ROBO-TRADING MENACES FINANCIAL MARKETS, EXPERT WARNS
- by Gino Vicci
It was as sudden as a terrorist attack. It came out of nowhere. For a moment in time, a shock wave sent panic through the most sophisticated financial markets in the world.
Virginia Mcgathey, owner of McGathey Commodities, remembers the May 6, 2010 "Flash Crash" very well.
The stock market had been relatively stagnant most of the day, she recalled, but in the afternoon, the Dow Jones industrial average suddenly plummeted 600 points in only a few minutes. That had never happened before. It is not supposed to happen, either.
"People were thinking… uh oh.. there is more fallout, not again," McGathey said, referring to concerns that the market was about to see a replay of the 2008 meltdown. But that earlier decline was slower, and people understood why it was happening.
With the Flash Crash, "There was so much confusion about what was happening. When you looked at the screen… it was moving so fast you couldn't even see what the numbers were," she explained. The trading screens were blurred in the same way numbers are blurred when gamblers pull the levers of a slot machine, said McGathey, who trades commodities at the Chicago Board of Trade.
And then, almost just as quickly, the market reversed course to recover from most of the damage in 20 minutes. Nobody knew exactly what happened, or why the blowoff had stopped.
The scary, unexplained mini-crash spurred a joint investigation by the Securities and Exchange Commission and the U.S. Commodity Futures Trading Commission, and the report those regulators eventually issued suggested that algorithmic trading was partly to blame for the volatility of the "Flash Crash."
Algo-what?
Algorithmic trading, sometimes referred to as "high-frequency trading," is basically a computer program designed to set trades, which can be based on pricing, timing, quantity of an order or a combination of the former. In essence, the computer is doing trades without human input.
The joint report also found the crash was accidental, caused by a flaw in the execution of a giant computer-driven trade order that made use of an algorithm. As the stock prices dropped, other computer programs kicked in, creating a snowball effect.
Because the markets are vulnerable to these accidental flubs, a growing number of people contend that intentional attacks, like terrorism, could be even more catastrophic. And such attacks are far from impossible, some experts say.
"Wouldn't it be really bad if terrorists took over a hedge fund or hacked into a stock exchange and manipulated markets?" asked Dr. John Bates, chief technology officer of Progress Software. "It could really hurt us badly."
With algorithmic trading, a computer program could be set to outbid another potential buyer or algorithm. If two or more algorithms were designed to do the same thing, the programs could outbid each other infinitely, causing a chain reaction that could lock up the markets.
"It's really easy to see how disastrous this could be," Bates added.
Bates, who also sits on the advisory panel of the Security and Exchange Commission, has predicted that an exchange or trading destination will be hacked by financial terrorists in order to manipulate the markets for political gain this year.
"That's an area that I think is a blind spot, I would love to know who is thinking about that," Dr. Bates said.
The CFTC and the SEC have been examining high frequency trading more closely, in order to explore ways to police it, but the regulators acknowledge they're struggling to keep up with technology.
In a testimony to Congress last month, Director of National Intelligence James Clapper admitted that combatting cyber-terrorism is a priority and that the U.S. is "currently facing a cyber environment where emerging technologies are developed and implemented faster than governments can keep pace."
And clearly, algorithmic trading is one of those potentially dangerous technologies.
During a question and answer session for reporters last week, SEC Chairman Mary L. Schapiro signaled that there is not much more that can be done at this point.
"… We don't have enough data yet to really be able to justify significant additional steps at this point," Schapiro said. "We need to have a much deeper understanding of the impact of high-frequency trading on our markets."
In the end, fighting fire with fire may be the most effective way for regulators to instantaneously monitor potential threats or attacks against the market.
"The technology exists… it's the technology high frequency traders are using," Dr. Bates explains. "If they can use it to spot trading opportunities, you can certainly use it to spot market abuse," he argued.
There are currently measures in place by federal agencies to rein in high frequency trading abuses, but none of of them monitor in real-time.
Because high frequency traders are often blamed for lightning fast market moves, most trading exchanges such as the IntercontinentalExchange have implemented their own measures to police high frequency trading.
Exchanges may assess fees to trading firms that do not fulfill a certain percentage of orders placed on the exchange, for one example.
While dangers do exist, others feel the technology is misunderstood.
"Unfortunately, high frequency trading is not well understood and, as a result, is often blamed, in absence of facts and analysis, for a multitude of market problems, whether real or perceived," said Chuck Vice, ICE president and COO.
Nevertheless, the debate over whether the advantages of computer-driven trading technology outweigh its dangers is likely to get more heated as regulators try to catch up to its speed.
