Movie Stars, Super Models, and Alternative Beta

The worlds of filmmaking, fashion and finance converge on London this week. Dame Helen Mirren was crowned queen of the world at last night’s BAFTA awards (Britain’s Oscars), London Fashion Week has flooded this city with (highly controversial) size-zero models, and Professor Bill Fung presides over IRC’s Hedge Fund Replication and Alternative Beta conference today. Only able to attend one of these three events, Alpha Male eschewed the movie stars and super models and chose instead to spend the week with hedge fund stars such as Fung, David Hsieh, and Lars Jaeger. And while there were no catwalks or super models, we did enjoy plenty of super factor models.

While this event is ostensibly about hedge fund replication, it really speaks to alpha-centric in all its forms.  Fung, who coined the term “alternative beta” in a chapter of a 2003 book edited by Lars Jaeger, kicked off the day by asking the audience, Why pay 2 and 20 fees for what you can do yourself?  Whether it is possible to “do it yourself” remains to be seen.  But regardless, this is the central question in alpha-centric investing today.

The Failure of Investable Hedge Fund Indexes

Fung’s research corroborates what the industry has sensed for some time now: that investable hedge fund indexes under perform their non-investable peers.  Even when adjusted for management fees, the investable CSFB, HFRI and MSCI still under performed the non-investable versions of these funds (see recent post on one manager who royally panned these investable products).  Fung proposes another method of capturing alternative betas that might not succumb to the certain balls and chains that seem to stymie the existing investable indexes. The manacles to which Fung refers include: high fees, opacity, illiquidity, and hedge fund risk

Hedge Funds Already Cloned?

One of the first studies to explore hedge fund cloning was written by Mitchell & Pulvino (Journal of Finance, 2001). They showed that merger arbitrage could be approximated by a combination of long positions and put options. Fung’s own research shows a remarkable high correlation between one particular long-only mutual fund in the United States (called The Merger Fund) and the HFR Merger Arbitrage Index.  He shows the same correlation between emerging market hedge funds and the IFC Emerging Market Index.  Finally, Fung’s research shows CTAs are highly correlated with 10 year treasury volatility (i.e. they go up when the 10 year is either up significantly or down significantly).

Fung makes the point that replicating the average hedge fund leads to lackluster results since it ignores the sub-segments of funds most worth looking at and captures funds that are so bad, they eventually fall victim to poor performance.  But on the other hand, he says, actively picking the sub-segments in a replication strategy is tantamount to managing an active fund of funds – the very products hedge fund replication aims to replace.

Fung made another interesting observation. He asks, If someone could perfectly replicate the returns of the best hedge funds or strategies, then why would they charge you less than those active funds?  He is basically pointing out that the hedge fund replication industry is obviously not motivated by altruism and is instead incented to set fees only marginally below the prevailing market rate.

In our view, the assumption that replicated hedge funds will necessarily be cheaper than real hedge funds is based on the assumption that traditional hedge funds price their services based on their costs.  But the high profits earned by hedge fund managers and the extreme scalability of their business models clearly shows that hedge fund fees are currently based on their value to investors, not their underlying costs. So why would these new players want to set bargain prices for their services?

After all, this is what movie stars and super models do.

– Alpha Male

Be Sociable, Share!

2 Comments

Leave A Reply

← Conference Notebook (Day Two) Alternative Beta Conference Notebook →