Saturday, August 18, 2007

Even Quants Wobble When Markets Quake

Talking Business
The New York Times
August 18, 2007

“We have to atone to our clients, but we have no right to whine for ourselves,” said Clifford Asness, the co-founder of AQR Capital Management, a money management firm that has been much in the news recently. “When we succeed, we make a boatload of money, we get imitators, and our risk increases. That’s how capitalism works.”

We were speaking on Thursday, a week after one of the lousiest market days of his life. Along with James Simons of Renaissance Technologies and David Shaw of D. E. Shaw, Mr. Asness is one of the leading practitioners of what is called quantitative investing, using computer models to buy and sell thousands of stocks (and bonds and derivatives and commodities and currencies and country indexes and just about anything else that can be traded). Mr. Simons, Mr. Shaw and Mr. Asness, in particular, use these quant models to run what are called in the business “market neutral” hedge funds, meaning that their gains (or losses) are not dependent on whether the market goes up or down.

AQR has about $37 billion under management, with $27 billion of that in plain vanilla equity funds. The rest is invested in its quant hedge funds, its best-known operations. Indeed, over the last seven years, AQR’s flagship hedge fund has been up, on average, 13.7 percent a year, after fees, handily outperforming the Standard &Poor’s 500, which gained only 1.9 percent annualized during that time.

Mr. Asness himself is known for being ridiculously smart, highly engaging, and funny, with more one-liners than Henny Youngman. I got to know him, and to like him, a few years ago, when I wrote about him for The New York Times Magazine. He can get a little full of himself, but most of the time he can be brought back to earth with a small, friendly jab. In other words, he’s the rare hedge fund manager you’d like to have a beer with.

Anyway, back to that awful Thursday. As you may recall, the market ended Aug. 9 down more than 380 points. That kind of day isn’t fun for anybody, but it was an especially brutal day for firms that are supposed to be indifferent to market ups and downs — namely, quant funds like those run by Mr. Asness and his partners. What made it especially painful is that their troubles on Thursday really had nothing to do with the market’s fall.

In the days leading up to Thursday, Mr. Asness’s fund — and most other quant funds — had gotten clobbered. When the AQR flagship fund opened for business on Friday, Aug. 10, it was down 13 percent for August. Mr. Simons’s famed Medallion fund, which has rarely had a down month during nearly two decades of incredible performance, lost 8.7 percent in early August. By mid-August, Goldman Sachs’s flagship Global Alpha fund was down 26 percent for the year. Everywhere you looked in the little town of Quantsville, it was ugly.

And then, in the blink of an eye, it turned around, at least for the moment. As of today, Mr. Asness’s fund had gained back half of what it lost in the previous two weeks, and was at break-even for the year. I hear through the grapevine that Mr. Simons has already made back every penny Medallion lost in early August. During its conference call earlier in the week, Goldman announced that it had rounded up $3 billion for one of its battered hedge funds; I’ll bet a steak dinner that that fund has seen some gains this week as well.

All of which poses some big questions: What really happened during the Great Quant Meltdown of early August? More to the point, should it scare us or reassure us?

Let’s be honest here. You hear the words “quant fund meltdown,” and one firm comes to mind: Long-Term Capital Management.

Back in 1998, that now infamous quant fund really did melt down, not only liquidating, but shaking the entire global financial system. Long-Term used complex computer models that failed to anticipate some severe once-in-a-lifetime market events, and it was shockingly leveraged — it was using $100 of borrowed money for every dollar of its own capital — which magnified its losses. It was also run by some of the smartest people on Wall Street. “When Geniuses Fail” was the apt title to Roger Lowenstein’s fine book about that fiasco.

Ever since, whenever quant funds stumble, it’s “When Geniuses Fail Redux.” Wall Street wags begin to wonder if those losses will lead to something truly cataclysmic, while newspaper reporters take a certain undisguised glee in reporting on really smart people losing money. Even now, there’s enough Luddite schadenfreude in the air that rumors continue to circulate that AQR is continuing to absorb substantial losses — which is the exact opposite of the truth, Mr. Asness says.

