What good are hedge funds?

How can they beat the market consistently, especially if we take EMH seriously at all?  And if they don’t beat the market, how is 2-20 to be justified?  Here is a snippet from an interesting Amazon review:

…this kind of comparison misses the entire point of most hedge funds. A market-neutral fund is not designed as a stand-alone investment, but as a diversifier for an equity portfolio. It can have half the return of equities with the same volatility, and still be valuable. The question isn’t whether putting 100% of your money in hedge funds did better than putting 100% in stocks, it’s what portion of assets an investor should allocate to hedge funds. Using the author’s own numbers, an investor would have done best to have 30% of assets in hedge funds, rebalancing annually, from 1998 to 2010. That produced 4.2% annual alpha (return in excess of what you could have gotten investing in stock index funds and t-bills with the same volatility). That number is certainly overstated, hedge fund investors typically do worse than the index suggests, but it demonstrates that you can’t consider only stand-alone returns. This point is borne out by the finding that endowments and pension funds that make use of hedge funds have consistently better risk-adjusted performance than those that do not.

The review, by Aaron C. Brown, offers other points of interest.  I’ve ordered the underlying asset itself (the book) and I will report back on it.  It was reviewed in today’s FT, still no permalink.

Here is a recent story on hedge fund closures.


"How is 2-20 to be justified?"

Insider trading on Congressional secrets?

I would be interested in seeing a deeper analysis of the 30 percent number. The reviewer is absolutely correct that one should routinely diversify and rebalance out of equities. My biggest issue with the EMH is that it readily plays into the dogma that stocks can NEVER be overpriced and therefore investors should just keep plowing all money into index funds. That analysis may be correct for index funds vs. actively managed funds, but if all stocks are overpriced then putting money in bonds will do better.

The 30 percent number seems to be a portfolio that is 70 percent equities and 30 percent zero-Beta hedge funds. I can definitely how that would do better than a portfolio with 100% stocks. However, investing that 30 percent in bonds without the extreme 2-20 management fee structure probably would have done even better If that's what the author means by a portfolio with 30 percent hedge funds, then it is hard to see where hedge funds would fit in with a portfolio balanced with stocks and bonds based on the relative values of P/E and interest rates. In that case, then hedge funds really do only exist to let hedge fund managers play off of rich investors wishes to be in some exclusive club while charging 2 and 20. In the worst case, that's exactly what Madoff did. Even as he needed new investor money to keep the scheme going, he did an excellent job of making all the investors feel like they were in an exclusive club.

At the end of the day, investors need to have an understanding that investments only pay off if they have higher real, fundamental cash flows then just keeping money in Treasuries, including longer-dated Treasuries if they definitely do not want to sell early. Stocks at least have GAAP and all that so investors can get a guesstimate of their future cash flows by looking at P/E. Corporate bonds also have interest rates and ratings for credit risk. An investor should only become more sophisticated than that if they can personally understand how the hedge fund manager is making such outsize returns even after taking out 2 and 20, or for that matter how a hot stock tip would do better than a portfolio of bonds or an index fund. That means staying on the sidelines until they understand the nature of cash flows, even as the sophisticated hedge fund or the hot stock shows outsize gains, which is understanably very difficult. But it's also how the investor will maximize their long-term gains.

The argument for index funds is not really the EMH, but something that is only similar: It does not matter if markets are actually efficient, just that investment funds do not beat them in a consistent enough basis to outweigh the fees.

If some wizard could beat the market every time, but I could not realistically copy his strategy, markets would not be efficient, but my best bet is still an index fund, as I could not invest on the wizard.

The weak form of the Efficient Markets Theory they taught me in 1980 was that you couldn't beat the market without insider information.

You're describing semi-strong form EMH. Weak form EMH does not preclude beating the market with fundamental analysis.

"an investor would have done best..."

No. With the same hindsight I can come up with a of lot of tremendously better investments over that period. And what a period 1998-2010 was! I suspect any market neutral component would have lowered one's overall risk adjusted return over the 1982-1998 period. Are we supposed to be surprised that a market neutral investment might have done OK over a period when the market stayed neutral?

Yes, but if enough people believed the EMH, then it would lose the mechanism for it's own enforcement because the EMH believers would stop doing research.

Personally I think EMH holds in the short term and should be the default assumption for the long term, but that there are some exceptions -- systemic biases that are very hard to professional money managers to exploit that allow smart investors with a long time horizon to beat the market consistently.

