As a stock-futures trader, George Pruitt used to spend the day hovering over his computer screen.
Now, he just stays within earshot. When his computer identifies an attractive time to buy or sell, it emits a horn blast. It does this often, including once last week while Pruitt was on the phone discussing his approach to automated trading.
Future Stock
“You hear that going off in the background? That’s saying the price of the e-mini just went up to a certain level,” said Pruitt, director of research at Futures Truth, a trading company and magazine publisher in Henderson, North Carolina. The e-mini, a futures contract that tracks the S&P 500 Index, is one of several securities the company monitors almost exclusively online.
While successful trading strategies still usually involve some subjective human analysis, traders are entrusting a growing share of the work to their PCs. Most of the time, individual investors authorize each trade before it goes through. In some cases, however, even solo investors are cobbling together systems that are 100 percent automated.
Call it the age of the machine trader. Computers have some major advantages over people in the trading game: They don’t sleep. They don’t ask for pay. And they aren’t subject to the equilibrium-killing forces of emotion, which have been known to provoke even the most disciplined human trader to abandon logic from time to time.
“They’re my partners,” said Greg Ballard, a trader based in Fort Smith, Arkansas, of his PC collection. “I make them work 24 hours a day, don’t give them vacation or anything.” They’re programmed to execute trades based on a periodically changing set of algorithms that track stock-index futures.
Ballard is among a small but growing cadre of investors who develop and program their own mathematical formulas to identify trades. Each new program, Ballard said, is subjected to months of simulated trading before it’s used to manage actual money. Ballard won’t share how the algorithms work or how much income they’ve generated over his four-year trading history, except to say they’ve been profitable.
More commonly, individual investors who use automated trading track established indicators for stocks and indexes. Ralph Cruz, co-CEO of TradeStation (TRAD), a trading platform for active traders, says the following metrics are particularly popular:
The moving average: Traders look at the mean of a stock’s closing price over a specified time frame. The particularly prevalent 10-day moving average is the mean of a stock’s closing price over the past 10 days.
Relative strength index: RSI compares the magnitude of recent gains to recent losses in an attempt to determine if a security is oversold or overbought.
Moving average convergence divergence, or MACD: This popular, fairly complex indicator compares moving averages to determine buy and sell opportunities.
All such methodologies fall under the catch-all title of technical analysis, the subject of countless books, newsletters and “how-I-got-rich” infomercial testimonials. The central idea — reducing securities markets to mathematically crunchable datasets that can be objectively analyzed — is as old as trading itself.
What’s newer is the ease and sophistication with which individuals can put such methods to work on their own behalf.
Nelson Freeburg, editor of the algorithmic trading newsletter Formula Research, estimates that millions of people worldwide use computerized systematic trading techniques today. No cumbersome barrier keeps newcomers from entering the field. A trader could probably get started, he said, with $10,000 to $15,000 dollars.
Freeburg advises novice traders to test their skills using one of a handful of sites and trading platforms that offer trading simulations, such as Fidelity’s Wealth-Lab or TradeStation. He also warns traders to be selective about the indicators they follow. When a market-timing signal becomes too widely known and implemented, it may cease to be effective.
On the New York Stock Exchange, trading is more mechanized than ever. Program trading, which NYSE defines as transactions involving the purchase or sale of baskets of 15 or more stocks, accounts for more than half of all shares swapping hands on the exchange. Such trades accounted for less than a quarter of trading volumes five years ago.
Program trading isn’t exactly robotic trading. Much of it is initiated by humans, who use automated trading systems to separate buy and sell orders into smaller batches. Even individuals who use automated trading platforms, Freeburg believes, are ill-advised to allow their computers to actually execute trades.
“I would have a little human element just to check on things before surrendering everything to an electronic algorithm,” he said, noting that computers, like the people who make them, still make errors.
Ballard, however, says he’s willing to entrust the bulk of trading responsibilities to his PCs. In the interest of reducing risk, he closes his positions at the end of every trading day, even if it entails taking a loss.
The effect of automated trading on markets is that they move faster than ever. That’s why inexperienced traders, according to Ballard, probably ought not to risk substantial sums of money before testing trading strategies for a few months.
First-timers should also be prepared to do poorly, said Cruz, who rapidly lost all his money when he and his brother began trading in the 1980s. Since then, Cruz said, he has learned to take a more disciplined approach, basing decisions on numerical analysis rather than subjective judgments.
“Typically, emotions are your worst enemy,” he said. “You don’t want to accept the fact that you made a mistake, so you let that mistake grow. One of the benefits of having a more rule-based approach is you take emotions out of your decisions.”
I'm a former general partner at google ventures, head of Google for Entrepreneurs in Europe and founding head of Campus London, Google's first physical hub for startups.
I'm also the founder of Techbikers, a non-profit bringing together the startup ecosystem on cycling challenges in support of Room to Read. Since inception in 2012 we've built 11 schools and 50 libraries in the developing world.
