He describes a friend, labelled Greg, who manages a superb long/short hedge fund. Greg is tough and obsessive, eats too much, and screams at analysts who do shoddy work. "Above all," Biggs says, "Greg doesn't like to be wrong or lose money anytime, anywhere, anyhow, and when he does he goes nuts."
The world's investors are flooding into hedge funds in search of managers like Greg who they hope will make consistently good money for them, regardless of where markets are going. Google the term "hedge funds" and you'll get 10.4 million replies – about 20 times as many as five years ago. And for good reason. Many hedge funds have performed stunningly well for a long time, and their clients wouldn't consider investing anywhere else. Some excellent managers have delivered 20-25% per year over a decade or more, with maximum drawdowns (cumulative losses) in single digits. Unfortunately, most of these Greg-like superstars close their funds to limit their size and keep returns healthy. So would-be investors lack access to some of the best hedge fund talent. But while hedge funds as a group are impressive, their past returns have not been as strong as is widely believed, and the future is not entirely rosy. While hedge funds are part of the investment answer for most people, they're certainly not the whole answer. Investors need to understand their strengths, limitations and likely future returns, and position them in portfolios accordingly.
Many a hedge fund presentation begins with the graph below, or a close relative. The solid line on the graph shows the cumulative value of $100 invested in hedge funds since the beginning of 1994, according to the CSFB/Tremont database, one of the best in the industry.
The results are spectacular. It looks as though hedge funds as a group made money amazingly steadily over this period, at 10.2% per year, or 7.7% in real terms (after inflation). This graph is as close to a smooth line as one finds in investments, and rises even during the equity crash of 2000-02. If you believe returns like this will be available in the future, then the implication is clear: sell everything, invest in a diversified portfolio of hedge funds, and relax. But this conclusion is flawed, for a number of reasons. One is simple: survivorship bias. Hedge funds voluntarily submit their numbers to databases, which can lead to various shenanigans when they run into trouble. If your fund blows up and loses 63% this month, will you tell Tremont all the shameful details? Probably not – you'll have far more important things to worry about. Or maybe you'll instruct the databases to delete your fund's entire history. Dead funds are remarkably quiet. Academic studies of survivorship and other statistical biases indicate that they boost headline returns in hedge fund databases by about 3% p.a. So the 10.2% from earlier falls to 7.2% or thereabouts. Not bad, by any means, but no longer spectacular. A related problem is that one couldn't have invested directly in the CSFB/Tremont Hedge Fund Index. Instead, one would have had to invest either via a fund of funds (with costs of 2% or so per year), or via a range of underlying funds – and had to spend time and money to monitor them and ensure they didn't blow up.
To make matters worse, the golden days for hedge funds as a whole may be past. Taking into account survivorship bias, which we conservatively estimate at 2.4% per year, the CSFB/Tremont hedge fund index rose at 9.3% p.a. 1994-2000, and at 5.5% p.a. between 2001 and July 2006. Why the fall? The earlier period was uniquely favourable for hedge funds – global equities soared, inflation fell, political risk seemed subdued and the world economy internationalised steadily. All these factors weakened a little in the later period.
Moreover, researchers from J.P. Morgan have shown that huge inflows of capital into the hedge fund arena have eroded returns in some strategies. Hedge funds are now big players in many markets, and it's becoming harder for hedge fund managers as a group to beat the rest of the financial sector. According to Vanguard's John Bogle, back in 2001, "I think it's inconceivable that you could take $500 billion run by 6,000 different managers and expect these managers to be smarter than the rest of the world." Since then, the hedge fund industry has more than doubled in size. Bogle is probably exaggerating – the managers in question can be smarter than the rest of the world, especially when managing risk … but only slightly. In short, the bigger the pack of hedge fund wolves roaming the forest, the less each of them eats. In some areas, like convertible arbitrage, hedge funds at times have made up half or more of total activity in the sector, and end up trading with each other much of the time. It's tough to make money under such conditions: wolf eat wolf is not a great business model.
Looking ahead, then, the hedge fund industry may deliver mildly disappointing returns. It's most unlikely to blow up, but may gradually grind down. Unless you're exceptionally lucky or skilful, then, it doesn't make sense to have 100% of your portfolio in hedge funds.
So what are hedge funds good for? A great deal, actually. One shouldn't expect miracles from them, but if they're used selectively they can contribute to almost any investment portfolio.
Analysts often write about hedge fund alpha and beta, where beta is the portion of the fund's return that is linked to markets (equities, bonds or whatever), while alpha is the return that's independent of market movements – and can be ascribed to manager skill. Investors can buy most kinds of beta cheaply via index funds, exchange-traded funds and the like, but really value the alpha or absolute returns that hedge funds can provide.
A number of academic papers have analyzed historical data to investigate the determinants of hedge fund returns over time. They find that a majority of hedge fund returns can be explained by market movements. One study found that around 60% of net (after cost) hedge fund returns between 1995 and early 2004 could be explained by returns to US stocks and bonds. The authors reckoned, in fact, that funds grabbed almost half the gross alpha they generated via their hefty fees. Though these studies may be biased for technical reasons, we agree that markets are pretty efficient, and there isn't an enormous amount of alpha out there to be captured. Moreover, for the finance industry as a whole, alpha is almost zero sum, so investors should be cautious of relying on it too much.