For example the smart money skipped out on the dot com boom and bust, but imagine that you're fund manager in January of 2000. You've gotten killed by the index for the last three or four years, half you money has fled and you're probably lucky to still have a job. So if you're mostly investing your own money and don't really care what your investors think, you'll do fine.

So strong EMH would say that the dotcom bubble wasn't really a bubble because the smart money would otherwise have been making highly leveraged bets against the dotcoms and preventing a bubble. I don't agree because the reality is that there just isn't enough money with that long of a time horizon.


The "dotcom bubble" wasn't a bubble. There is probably no such thing as asset bubbles. Big booms and busts are probably better explained by increased uncertainty than bubble theory. Nobody had a good sense of where dotcom technology would lead us. Isn't it surprising that we haven't yet experienced much growth from the dotcom revolution? Were people fools because they expected the internet to be of much greater value to the economy than it has turned out so far?

Generate a random stock chart on a spreadsheet you will see plenty of booms and busts which look like bubbles growing and bursting but are in fact random movements. Try it! (For extra credit add in random but serially correlated changes in volatility (which describes real-world asset price movements better))

If the dotcom bubble is not a bubble, then bubble loses any meaning. For example, priceline.com was valued at more than American, United and Continental combined. Meanwhile, priceline.com's whole function was a financial intermediary in a market with no network effects and extremely low barriers to entry (travel sales).

The only way the pricing of Internet stocks made sense is if somehow:

1. Interest rates stayed very low for a long time.
2. The firms had almost pure monopolies on their markets as well as monopolies on their input costs, which was almost entirely the compensation of software engineers.

Furthermore, even priceline.com's competitors had the same stratospheric pricing! In other words, there existed literally zero circumstance where the pricing of dot.com stocks made sense. Warren Buffett and others said exactly this during the bubble and then, surprise, the bubble popped due to low or no cash flows exactly like they said.

Remember stocks don't just move up and down in stochastic fashion in a predictable way, like molecules diffusing in water. Stocks are in fact legal claims on the assets of a company minus their liabilities. The news that affects the asset prices may be somewhat predictably stochastic and so you would expect stochastic, predictable movements if prices reflected all publicly available information.

And that's the whole point of the tech bubble: under no circumstances could every tech company have monopolies on their market. And even with computers, consumers are pretty prickly about choosing the lowest price and pushing down margins, especially in areas like travel purchases. This was obvious to anyone who bothers to actually see if the NPV of the stock market made any sense. How the pure form of the EMH still has defenders is beyond me (I.e. that the market always reflects all publicly available information). Sure, prices are usually something like their fundamental value, but A LOT of times they're not.

I view bubbles as unsustainable spending based on massive leverage. The bubble was in telecom equipment, the dot coms were almost entirely equity financed, but came public much too early. I've read, that true dot coms were worth less combined than Ebay, Amazon and Google today, which wouldn't surprise me, everyone knew there were going to be real winners coming from the internet, no one knew which firms were going to be the winners.

The telecom equipment (and telecom service) firms that were all built on business models that relied being or supplying one of 20 or so firms who wanted to capture 10% of AT&T's business are today worth a small fraction of peak values (and responsible for most of the biggest declines in the NASDAQ indicies (worldcom, Nortel, Lucent, etc are all worth a very small fraction of peak market value).

Smith, Suchanek, and Williams (Econometrica, 1988) and many subsequent studies demonstrate how easy it is for a bubble to form in a transparently simple environment where everyone at all times knows the expected value of the asset. Asset bubbles exist. QED.

The difference between a bubble and a momentum trade may in the number of people who need to apologize to their clients.

For hedge funds outside of the equity space it's a lot easier to justify. There are large asset classes which are difficult to invest in. Hedge funds give you exposure to those investments.

the reviewer is certainly right that you have to look at their value as a part of a portfolio. but here's a question to consider: how many of these funds, charging 2&20 under the marketing guise of diversification, blow up when the $h1+ hits the fan? even market neutral funds, which claim they can escape systemic risk, blow up during market crises.

I'd like to see the risk stats; I find them hard to believe. Hedge funds can be extremely in terms of what they choose to invest in and can be flighty with the use of withdrawal gates. Phil Falcone's Light Squared is just one example of hedge fund investors not knowing what they were getting themselves into.

The story about hedge fund closures is, I assume, designed to draw the readers' attention towards survivorship bias. If, as the Economist claims, hedge funds open *and close* more often than other funds (and certainly more than index funds), then one has to be very careful when measuring the average returns of hedge funds.

