Hedge fund engineering

The New Yorker's John Cassidy has a great piece on hedge funds here. He makes a number of important points, largely based on academic research.

1) most funds don't generate alpha, net of fees. Hey, what happened to the efficient market? Why, then, is so much money flowing into hedge funds? If qualified investors (multi-millionaires) can't sort out the alpha question, who can? Apparently, it's left to a few academics...

...only eighteen per cent of the funds outperformed their benchmarks, and returns even at the most successful funds tended to decline over time. “Our research has shown that in at least eighty per cent of cases the after-fee alpha for hedge funds is negative,” Kat told me. “They are charging more than they are adding. I’m not saying they don’t have skill; I’m just saying they don’t have enough skill to make up for two and twenty.”

2) most funds are implementing a strategy that can be replicated much more cheaply (previous discussion)

3) although the very best funds do generate alpha, most of them are closed to new investors

4) finance professors, although among the best paid academics, make an order of magnitude less than financiers :-(

The article also discusses a program called FundCreator, which seeks to replicate any particular fund based on vol and correlations. The general theoretical framework is nice, but the idea of cloning, say, SAC Capital, seems nutty given changing market conditions and limited statistics.

However, Kat remained skeptical. As he conducted his research on hedge funds, he became convinced that it might be possible to generate similar returns in a mechanical way and with much less effort. Two years ago, he and Palaro began to sketch out ideas for a software program that could mimic the returns of individual hedge funds by trading futures. “We may be able to do without expensive hedge-fund managers and all the hassle, including the due diligence, the lack of liquidity, the lack of transparency, the lack of capacity and the fear of style drift”—changes in a fund’s strategy—“which comes with investing in hedge funds,” Kat and Palaro wrote in a working paper about the project which they published last year.

Kat provided many of the mathematical ideas. Palaro, an experienced programmer, did most of the computer work. Rather than trying to emulate a hedge fund’s monthly return—a nearly impossible task—the researchers sought to match the fund’s results over a period of several years, as well as the other statistical properties of its performance that investors were likely to care about most: the volatility of the returns, their correlation with the stock market, the likelihood of suffering extreme losses.

In the spring of last year, Kat sent me an e-mail in which he expressed confidence that he and Palaro would succeed. “It is possible to design mechanical futures-trading strategies which generate returns with the same, and often better, risk-return properties as hedge funds,” he said. “This means investors can have hedge-fund returns but without the massive fees and all the other drawbacks that come with the real thing.”

By the end of 2006, Kat and Palaro had finished writing their software program, which they called FundCreator, and have conducted several successful trials...

...Some scholars remain skeptical. “As a renegade statistician, I am a little bit suspicious of Kat’s methods,” Stephen Brown, of N.Y.U., said to me. He pointed out that, unlike Quantum, many hedge funds have been around for just a few years and there is little information about their performance. “On the basis of very limited data, it is a real challenge to construct an accurate and robust model of hedge-fund returns,” Brown said. Andrew Lo said that using FundCreator may not be as straightforward as it seems. “From the point of view of theory, there is nothing wrong with what Kat is doing,” he said. “But all dynamic trading strategies involving derivatives carry some risk. They rely on very specialized mathematical assumptions. If the assumptions turn out to be wrong, you can be mis-estimating the risks in a big way.” Faulty assessments of risk contributed to major financial losses suffered by Long-Term Capital Management and several other companies that have encountered problems trading derivatives.

Veryan Allen, an investment adviser and former hedge-fund executive who writes a blog about hedge funds (hedgefund.blogspot.com), said in a post last December, “If Goldman Sachs, Dow Jones, Merrill Lynch, Andrew Lo, or Harry Kat think they can do it, great . . . but I suspect investors will end up disappointed if they think the returns from hedge-fund clones will be anywhere near the performance of the best hedge funds.” Allen went on, “No matter what occurs in the markets, well-managed ‘expensive’ hedge funds operating proprietary strategies with skilled traders, robust risk management, and technology will perform, even under pessimistic economic scenarios. . . . That is why it is worth paying the two and twenty. . . . Average or generic hedge funds can certainly be replicated, but not the best hedge funds.”

Also, some interesting stuff about Renaissance and others:

It is notoriously difficult to distinguish between genuine investment skill and random variation. But firms like Renaissance Technologies, Citadel Investment Group, and D. E. Shaw appear to generate consistently high returns and low volatility. Shaw’s main equity fund has posted average annual returns, after fees, of twenty-one per cent since 1989; Renaissance has reportedly produced even higher returns. (Most of the top-performing hedge funds are closed to new investors.) Kat questioned whether such firms, which trade in huge volumes on a daily basis, ought to be categorized as hedge funds at all. “Basically, they are the largest market-making firms in the world, but they call themselves hedge funds because it sells better,” Kat said. “The average horizon on a trade for these guys is something like five seconds. They earn the spread. It’s very smart, but their skill is in technology. It’s in sucking up tick-by-tick data, processing all those data, and converting them into second-by-second positions in thousands of spreads worldwide. It’s just algorithmic market-making.”

Almost by definition, there can be only a handful of genius investors, Kat continued. “And even if they are there, the chances that you will find them and that they will let you in are very, very slim,” he said. “That’s what I tell people. If you are really convinced that you can find those super managers, then don’t waste your time with our stuff. Go look for them. But if you are a bit more realistic, if you know that eighty per cent of hedge-fund managers aren’t worth the fees they charge, then the rational thing to do is to give up trying to find a super manager, and just go for a good, efficient diversifier instead.”

Not so long ago, Kat recalled, one hedge-fund manager, a “global macro” investor who specializes in betting on currencies and stock markets around the world, approached him with an offer. “He said, ‘Harry, I want to buy your thing so I can replicate myself. Then I’ll be able to enjoy life a bit more and keep sending my clients bills for two plus twenty. It’ll take them years to figure it out, if they ever do.’ ”