What is scary in this case is not that the quant funds were the initial source of a ripple effect on the rest of the market; they weren’t. The quant funds were the recipients of a ripple that began in a corner of the market that they had little to do with —namely, the subprime mortgage crisis. It’s the way the subprime contagion shook the quants, whose subsequent downturn then added to the ripple effect, that’s what is nervous-making.

Mr. Asness’s hedge fund offers a case in point. Does his fund deal with the subprime business? Not in any significant way. Rather, the securities that cost AQR so much money were good old-fashioned equities.

To oversimplify (sorry: you can’t explain this stuff without oversimplifying), AQR’s market neutral funds use computers to sort through a set of complex but common-sensical criteria to identify all sorts of assets — including stocks — that it believes are undervalued but gaining some momentum, which means that both price and fundamentals are improving. It buys, literally, thousands of those stocks. Then it seeks out stocks it believes are overvalued and starting to lose momentum. It shorts those stocks. What makes the fund “market neutral” is that it always tries to have the same amount long as short. Mr. Asness likes to say that it’s not really rocket science but intuitive investing; the computers mainly allow him to do it across thousands of stocks at the same time.

Mr. Asness does not suggest that he is going to be on the winning side of every trade. Not even close. Nor does Mr. Asness suggest that his strategy is risk-free. It’s not. “If you don’t take any risk, you won’t make any money,” he said. Even when things are going swimmingly, he’s going to have almost as many losing trades as winning ones. But over time the winning trades will add to better-than-average gains. In a down market, he hopes that his shorts will fall more than his longs, and in an up market, he wants the longs to rise more than the shorts.

As for risk, he adds leverage to bolster returns; indeed using borrowed money to calibrate risk is a major part of his strategy. But it’s not crazy stuff like Long-Term Capital Management, and it would be hard to argue with his results over time.

What happened in August is something that happens to every investor at times, even Warren E. Buffett: his strategy stopped working. So did Mr. Simons’s strategy and that of all the other quants. Mr. Asness’s trades weren’t just a little off — they were hugely off. The undervalued stocks he was buying were dropping steeply, but he wasn’t getting any help from the short side of his portfolio. Several “quants” I spoke to — market veterans who had been through the 1987 market crash and the 1998 Long-Term Capital disaster — told me they had never seen anything quite like it.

Why did it happen? In the immortal words of the market sage, James Grant, “On Wall Street, every good idea is driven into the ground like a tomato stake.” Quant investing, as practiced by the likes of Mr. Asness, Mr. Simons and others, has been enormously successful. And anything that’s successful on Wall Street is invariably going to be copied by others. That is exactly what’s happened in many cases at firms that did other things besides quant investing — like trading in derivatives built around subprime loans.

As these subprime instruments have cratered, investors have lost faith not just in them but in other credit derivatives. The holders of these securities had to meet margin calls and make other payments. So they had to start selling more liquid securities like, well, the kind of easily traded securities held in their quant equity portfolios, like Microsoft or I.B.M. or General Electric. And as they sold, other quant shops, like AQR, which held many of the same stocks, saw huge drops instead of small gains. Is it any wonder traders are calling this a contagion?

One line making the rounds on Wall Street is that the events of last week show that, just as with Long-Term Capital Management, the quants’ models didn’t work — that bloodless computers simply can’t anticipate events outside the norm. That line drives Mr. Asness bonkers. “In theory, what just happened is impossible, so if we stuck to the theory, we’d be dead,” he said. “We know this stuff happens.” Once they realized the magnitude, he and his partners quickly began a mild “deleveraging” to protect against even bigger losses. Eventually, AQR started buying cheap stock again — which had become even cheaper thanks to the short-term panic.

In the view of several big-time quants I spoke to, their big mistake was in not realizing that their little corner of Wall Street had become so crowded with imitators — and that when others were forced to sell, they were going to get hurt. Now they are all trying to figure out how to factor that into their thinking for the future — Mr. Asness very much included. “We have a new risk factor in our world,” he said.