So one serious problem is that the term "hedge fund" in at least some forums means "arbitrary investment mechanism". So some parties may gain from investing with hedge funds that don't invest in the US stock market at all. They may trade in private equity, distressed debt, privatization of government assets, etc.

A hedge fund that is strictly long and short stocks listed in the S&P 500 will have much more exposure to whatever the limits of EMH are than a fund allowed to buy any tradeable asset on Earth.

Other hedge funds claim to fame is generally being very close to market returns, with less variability - so they are more reliable sources of funds on any given day than an index fund.

EMH is basically bollocks, even in its weak form. Therefore, some funds and some traders and some investors have real alpha that persists for many many years and even decades. Renaissance Technologies, BridgeWater, DeShaw , etc. Systematic funds that employ algorithms. However, that does not mean that it's easy to find alpha or to realize alpha meaning turn it into actual profits. Therefore, many funds probably don't have alpha going forward, even if they did have it when they generated their past track record. So my advice is, do not invest in hedge funds unless

a) you know that a certain fund has alpha with extremely high probability and they actually accept you as an investor or
b) you have alpha at identifying which funds have alpha

Yeah but could you have predicted who the super performers would be beforehand? That's crucial.

Surely the leaders of those companies predicted it, they believed in their own success and made it happen.

So did the leaders of the companies that failed.

So you say that EMH (even weak) is "bollocks" but most funds won't have alpha going forward. What am I missing?

Some funds have true alpha, but there are many funds that don't
If EMH were to hold, no funds should have persistent alpha
The fact that alpha exists, doesn't mean it's easy to find

I think you don't really understand EMH. It certainly doesn't say anything about past alpha being impossible. Certainly someone must succeed due to chance alone.

How are you not just confusing average and marginal returns Tyler? They don't outperform the market on the margin, but they do on average. There are decreasing returns to portfolio management in the size of the portfolio. See Jonathan Berk's work on this.

"How can they beat the market consistently, especially if we take EMH seriously at all?"

So if a fund has been consistently beating the market, we should say 'But that's impossible, because ... efficient markets!' Traders didn't come up with the EMH, and most of the proponents (to paraphrase Ariel Rubinstein) don't quite know what a stock is.

How do you distinguish luck and skill?

Even if alpha is transient and gets traded away, there are certain individuals or teams that can consistently discover alpha faster than others.

There is bound to be alpha variance. Unless it is predictable in a systematic sense that doesn't help much.

Consider a model where alpha opportunities present themselves continuously though not predictably. You have many teams trying to discover and leverage the opportunities. Might it be the case that some teams are better at it than others? HFT is an example of this, by the way.

EMH is true in the long run. So is death. There's plenty to do in the mean time.

Ugh, anyone else extremely tired of the words "alpha" and "beta"? They're paeons to CAPM (and EMH by extension). I mean the very idea that a simple linear regression of returns on the market portfolio fully explains an asset or portfolio's returns is laughable. Everything other than what the market did is termed "alpha" -- we need far better parlance. Even go back to the original Fama-French paper; their regressions had r-squared of 0.7 or so at best. And we base our entire discussion of the stock market on that equation? Jesus.

well, we can add Size, Value, and momentum. If you find another factor that holds in the cross section, let me(and no one else) know. There's a lot of unexplained variation.

I thought hedge funds were simply what the name describes, funds that hedge your primary investment. If you're invested heavily in stocks, you invest part of your money in a fund that goes up when stocks go down. If you're invested heavily in markets that depend on oil prices being low, you put money in a fund that goes up when oil prices are high. Silly me if I'm wrong.

It's a little silly that people investors are flocking to hedge funds as primary investments, by their nature they should be way too risky as primary investments.

You are wrong, but you're error is instructive. The phrase "hedge fund" is a misnomer, and has been a misnomer for a long, long time. (Actually, a company might offer something like what you describe, as a side business, but that's not where the big fees come from. Google-search "currency overlay," for example.)

Hedge funds are simply investment managers that cater to sophisticated investors and are therefore minimally regulated in, for instance, their ability to use leverage. They earn returns one of two ways:

Category 1 (true alpha): They are just smarter than the market as a whole, and therefore obtain positive returns that are uncorrelated in the long-run to the returns of risky assets in general. An equity fund might, for example, have long positions in some stocks and short positions in other stocks, but have neither a long nor a short bias. This is extremely difficult (impossible if certain versions of EMH are correct without exception), so if a firm can pull it off then they probably deserve 2+20. My best guess is that a small number of companies really do do this, but most either try to do this and fail, or are Category 2 in disguise.

