The rise of DIY quants could spell the end of easy profits from computer-driven trading
With cheap data and software, machine-based investing is open to anyone with the math chops “Democratization of tools doesn’t necessarily mean… democratization of good judgment”
Clay pottery decorates the halls of a thatched, single-story adobe home in the desert town of Las Cruces, N.M. Out back, where scrub brush stretches into the arid plain between the mountains and the Rio Grande, is a 50-foottall wireless Internet tower.
Roger Hunter is in the kitchen, grinding hand-roasted coffee beans. The 66-year-old former math professor turned investor apologizes if he appears a bit out of sorts. As chief technology officer of a two-man startup called QTS Capital Management— his partner lives in Canada—Hunter pulled an all-nighter fixing a systems glitch.
It’s a long way from Wall Street, but for Hunter and investors like him, being apart from the frantic activity of New York is an advantage. “This is particularly true when developing code and exploring new strategies,” he says. His strategies are all automated and might include trades on anything from currencies to hog futures to options on market volatility.
With little more than open source software and an Internet connection, Hunter is one of a new breed of traders breaking into quantitative investing. Quants, as they’re known, crunch a dizzying amount of data from across global markets and write programs to trade on the patterns they spot in asset prices. Powerhouse firms such as AQR
Capital Management and Citadel have used fast computers and closely guarded algorithms to try to beat the market, and analysts estimate the 40-year-old industry runs more than $1 trillion in assets. Now low-cost, highpowered technology is razing virtually every barrier to entry.
The rise of DIY quants comes as the proliferation of machine-based strategies has made it harder for traditional players to succeed. In January, Martin Taylor of Nevsky Capital closed his 15-year-old hedge fund, lamenting the distorting influence of computer traders. But established quants, too, are feeling the heat. “Technological edge is harder to come by because the more egalitarian these tools have become, the more difficult it is to come up with something truly new,” says Andrew Lo, a finance professor at MIT and
chairman of AlphaSimplex Group, a quant research firm.
In other words, the growth in traders using quant strategies also tends to diminish easy profits, says David McLean, a finance professor at DePaul University. He cites research showing that three years after an academic paper on an automated strategy is published, returns based on that strategy fall by more than half as more traders catch on. Meanwhile, as more quant traders vie for an edge, some envision a world where so many algorithms are unleashed on electronic markets that sudden shocks—such as August’s meltdown in U.S. stocks—become more frequent.
While it’s difficult to know precisely how many quant startups there are,
Quantopian, a Boston-based firm that provides coders the tools and software they need to build quantitative trading programs, has seen its user base climb to 66,000 from 1,570 in 2011, the year it started.
Open source coding languages such as R and Python, which are building blocks for critical number crunching, are posted for free on online libraries. Boutique services such as Estimize provide crowdsourced earnings estimates. “There’s so much data, so much open sourced software and computing power available,” says Emanuel Derman, director of Columbia University’s financial engineering program and the former head of quantitative risk strategy at Goldman Sachs.
Back in Manhattan, the elite hedge funds still rule, having built up the status and the reputation that come with years of outsize returns. Point72
Asset Management’s Midtown-based quant business, with its large glass conference rooms and white walls adorned with founder Steven Cohen’s personal art collection, looks and feels nothing like a startup.
And according to Ross Garon, the head of Point72’s quant shop, big firms have little to fear from smaller competitors. They still have the best technology and brightest minds (not to mention the most money). “The democratization of tools doesn’t necessarily mean there’s the democratization of good judgment of what to research,” he says.
Despite those disadvantages, Hunter and his partner, Ernie Chan, who works from Niagara-on-the-Lake in southern Ontario, have held their own. QTS returned 12 percent last year, outstripping the U.S. stock market and the average for hedge funds globally. It runs $22 million for individuals and one large family fund.
To keep costs low, QTS uses a service called AlgoSeek to get access to price data for futures, pulling in an “astronomical” amount of information for $500 a month. Hunter himself wrote the code that QTS’s options trade on. The firm employs part-time contractors and uses tools such as Amazon Web Services to augment its computing capacity when a laptop won’t do the trick.
Billion-dollar-plus firms might not worry about two guys running a few million dollars. But “there’s a threat they’re missing,” says Dan Dunn, who oversees product management and membership at Quantopian. “The reality is there are brilliant people all over the world who they have never seen or heard of until they show up and eat their lunch,” Dunn says.
Then again, all those brilliant minds sometimes trip over one another. JPMorgan Chase’s Marko Kolanovic pointed to the role of crowded quant trades in the events of August, when U.S. stocks plummeted 11 percent in six days. Many blamed China and the Federal Reserve. Kolanovic told clients automatic selling by “price-insensitive” quants made everything worse.
Where Kolanovic sees danger, Hunter senses opportunity. On any given day, he tests 10 different models while executing eight strategies for clients. He and Chan are considering developing code to profit from distortions that show up in managed futures when too many quants trade the same strategies. “We’ve thought about trying to take advantage of it, certainly if the algorithm is clearly affecting the market,” Hunter says.
The bottom line Computer-driven trading strategies are easier than ever to execute, but that may chip away at their profitability.