Saturday, September 7, 2013

Quants-R-Us? Algorithmic Trading Trickles Down To Individual Investors

Six flights up in a dim, grungy office building in Manhattan's Union Square neighborhood, Christopher Ivey is working on what he thinks is the next step in the evolution of trading. His efforts also serve as a subtle reminder that the current bull market seems to be a bit frothy.

Armed with $4.5 million in funding, the 2011 Harvard grad recently launched a Web-based platform called Rizm, designed to let individual investors with no coding skills build computer programs that select and trade stocks automatically, similar to the trading programs used by quant funds and high-frequency trading firms.

The pitch: For $99 per month investors get quick cloud access to sophisticated algorithm-building tools and the capability to back-test strategies. You can easily generate rapid-fire executable trades, sans emotion, and place them with an e-broker. Suddenly the notion of blasting out math-driven trades like über-successful quant hedge funds, such as James Simons' Renaissance Technologies, are a few clicks away.

"If you can only follow five trades a day, because that's the mental bandwidth for a human trader, now run it against 500 stocks," says Ivey. "Or just let it run on your five, and go spend time with your daughter and hang out, because it's running for you."

If that sounds like trouble–amateurs turning their financial ovens to high and then leaving the house unattended–you'd better get used to it. This is the natural evolution of the self-directed trading boom that started with Charles Schwab and the deregulation of commissions in 1975. Cheap technology and sophisticated engineers have disrupted almost every other aspect of Wall Street. Now we get hedge fund tools for the masses. Besides Ivey's startup, EquaMetrics, there are at least four other companies offering programming-free algorithm builders that sport slick names like CoolTrade and Prodigio.

Sociologist Alexandru Preda, who studies traders, says he's seen an uptick in individuals building trading programs over the past four years. Some are young math and science graduates who may have learned some algorithmic techniques in campus trading clubs sponsored by investment banks. Others are veterans of quant funds and investment banks, retired or laid off, who are using the tricks they learned to manage their own money.

Individual investors are clearly gaining more confidence in the post-financial-crisis bull market. The number of self-directed brokerage accounts in the U.S. rose 5% last year to 40 million, according to the market research firm Celent. More tellingly, the ranks of active traders (10 or more transactions a month) jumped 14% over the same period, now encompassing 2.4 million accounts.

"It doesn't take a lot to get started nowadays" building trading programs, says Ernie Chan, who runs a systematic trading consultancy. "The price of platforms for retail traders is dropping, and more brokers are catering to them."

Exhibit A: Tim Richardson, a former pro football player and hospital executive from Raleigh, N.C., who is one of  1,250 Rizm account-holders. The 49-year-old now trades for a living and says that the DIY algo platform has allowed him to ramp up to almost 30 trades a day, from fewer than 10. His strategy revolves around tracking technical indicators and jumping in when stocks or ETFs appear oversold. "Before, I could only comfortably look at a couple of stocks. Now I can follow 10." Richardson says automating his strategy has also helped him curb competitive instincts that could keep him in positions too long. "When I first started trading everything for me had to be a win. I would ride it to the bottom."

Some websites offer finished products; investors can browse established trading programs and subscribe to them or download a robot, as they're sometimes called, to run on their own computer. Needless to say, though, the pros aren't sharing their alpha-generating secrets, and the independent newbies often follow suit. "No one wants to talk about their algorithms, and they don't want to talk about the rig they built to operate it," says John Fawcett, cofounder of a browser-based development platform called Quantopian.

Fawcett aims for Quantopian to become a hub for math and science grads who have programming skills and are looking to get into the game, and it provides standardized tools to do so. With $2 million in funding from backers such as the high-frequency trading firm Getco, his website offers a free back-tester, sample programs to clone, and forums and supporting documentation. Live trading is in beta.

Ivey's Rizm platform differs from Quantopian in that it offers algorithmic trading to people who can't code and don't want to learn how. In Rizm, users drag and drop modules to build strategies–throw in a Bollinger Band indicator, set the terms (buy at the lower band, sell at the upper), toss in a stop-loss module, enter the stocks to track and voila.

"It's infinitely scalable," says Ivey. "You can nest as many strategies as you want, and it will just keep getting bigger and bigger."

Rizm has its detractors. Programming snobs scoff at the inherent limitations of drag-and-drop algorithms. Big retail brokerages firms like Fidelity and TD Ameritrade are also skeptical about the market opportunity in general. Most have offered auto-trading tools, of varying levels of sophistication, going back as far as 2006. Nicole Sherrod, managing director of the trader group at TD Ameritrade, says she sees a higher payoff in focusing on options and futures trading, and improving the brokerage's mobile platform. Most investors aren't interested in taking on the complexity of formulating rules-based strategies, she says.

Still the new wave of amateur quants means more trades, more commissions and more market volatility. Will these geeky hyper-traders beat the buy-and-holders? David Leinweber, who ran a $6 billion quant fund at First Quadrant and now heads the Lawrence Berkeley National Laboratory Computational Research Division's Center for Innovative Financial Technology, doesn't think so. "Algorithmic trading frees you from the drudgery, but do you have good ideas? There aren't that many masterpieces out there." A lesson the day traders of the 1990s learned the hard way.

***

Quant Farms

Wall Street banks and hedge funds are constantly looking for young talent to improve their trading systems. These are the five U.S. universities that are the richest hunting grounds, based on a survey of quant recruiters Lou Ricci of the Hagen-Ricci Group and Todd Fahey of Sheffield Haworth; David Leinweber, head of the Lawrence Berkeley National Laboratory Center for Innovative Financial Technology; and Margaret Belden of AQR Capital Management.

Carnegie Mellon University

Master's in computational finance, 2 years.

Tuition: $57,450 per year

Columbia University

Master's in financial engineering, 1.5 years.

Tuition: $57,000

Massachusetts Institute of Technology, Sloan School of Business

Master's in finance, 1 year.

Tuition: $74,400

Computer science and engineering programs are also fertile recruiting grounds.

University of Chicago

Master's in financial mathematics, 1 year.

Tuition: $51,000

Undergrad CCIB program and physics program are highly regarded as well.

University of Pennsylvania

Penn-Wharton Management & Technology BAS/BSE, 4 years.

Tuition: $45,900

 

No comments:

Post a Comment