Category 2 (levered or exotic beta): They invest in risky assets, and therefore obtain a long-run reward for assuming that risk. These risky assets might be unusual assets (like emerging-market debt or CDOs or timber) that have small-to-negative correlation with investments like stocks and bonds that investors would already hold, or these risky assets might just be ordinary assets that the fund is able to lever up. Running this kind of fund may require expertise in trade execution, unusual markets, risk controls, tax optimization, etc. But conceptually it's an easy thing to do, and therefore should never command anything close to 2-and-20 fees (which doesn't necessarily stop people from paying that much!).

Category 3: Some mixture of Categories 1 and 2 — often, the marketing is mostly Category 1, while the returns are mostly Category 2. For instance, suppose a fund might have a strategy of borrowing in order to invest 150% of its equity in emerging-market stocks. To the extent that its portfolio's returns are just 150% of the EM stock market as a whole, it's just doing Category 2. But if it additionally engages in serious security selection in order to get an uncorrelated excess return beyond 150% of the whole EM stock market, then it's trying to do Category 1 as well. (Your actively-traded mutual funds fall into this category, although they can't use leverage.)

Hope that clarifies what it is that hedge funds do.

"your" not "you're" in line 1.

You are wrong, but you’re error is instructive

Public discussions (either online or on tv or elsewhere) would be greatly enhanced if more people started out their responses this way (by defusing reflexive defensiveness that all too often leads to closed minds in the followup discussion), and of course followed it up (as you did) in the same spirit of constructive criticism.

Characterizing hedge funds based on investment criteria is an exercise in dog chasing tail. "Hedge fund" does not define an investment approach. "Hedge fund" does not equate to a compensation scheme. A hedge fund is the absence of corporate governance. Discussions of hedge funds, therefore, only make progress when you address all the necessary principal/agent problems that arise.

Kyle's comment above, "For hedge funds outside of the equity space it’s a lot easier to justify. There are large asset classes which are difficult to invest in. Hedge funds give you exposure to those investments." is a good one. However, it poses a serious problem for someone trying to compute "volatility" of hedge fund returns and compare them to equity returns. How are hedge fund returns computed? Aside from the fact that some hedge funds don't report returns at all to the public, some hedge funds invest in exotic and thinly traded assets that it is not at all clear how they are valuing their assets to compute returns. Let's say no one should be shocked if they are valuing them in an optimistic and non-transparent manner.

Here's a pretty balanced view of hedge fund returns, adjusting for some of the index biases:


The absolute returns are not very high (7.70%), and there is a beta to the stock market of about .34 for an equally weighted hedge fund index.

The risk adjusted returns seems pretty good even after fees, but the portfolio gains quoted in the posting are surely overstated, unless they only replace bonds, which is for example what pensions should hold to match their essentially fixed liabilities.

IMO most of the demand from hedge funds does not come from some portfolio justification (although they may be sold to asset allocation committees that way), but as absolute return generators to pull underfunded pensions out of a hole. They don't seem to work that way very well.

There is also a serious return trampling problem, if you believe they are exploiting mispricings, as success will breed future failure.

The worst idea ever presented: risk-free rate....

I'm going to remain boring and buy index funds, which will continue to work great right up until enough people other people do it that it fails badly.

Consider the momentum factor, i.e. buying the top decile of stock performers from the last 60 trading days and shorting the bottom decile, then rebalancing daily based.

There is super strong evidence that this strategy makes produces positive returns over many different market conditions. It of course is uncorrelated to the equity market returns because it's equally long and short. Historical analysis says an investor should allocate 40% risk to market beta, 30% to the value factor (long low P/E stocks, short high P/E stocks) and 30% to the momentum factor. Historically this would return far superior returns to just holding market beta. There are strong economic explanations for this as well, so it's not just data mining.

However the typical investor cannot replicate the momentum factor in a retail account (transactions costs alone would be too high). Nor can he in a vanilla mutual fund because of short sale restrictions. The only real way for a typical investor to gain momentum exposure is in a hedge fund that can build the analytics, negotiate cheap commissions, and minimize transaction costs by using trading algorithms, dark pools, block trading, etc.

In this case the gains from diversification outweigh the 2-20 fees. Nor does this violate EMH, because momentum is just a simple risk factor, much like size, value, volatility, yield and carry.

Firstly, I wanted to link this story to the later Lin/talent story. There are many similarities to be explored in the HF business.