So how should the rest of us feel about what just happened? Even though the worst seems to be over, I still think we should still be worried. But not because computer-driven quant funds took a tumble. That’s a symptom, not a cause. The larger issue is the contagion itself — the fact that something so out of left field, like subprime, could wind up hurting the quants.

Richard Bookstaber, a former quant manager, has recently written a book, called “A Demon of Our Own Design” (Wiley, 2007), which has become a small sensation on Wall Street. In it, he argues that the proliferation of complex financial products like derivatives, combined with use of leverage to bolster returns, will inevitably mean that there will be a regular stream of market contagions like the one we’re having now — one of which, someday, could be calamitous. To him, last week’s quant crisis is a classic case in point. “I think crises become inevitable when you have a financial structure like ours,” he said. “How deep or how frequent they are, I wouldn’t want to predict.” Well, who would?

So yes, it really is a scary world out there. But quants like Mr. Asness aren’t the reason.


Anonymous Anonymous said...

Couple of comments!
1. If as Asness says, "we get imitators", then it must mean that what he/they do is not unique and can be done by many others. So simple common sense dictates that he/they should not get those big fees or deserve the 'boatload of money'. There are other people in society who do some real unique thing (which only a select few can do) and get paid much much less.
2. The problem with this hedge fund industry is the asymmetry of the payoff. Over the long term (5-10 years), it is very well established that 99% of these 'great' funds underperform the indices. But the managers take the 20% cut in the years they 'accidentally' beat the market but don’t return it back when they 'accidentally' underperform. All the high water marks etc in the contract are easily avoided by just closing the old fund and starting new.
3. Obviously anyone who gets the chance to manage money on these asymmetric terms would be stupid not to take it. The real question is why the investors (pension, endowments, etc) fall for this - especially given that their mandate and objectives are over decades and a few years up and down don’t make a difference on their long term returns.
4. The whole point of several of these funds which they tout incessantly is the 'market-neutrality'. Then how come, when it really matters (i.e. when the broader indices drop), they follow suit and drop even more!
5. The last point is about the returns of even the 1% who 'outperform'. This is not strictly true if you consider the risk they take on (higher leverage). It is possible to get these same returns by holding an index and levering up to the same levels as these funds do. Would be much cheaper, less riskier (volatile) and even may perform better. This is more or less what Mr. Swenson of Yale says.
In fact it is surprising that one of the most glorified of these (Goldman's hedge fund) returned just 40% in 2005 given the leverage they take on. This is amply shown by the 30% drop in 2 weeks which very clearly shows the leverage they are employing. Give this multiple, how come when they are 'lucky' as in 2005, they didn’t outperform by much more.
6. A handful of ‘normal’ fund (especially value oriented) managers have done equally well over the long term with much simpler and more fundamental bottoms up research and hence rational investing philosophies. Some of these quant funds may have done a few percentage points better but that’s a dubious achievement given all the math, computational and intellectual prowess/superiority they claim and delude themselves and others into. It is like a student who gets 70% with simple text book study and a few hours work vs. another who gets 75 with 10 times the effort and time and attending all kinds of coaching classes and creating general disturbance and noise and chaos all over the house and neighborhood.

11:33 AM  
Anonymous Anonymous said...

1. There are some big factual mistakes in the article. From Yahoo Finance info on S&P, I get an annualized return for S&P 5 years as 9.25%

Date S&P Ann Ret
8/16/2002 928.77
8/16/2003 999.74 7.64
8/16/2004 1,095.17 8.59
8/16/2005 1,219.02 9.49
8/16/2006 1,302.30 8.82
8/16/2007 1,445.94 9.26

2. The 5 year Asness return of 13.7 is not that amazing given that several mutual fund managers have performed close to same/better (Nygren, Weitz, Hawkins, DodgeCox Funds etc not to mentione several Motley fool portfolios) and not to mention the best of all - Buffet.

12:04 PM  

Post a Comment

Links to this post:

Create a Link

<< Home


Web Site Hit Counters
High Speed Internet Services