And hedge funds are expensive. With fees of (say) 1.5% annual management fee and 20% performance, sometimes more, they can cost five or ten times more than long-only funds – especially index funds. So hedge funds are most valuable when they provide returns streams or portfolio protection that you can't get elsewhere. The way we see it, hedge funds can play several roles in portfolios. Two especially intriguing ones are as True Diversifiers, and as Return Enhancers.
Many hedge funds move quite closely with equity or other indices, and don't diversify conventional investment portfolios much. But some funds follow unique investment strategies and their returns have little or no relation to the market. These are the most valuable to long-biased investors. A fund that moves negatively with equities, for example, but makes money in its own right can be a wonderful hedge for a long-only portfolio.
The problem, we find, is that True Diversifier funds of this kind are scarce and often don't make much money. In principle, hedge funds can add the most value to conventional portfolios by having low or negative correlations with conventional assets. But only a few strategies, like managed futures trading and fixed income, offer consistently low correlations of this kind – and these tend to be the lowest-return areas of the industry. Using the CSFB/Tremont data referred to earlier, the relation between hedge fund strategy returns and the correlation to a mixed equity-bond portfolio since 1994 is shown in Fig 2.
In general, hedge fund strategies that don't move with conventional assets – to the left of the graph – earned somewhat less than the higher-correlation assets on the right, as indicated by the regression line. Moreover, the graph excludes (equity) short sellers, which had negative correlations with conventional assets – and lost around 5% p.a. over the period. They probably earned positive alpha, but their beta to equities was large and negative. No wonder: very few equity hedge managers consistently make money on their short book, and for most it serves merely as an expensive hedge.
So true diversification comes at a price. Funds that are truly uncorrelated with equities and bonds may earn some alpha but lack the beta kicker and struggle to earn even modest, bond-like returns. It seems that funds that take on a moderate amount of beta (market risk) are likely to earn the best returns over time – the kind of returns that equities have delivered in the long run. Nonetheless, uncorrelated funds can benefit most portfolios, despite their modest returns, by smoothing out drawdowns and enabling more risk to be taken overall. True Diversifiers may be valuable to investors, even if they don't earn much.
A second role for hedge funds is as Return Enhancers. These are funds that combine alpha and beta in clever ways, so the outcome is better than could be obtained elsewhere. Some funds manage to combine solid returns with selective market timing and good downside risk management, such that they earn as much as conventional assets when markets (usually equity markets) are strong, and lose less when markets are weak. Often this entails getting the Big Picture roughly right: investing heavily when markets are trending upwards, and being conservative or net short when they're falling. Over five or 10 years they can beat equity indices by far, even after fees. Indeed, the most successful hedge fund strategies and funds seem to combine alpha and beta components in the long run.
This conclusion is reassuring, in a way, because beta is persistent but alpha is (near) zero-sum, hence if you're chasing alpha alone, you have to be consistently cleverer than the competition – perhaps increasingly so, as markets become more efficient over time. Active beta management – seeking promising markets at roughly the right time – can be a large and persistent source of alpha.
The best hedge funds develop remarkably cunning ways to make money and avoid losing it – in areas ranging from exotic derivatives to high-frequency trading to equity long/short. Barton Biggs writes about a friend who's excited about Iranian consumer goods companies. Why? Because the country has 70 million future consumers, fairly widespread education, some democratic institutions, and a westernising population. This is a highly contrarian view, of course, but one that may pay off eventually. Moreover, if he gets his timing (beta) right and begins investing early enough, in the right areas, the risk of losing money should be small.
Conventional long-only investing tends to be tightly constrained, with managers facing strict limitations on where and how they can invest. By contrast, hedge funds have the freedom to do almost anything they wish. Via gearing and shorting, trading exotic derivatives and markets, and operating with very broad investment mandates, hedge funds have the opportunities to make huge amounts of money – or lose huge amounts when they go wrong.
For example, Amaranth Advisors, a $9.25bn US multi-strategy fund, announced in September 2006 that it had lost 50% in the last few weeks on ill-advised natural gas trades. Despite their size and record over several years, they'd lost sight of the First Law of Risk Management – don't put too many eggs into one fragile basket.
Between 1998 and 2001, for example, the average hedge fund in the CSFB/Tremont database earned 11.2% per year. But this average conceals a wide range of returns. Almost five percent of the funds on the database lost more than half their value over this period – and a handful (some of which vanished from the database) lost everything, either because their risk management proved hopelessly inadequate (e.g. Long Term Capital Management) or because of fraud (e.g. Manhattan, a big US shorting fund that went bust in 2000). This is one of the reasons why headline returns data over-estimate hedge fund returns and alpha.
So manager selection with hedge funds matters far more than for conventional asset classes. In the manager-selection arena, we focus on managers who are likely to deliver positive alpha (who have understandable, repeatable ways of making money, and where we understand the downside risks.) The latter is crucial – there's not much worse than getting to work in the morning, opening your newspaper and discovering that your favourite fund lost 30% last month and is still struggling.
For private investors, funds of funds are usually the best way of getting exposure to hedge funds. Goodfunds of funds do an enormous amount of due diligence on their managers, to make sure they won't blow up, and construct diversified portfolios covering various strategies that should make money through most market environments. They function as shock absorbers to smooth our portfolio returns when equity markets are falling, while mostly providing some upside when markets are strong.