Secondly, comments on the HF business, especially with respect to alpha as a metric. Alpha is extremely difficult to achieve, but it exists (and is certainly not portable). It is used in the wrong context above, which is understandable as that is encouraged by HF marketing. Alpha only pertains to those trading the exact same universe as the benchmark. If you want to claim alpha to the SP500, you must only be long trading those stocks and nothing else. Also, any metric or combination of metrics that does not take into account drawdown is useless.

BTW, EMH is nonsense.

There are many examples the market isn't efficient. Here is one recent one:


That article doesn't even try to prove the headline.

In some circumstances EMH makes sense but never in the strong form as we can see from market-making activities, they clearly have inside information because they have access to level 3 quotes and they’re price makers not price takers. With that being said, they’re compensated for taking the opposite side of a trade to inject liquidity into markets – that’s perfectly fine and acceptable (note: Volker rule may have a negative effect on market making/prop trading of banks thereby increasing the risk premiums sought after by those legally required to take the opposite side). Secondly, EMH even in its strong form still allows for inefficiencies resulting from behavioral bias, liquidity needs and information asymmetry (which shouldn’t be the case, but it is) Even if the markets are perfectly efficient, a person holding a large position of a stock may affect the price due to a liquidation of stock to buy a house, an island, etc. We also see behavioral bias such that, people tend to sell winners and hold losers, they’re afraid to take a loss. It happens. And obviously information asymmetry, not even fundamental or insider trading but information that an individual and most professionals wouldn’t be aware of outside of a given strategy, I believe the year was 1997 or 98 or 2007-08 when a market neutral fund had a massive liquidation ( not LTCM, if that’s what you’re thinking) the first two days of the week had a sell off with the last days closing out the week higher, the down moves those first two days weren’t by any stretch of the imagination bad days, but for those funds in the market neutral strategy, you would’ve experienced a 5-30% loss that week as a fund. No one really knows why there was a massive loss on only market neutral funds. Some speculate the losses were the result of a liquidation of 1 portfolio to cover losses on another portfolio of illiquid investments. The case I’m making is, even when all information is known, there’s a chance that you need to be invested exactly like another person to suffer from the knowledge of their non-market priced portfolios.

Risk free rate is a great idea, but can only be said of near-dated US treasuries. Anything beyond the on-the-run bills have risk, some cases very, very little, but nonetheless still there.

Other arguments: size… size is what makes or breaks us, whether we’re a company, a fund or a guys ego. A $200mm fund is much more versatile in using manager skill to out-perform than a $30mmm fund which is better than a $65mmm fund. At a given strategy capacity, a fund should not take in any more assets and should build their brand on effective management. Alpha and beta are useful measures as is the lesser quantified opportunity cost of execution and mobility of a large fund. 2/20, I’d pay that if they could consistently net 15%/annum compounded. Some funds are able to very easily surpass 15, I know a manager who consistently does 60 we’ll say his sharpe is about 5, very dangerous! Anyhow, it’s possible to consistently outperform the market. And even more interesting is defining the market, retail investors use DJI, some professionals use SPX others use IXIC. Point is, define the market then determine an approximate yield necessary to achieve your results net of fees, find a manager (not john Meriwether) to manage per that strategy to achieve your goals and don’t give them 100% of your assets.

And internet stocks… yes that was a bubble, it was irrationality, it was a change in the composition of the index, it was mutual funds having to keep up with a momentum-driven benchmark. Someone mentioned leverage… leverage is so little in the cash markets. Fed Reserve tracks margin balances to determine leverage in the cash equity markets and let’s say its around 2%. That won’t cause PEs of 200 or 400 or.. n/a cause the firm has NEVER had positive cash flow!

From 1998 to 2010 the US market returned about 4.35% annualized (with a std. dev. of 20.35%). So the claim is the 30% hedge funds/70% market returned 8.55% with the same volatility.

But it's easy to come up with even better portfolios. For example 50% emerging markets/50% TIPS returned 10.35% over that time period with the same std. dev. (20.1%). Or 25% small cap value/25% emerging markets/25% T-bills/25% TIPS returned 9.28% with a std. dev. of only 12.4%, giving you higher returns and must less volatility.

But is this, as the writer claims, "alpha"?

If a HF consistently returned 15% (assuming a reasonable SD and DD), you won't be paying 2/20, you'd be paying a minimum of 3/30. (Yes, the higher fee would affect future returns.)

I have a friend that charges 2/20 on 60% returns. Hes to the point where he can stop managing OPM and the returns he makes on his own capital keeps him in the 1%. Although you are right, generally the higher the returns (assuming reasonable SD, DD and other ratios) the higher incentive fees. Last I read SAC was asking upwards of 40%, but the guys a legend.

Comments for this post are closed