This week’s FTfm (the FT’s financial market supplement) ran an article headlined Hedge fund gains are other funds’ losses (access reserved to subscribers). It finds that hedge fund managers outperformed over the past thirteen years and that this outperformance was “financed” by a corresponding underperformance of traditional asset managers. This would be very good news for the somewhat beleaguered hedge fund industry if only it were true. But despite a valiant effort by the author, which includes using results based on those of a popular book and citing a Nobel prize winner as his authority for the interpretation of the main result, the article’s main conclusions are simply wrong and the seemingly precise performance results essentially meaningless. This article is a rather drastic example of what can happen when misunderstood theories are applied to empirical results. It is this kind of study that has driven some observers to the view expressed in the title. One would have expected the FTfm to know better.
The centerpiece of the article is a comparison of the net performance of a broad hedge fund index (the DJCS asset weighted composite index) and of what is called “market return”, arbitrarily defined as a mixture of US and UK equities, for the 1999-2011 period. The initial conclusion of the article – that hedge funds outperformed developed market equities over the past 13 years – is hardly front page news. But the real problems arise from the use that the author seeks to make of this result.
He claims that his combination of US and UK blue chip equities is a sensible proxy for the capital market as a whole. In other words, he is perfectly willing to ignore every other asset class under the pretense that hedge funds invest mainly in equities – itself a rather loose statement. The period under consideration was a miserable time for developed market equities but other important asset classes that are also open to traditional asset managers such as emerging market equities and debt, small caps, credit, both investment grade and high yield – to name but a few – did much better. Only if the author’s combination of US and UK equities were truly the only asset class available, would it be justified to interpret the outperformance of hedge funds as evidence that their managers are able to add value, i.e., that they generate skill-based “alpha”. But given that there is no shortage of other investment opportunities, it is a pure nonsense to insist that the difference between hedge fund returns and the article’s “market return” is due to managerial skill. Most or all of the difference – particularly at the level of a broad index – is likely to be due to the exposure of hedge funds to other, better rewarded directional bets.
As the author correctly states, Bill Sharpe (and many others) have argued that the quest for alpha by active managers is a zero sum game. However, Sharpe and the others were all talking about active bets within an asset class. But it is nonsensical to argue that active managers of a specific asset class collectively underperform just because some other asset class had better returns, as the author does in this article. The relevant measure of success for managers focusing on a single asset class is whether they outperformed a suitable benchmark; one that is representative for their investment universe. Thus, the claim in the article’s title that the underperformance of the US and UK equity markets directly financed the outperformance of hedge funds is simply ludicrous. Of course, it is possible to seek alpha with active (tactical) asset allocation decisions but this is not what the article is about. To summarize, the author arbitrarily assumes that equity risk is the only risk premium that matters, that hedge fund returns are dominated by alpha, and that, consequently, traditional managers must have been the losers, hedge fund managers the winners. But as Merton Miller, also a Nobel laureate, once claimed, it is surprisingly easy to employ a sufficiently ridiculous set of assumptions and test them using inappropriate statistical tools to find almost any desired “result” in finance. The FTfm article is a most impressive illustration of that fact.
There is a second major issue in the article: the claimed level of hedge fund returns. The author talks about a “Hedge Fund Community” (HFC), made up of investors, managers and prime brokers. Starting from net returns to investors, he imputes gross returns, assuming a 2+20 fee structure and that prime brokers receive three percent for their services. This leads him to postulate an HFC return, i.e., the total return that goes to this so-called community, made up of investors, managers and prime brokers. According to the author’s calculations (more on those later), this HFC return is more than double the net return. It is this return that he compares to the “market return”. However, the HFC return is of no interest whatsoever to investors. They are interested only in that part of the HFC return that they receive. If managers and prime brokers were to extract more than 100 percent of the value added, clients would find no consolation at all in the magnitude of the gross return. And on this point, the evidence is quite clear: both academic studies and the results of hedge fund replication strategies show that, as a group, hedge funds add no value. Hedge fund investment is at best a zero sum game after fees, on average, just like traditional active management. Hedge funds, as a group, have been winners in terms of gathering assets and remuneration but their clients, again on average, have not benefitted from the managers’ success – there is not much community spirit in that HFC.
There are a number of additional issues with the article. Moving from net to gross returns as proposed in the article only works if all hedge funds that make up the index have positive returns in years when the index has positive returns and vice-versa. Otherwise, the hedge fund fees would be a higher proportion of the net return (this is the well known fund-of-fund fee problem). The author claims that his numbers reflect money-weighted returns. But simply multiplying hedge fund returns by industry assets for every year and dividing the sum of those dollar returns by the average industry assets will not do. If we are interested in money-weighted returns, an internal rate of return calculation would be required. But it is actually not clear why money-weighted returns should be of any interest in the context of the message that the author seeks to convey.
The one interesting result in the article is actually the information on the sharing of the generated gross returns. On average, hedge fund investors receive less than half of the gross returns generated. If we consider that risk premia from directional bets make up most of the gross returns, then it is difficult to escape the conclusion that hedge fund managers extract more than their value added in the form of fees. Consider a simple example: if a hypothetical hedge fund generates twelve percent gross return, its investors receive eight percent net with a 2+20 fee structure. Now assume that the cash return is two percent and that a further six percent are risk premia harvested through strategic exposures to directional bets, which leaves four percent for alpha. Obviously, since the manager extracts all of the value added as his fee, investors might be better off with a replication strategy, particularly if one considers that they probably paid a fund-of-fund manager a additional fee to get access to this hedge fund. The point here is simply that the average hedge fund is likely to be as mediocre as the average traditional active manager. Successful hedge fund investment requires skills in selecting managers and that is far from easy. But the point of this post is not to discourage investors from considering hedge funds. There are a number of good reasons why they ought to give hedge fund investment serious consideration, assuming that they have access to the requisite skills. But the arguments put forward in the FTfm article are simply not among those good reasons.