alpha.
But for those who aimed to simply represent the average returns of hedge funds, the question then became: which index are you trying to clone? Several speakers pointed out that the difference between different hedge fund indexes could be as much 4% per annum. So a passive hedge fund replication technique with a high correlation to one of the hedge fund indexes might also have a relatively low correlation to another of the major hedge fund indexes.
Hedge Fung Replication
Fung believes that the reason hedge funds exhibit such a high correlation in times of distress (e.g. August, for some strategies) is not because they share the same assets, but because they share the same liabilities. In other words, Fung says “they all borrow from the same investment banks”.
Ergo, says Fung, any successful hedge fund replication model must integrate elements of the hedge fund industry’s liabilities, not just their assets.
Fung also presented several interesting slides showing the performance of the major hedge fund clones over the summer. We’ll try to wangle those from him tomorrow. But the bottom line is that after a pretty good first half, most lost a tiny bit in June, a bit more in July and even more in August (although they recovered from an average drawdown of several percent at mid-month).
“No Desire to Replicate the Downside”
Gavyn Davies, former head honcho at the BBC, Goldman Sachs banker, columnist for the Guardian newspaper and the founder of Fulcrum was one of the attendees that expressed no desire to fully replicate the average hedge fund. Taking a more active and liberal view of the term “replication”, Davies illustrated to the audience how his “risk stream” and “active trading” approach blew the pants off the traditional hedge fund indexes.
Davies also had some choice words for funds of funds. According to his research, it would take a fund of hedge funds about a 20% gross return to produce the same net return as a passive replication product with a 14% gross return (mainly due to the lower fees of most clones).
On the other hand, argued Davies, factor replication models “have trouble with market turning points” since they generally look back 24 months or more to calculate the appropriate factor weightings for the next month. Therefore, when the tech bubble blew up, (back-tested) factors models didn’t clue-in to the fact for over a year.
In contrast, he says, his “Fulcrum Alternative Beta Plus” (FAB-Plus) model reacted much quicker to such events and therefore had a much lower market correlation than many factor models.
Still, Davies echoed Fung when he told the audience that alternative beta funds and real funds of funds “will both have a big future”.
JP Morgan Stanley
The Fung & Hsieh-advised JP Morgan Alternative Beta Index (ABI) then went head to head against Morgan Stanley’s new “Altera” offering (Alt-era. Get it?). Both funds were given a performance-agnostic spin. JP Morgan’s Lakshmi Seshadri and Morgan Stanley’s Yazid Sharaiha both emphasized that their objective was to mimic the hedge fund universe, not to beat it. Sharaiha said his idea of a great hedge fund replication model was one that worked “out of sample” (i.e. one that actually explains hedge fund returns in different time periods). And Seshadri said that the “alpha-centric” industry needed an objective benchmark.
A Natural Extension
Philippe Schenk of Credit Suisse Tremont Index LLC rounded out the morning by proposing that “replication is a natural extension of the index business”. But he also asked why anyone would want to develop another product when most already produce much of the alternative beta embedded within investable hedge fund indexes.
In answering his own question, Schenk said that “now we can focus on trying to replicate the (better performing) non-investable hedge fund indexes.”
The Contrary View
Just when everyone – funds of funds and clones alike – were starting to get along, the afternoon speakers rocked the boat.
Dirk Sohnholz of FERI Institutional Advisors teed things up by accusing providers of hedge fund clones of cherry-picking the benchmarks that made their products look best (whether or not the clone’s objective was to beat or simply match the indexes). Reflecting a modest level of cynicism, Sohnholz told the audience that “marketing-wide, hedge fund replication is a great story.”
But if Sohnholz teed it up, BGI’s Stan Beckers hit the 400 yard drive – landing it right between the eyes of the hedge fund replicators gathered in front of him. In a blunt tirade that some in the audience actually admitted was refreshing in its candor, Beckers railed against any form of hedge fund replication and predicted the industry would all but disappear in 5 years. He saved his best barbs for the hedge fund indexes that replicators aimed to mimic, complaining that the industry was “targeting an illusionary, flawed and poor reflection of the hedge fund universe.”
Fung agreed and raised questions of his own about the accuracy of hedge fund indexes. However, the CISDM’s Hossein Kazemi (who ran such a database) said his research indicated that hedge fund indexes were actually pretty accurate reflections of the hedge fund universe after all. An audience member – I think it was Deutsche Bank’s Tony Chapple – told Beckers that the problem of the missing mega-funds was essentially solved by mimicking a fund of funds index since such an index would essentially reflect the performance of the world’s largest (non-reporting funds) via the funds of funds in their databases.
Flying Dutchman Fights Back
Harry Kat, the human lightning rod for strong opinions on this topic, then addressed the gathering and responded to “10 Unjustified Criticisms of FundCreator” (his distributional replication tool). We’ll leave the details for a future posting. But suffice to say, the panel had mixed feelings about Kat. At one end of the spectrum, Fung continued to attack things such as the model’s reliance on monthly data to make daily trades and the fact that an investor might have to wait 2 years or more to find out if FundCreator was actually working properly.
At the other end of the spectrum was Nicolas Papageorgiou of HEC Montreal Business School. Papageorgiou studied under Kat and along with his client, Canadian giant Desjardins Global Asset Management, he is a proponent of distributional replication (you may recall his paper on the topic).
In the middle were Kazemi and author Nassim Nicholas Taleb (speaking tomorrow) who both expressed reservations about distributional replication, but never really took a swipe at the Kat.
And that’s how things ended up on day one. With impassioned opinions, colourful debate and not an empty seat in the house.