Hedge funds, alpha, and beta

Hedge fund managers and investors in hedge funds are generally very smart and

very sophisticated. They like to talk a lot about risk-adjusted returns, and

especially about "alpha" and "beta" – central components

of the Capital Asset Pricing Model. And yet, after all that highfalutin’ talk,

the way that hedge fund managers are paid is very unsophisticated indeed. There’s

no risk adjustment there: just a percentage of the funds under management, and

a percentage of the raw profits.

It’s easy, if you’re a hedge fund, to bump up returns a lot using simple leverage.

Let’s say I think the S&P is going up, and I want to invest $1,000 so that

it returns more than the S&P. All I have to do is borrow $1,000 from the

bank, add it to the $1,000 that I have in cash, and invest $2,000 in an S&P

500 index fund. At the end of the year, let’s say the index has gone up 10%,

so my $2,000 has become $2,200. I pay off my bank loan – let’s say I borrowed

at 4%, so that’s $1,040 – and have $1,160 left: a 16% return on my original

$1,000.

Now let’s say that rather than going through that operation myself, I hired

a hedge fund manager to do it for me. I’d pay him 2% of the $1,000 – that’s

$20 – and another 20% of the $160 profits – that’s $32. So the total

fees would be $52, leaving me with $1,108. If I’d invested in the S&P 500

directly, I’d only have $1,100, so even after fees, I can tell all my friends

that my hedge fund manager has beat the market.

Of course, there’s no reason why my hedge fund manager should stop at borrowing

just one dollar for every dollar he invests. Let’s say he takes my $1,000, borrows

$4,000, and invests $5,000 in index funds. At the end of the year, he’s made

$500 in profits, and has paid $160 in interest, leaving him with a $340 or 34%

return. His investors pay fees of $20 plus $68 for every $1,000 they invest,

leaving them with total profits of $340 – $88 = $252, or more than 25%.

Whee!

In the real world, hedge funds, and their investors, are much more sophisticated

than that. They know the difference between smart investing and leveraged investing,

and they measure that difference using their beloved alpha and beta. Alpha is

an attempt to measure how smart an investor is: to work out what his returns

are after taking into account the riskiness of his investment portfolio. Beta

is simply the riskiness of the portfolio.

In my examples, the S&P 500 has a beta of 1, or 100%; the first hedge fund

manager has a beta of 2, or 200%; and the second hedge fund manager has a beta

of 5, or 500%. The riskier and more volatile the assets you’re investing in,

the higher their beta. And, of course, as you increase your leverage, you increase

your beta proportionately. What hedge fund managers are looking for is alpha,

which is return over and above whatever the beta of their portfolio

might be. So if the beta of your portfolio is one, and you manage to do better

than the S&P 500, then you’re a smart investor with positive alpha. In my

cases above, the hedge fund managers might have managed to beat the S&P

500, but they did so by increasing their beta and decreasing their alpha. In

fact, their alpha in both cases is negative.

Hedge funds, then, are always on the lookout for ways to increase their alpha.

One way to do that is to get high returns; another way to do that is to decrease

your beta. If one manager manages to get a 16% return with a beta of 110%, and

other manager gets that 16% return with a beta of 85%, then the second manager

has a much higher alpha, or risk-adjusted return, and is likely to be much more

popular with investors.

At the end of the year, however, the fees that those two hedge fund managers

charge to their investors will be identical. Hedge fund managers might spend

their lives chasing alpha, but that’s not what their pay is tied to. Their pay

is simply tied to total return.

This, I think, is the point that Jenny Anderson was trying to make in her

column on Friday. Jenny is an excellent reporter, as I said

on Tuesday. She covers the hedge fund beat for the New York Times, she has

great sources, and she’s generally dead-on in knowing who’s bullshitting her

and who isn’t. The broad thrust of her column is quite right. But the substance

of it is way off base: in order to make a perfectly valid point, she uses all

manner of erroneous arguments.

Firstly, she doesn’t seem to really understand alpha and beta: a problem, when

the headline on the column is "Hey, You Have a Problem Paying Alpha Fees

and Getting Beta Returns?". Here’s the meat of her argument:

Where are the "alpha" generators, those that are supposed to create

performance above the market? (Market returns are called beta.) From May 1

through May 19, the Standard & Poor’s 500-stock index dropped 2.9 percent.

… The GLG European Long-Short fund fell 5.13 percent through May 19, posting

a 13.70 percent gain for the year. …

To be fair, one month of data is not an indication of how closely a fund is

tied to a particular market or how that fund will perform for the year. Any

number of things can happen to turn around the performance. But the question

that many investors should be asking is: Am I getting alpha or beta? And if

the answer is beta, why the high fees?

