Correlation chart of the day, hedge-fund edition

By Felix Salmon
November 21, 2011
Mark Gongloff finds this astonishing chart in a Morgan Stanley note this morning.

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">


Mark Gongloff finds this astonishing chart in a Morgan Stanley note this morning. It shows the degree to which hedge fund returns, in aggregate, are correlated with our old friend the S&P 500. And it turns out that even though the correlation has never been higher, it’s still somehow rising.

Now this isn’t all hedge funds — it’s just looking at the HFRI Equity Hedge index. That index covers, broadly, any hedge fund which invests mostly in stocks — but is entirely agnostic as to whether you’re long or short. It includes market-neutral funds, which aspire to be uncorrelated with the market; it also includes short-bias funds, which aspire to a negative correlation with the market, at least in times when the market is rising. And it also includes things like quantitative directional funds, which are algo-driven and tend to make momentum plays rather than long-term investing decisions.

And yet, look at the performance of all those funds put together, and it turns out to almost exactly mirror the performance of the S&P 500.

The result is predictable:


Once upon a time, according to this chart, hedge funds playing in the stock market actually outperformed the S&P 500. They did so during the dot-com bubble, and they did so after the bubble, too. You can quibble about what exactly is being measured here, and survivorship bias, and performance fees, and things like that. But whatever’s being measured, it’s now gone negative. Put all the smartest hedge-fund managers together in a room, tell them to go out and make lots of money, and over the past five years you would have been better off in an index fund.

As Gongloff says, “the current market environment of extreme correlation has made it nearly impossible to pick stocks well”. And this has lessons even for those of us who simply invest in index funds, too. A lot of financial advisers get quite excited about diversifying index funds — rather than just investing in the S&P 500, you should invest in a growth fund here and a value fund there and an emerging-markets fund and a large-cap fund and a small-cap fund and probably a RAFI fund or two just for good measure.

But the one thing you can be quite sure of, when it comes to all these clever ETFs and index funds, is that they have significantly higher fees than your bog-standard S&P 500 index tracker. And with correlations where they are, it’s increasingly difficult to believe that there’s much if any reason to pay those higher fees. Even if they’re a lot lower than the 2-and-20 charged by the HFRI crowd.

Update: It’s not a thing of beauty, but here are the correlation charts for other hedge-fund strategies. The thick line is hedge funds generally; the only strategy which doesn’t have high correlation right now is Macro.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

Why is anyone surprised at this correlation? Most stock-market trading volume today *is* hedge-fund trading volume.

HFT shops alone account for around 70% of volume ( e.g., see rcentage-us-equity-trades-are-high-frequ ency-trades ). Add in all the other equity hedge funds, and the figure is even higher.

In contrast, index funds and ETFs trade almost exclusively to rebalance or accomodate inflows/outflows, and traditional mutual funds are typically very-low-frequency traders.

To a close approximation, hedge funds *are* the market.

Posted by callesantiago | Report as abusive

For further reading, Kenneth French had a good summary on the costs on active investing:  /322/ECON322(2008)%20Kenneth%20R%20Fren ch.pdf

(I’m not sure what you think the correlation should be, actually; I’m more surprised by the low values at the beginning of the sample than the 0.9 in the end.)

Posted by Th.M | Report as abusive

Does this mean that the smart money is getting dumber?

And the dumb money socking money away in low-cost Target Date funds in 401ks is getting smarter?

Posted by ErnieD | Report as abusive

Trading is essentially zero-sum. There is a certain profit associated with any given market inefficiency, which necessarily must be divided up amongst all the people attempting to exploit that inefficiency.

Lim P/M as M->inf = 0.

Posted by TFF | Report as abusive

The equity markets are casinos, not places of investment. That they are a zero sum game would result in the correlation you cite. Minus, of course, the vig extracted by the house managers.

Posted by Beezer | Report as abusive

Note the spike in correlation in late 2008. This simply shows that the recent high levels are somewhat due to most firms taking losses during the financial crisis: the S&P and their own portfolios were both subject to large downward movements, thus raising the correlation. I wonder what this would look like on an (e.g.) two year horizon.

Posted by Mike.Abrahams | Report as abusive

I’m biased but one of the best (only) ways left to get true alpha is to invest in startups/private growth companies… (of course that doesn’t mean it is risk free, just that returns will have a very low correlation to public markets…)

Posted by parkparadigm | Report as abusive