Simply put, hedge funds that claim they are not correlated to the markets

should not be correlated to the market. In bull markets, investors do not

question correlation.

There is a lot of very confused logic here. Firstly, alpha is not the same

as "performance above the market," as we’ve seen: you can have performance

well above the market, but if your beta is high enough then your alpha can still

be negative. And beta is not a measure of "market returns": it’s a

measure of risk, or volatility. Anderson seems to be assuming here that all

hedge funds have a beta of one, before rephrasing her assumption, mildly garbling

it in the process, and then giving it to us as a definition of beta.

Anderson then spends a lot of space (I snipped out most of this) running down

a list of hedge funds and looking at how they performed over the period May

1 to May 19. What on earth is this supposed to prove? For starters, we have

no idea of these hedge funds are representative of hedge funds as a whole, or

if Anderson picked them precisely because they fell significantly in

that short time. But in any case, what point, exactly, is she trying to make

when she says that the GLG European Long-Short fund is down 5.13% in May so

far but up 13.7% over 2006 as a whole? And why is she comparing a European fund

to the performance of the S&P 500?

Looking at various hedge funds’ performance over a period of three weeks will

tell you absolutely nothing: most smart investors won’t even look at a fund

manager until he can demonstrate his performance over three years at

the minimum. Hedge funds are difficult to get in to and difficult to get out

of: the people who invest in them do so for the medium to long term. Looking

at short-term performance is something that market analysts do in order to get

a feel for where hedge fund money might be positioned. It is a very silly way

of judging hedge fund managers. Anderson admits as much with her "to be

fair" sentence, but if she really believed it, she wouldn’t have spent

the previous two paragraphs detailing a series of short-term returns.

Anderson is quite right that investors should be asking whether they’re getting

alpha or beta: whether their funds’ returns are positive, on a risk-adjusted

basis, or whether they’re simply following the market as a whole. But this isn’t

news to any investor, or to any hedge fund. While fees might be tied to total

return rather than alpha, investors tend to pull their money out of any fund

whose beta is higher than the return that they’re getting. When hedge funds

report their performance to their investors, they tend to emphasize risk-adjusted

returns over nominal returns for precisely that reason.

And none of this has anything to do with correlation. Yes, Anderson is trivially

correct that "hedge funds that claim they are not correlated to the markets

should not be correlated to the market". But neither alpha nor beta is

related to market correlation. In fact, Anderson implies that "hedge funds

that claim they are not correlated to the markets" are exactly the same

as "the ‘alpha’ generators, those that are supposed to create performance

above the market". But of course that’s impossible: if you’re not correlated

to the market, you can’t promise above-market returns.

I once wrote a profile of a large, conservative hedge fund called Elliott Associates,

which is very much one of those funds which claims that it is not correlated

to the market. The idea is to get consistent good-but-not-spectacular returns,

no matter whether the stock market is going up or down or sideways, and to do

so with a lower risk than the S&P 500 to boot. Elliott definitely promises

alpha, but it never promises to beat the market. If Elliott returns 20% with

low beta in a year when the S&P 500 returns 30% with high beta, then Elliott’s

managers will be happy. Over the long term, Elliott has outperformed the S&P

500, but not by an enormous amount: the really impressive numbers emerge when

you look at its risk-adjusted returns, since it managed to outperform the market

while taking on substantially less risk than someone who simply invested in

the market as a whole.

Anderson, on the other hand, concentrates in her column on "emerging-market,

commodity, small- and mid-cap and currency funds": hedge funds which are

specifically designed to give investors exposure to emerging markets,

commodities, etc. If I invested in Elliott Associates and it underperformed

the S&P 500 in a good year for US equities, I wouldn’t be upset, since no

one invests in Elliott Associates in order to get exposure to US equities. On

the other hand, if I invested in an emerging-market fund and it did badly in

a year which was good for emerging markets, I would be upset: investors

in such funds want correlation with those asset classes.

When hedge fund managers or investors finish reading Anderson’s column, then,

they’re likely to have less respect for Anderson personally, and for the New

York Times business section generally, than they did at the beginning. And that’s

bad for both franchises. The Times wants desperately to be taken seriously in

the financial world, and "added-value" columns like Anderson’s are

a large part of that effort. But it’s worth noting that the Wall Street Journal,

the gold standard of daily financial journalism, has precious little in the

way of journalist-written commentary. Opinion is cheap; news is expensive. If

the Times continues to emphasize the former in its own pages while outsourcing

the latter on its website, it will never join the likes of the Journal or Bloomberg

News as a respected source of financial information.