This post has been updated.
For about a quarter of an hour in 2010, the entire financial world was in chaos. The stock market was plummeting, traders were scrambling. It was as if an earthquake had happened in the middle of Wall Street
But the May 6, 2010 “flash crash,” where the Dow Jones Industrial Average plunged wildly by 600 points, and then shot back up in a matter of minutes, wasn’t the result of tectonic forces — allegedly just one guy sitting in his house in Britain while a computer program boldly manipulated the market for him.
That’s according to U.S. prosecutors, who filed criminal charges Tuesday against 36-year-old British trader Navinder Singh Sarao alleging that he “spoofed” the markets on more than 400 occasions, including May 6.
In an announcement about his arrest today, officials said that Sarao was “significantly responsible” for the 2010 flash crash, which began after Sarao entered and then quickly withdrew a series of large orders using an algorithm intended to make prices drop fast.
Sarao’s spoof wasn’t the sole cause of the crash, officials said, but he “helped create the imbalance and the circumstances by which the flash crash happened.”
Eric Hunsader, CEO of the data company Nanex, which monitors all market transactions, believes that the “flash crash” goes beyond just one guy with an algorithm. Sarao’s “spoofing” may have precipitated the crash, but he was able to do so because of the inhuman speed at which trading now happens.
“It’s hard to really grasp exactly how far it’s come into the land of the silly,” Hunsader said in a phone interview.
The old-fashioned vision of Wall Street — ticker tape, slamming telephones, Charlie Sheen — has been replaced by buildings full of humming servers. Via these machines, high-frequency traders operating on algorithms rather than human decision making can execute thousands of orders faster than a traditional trader can blink.
[This animation, via Nanex, shows how the rate of trading has increased since 2006]
That’s what happened on the day of the flash crash, Hunsader said. The market was already having a down day, thanks to concerns about financial problems in Europe. And high-frequency traders were already straining the capacity of Wall Street’s computer systems, which were due for an upgrade.
That afternoon, Sarao allegedly put in a bogus order on a futures contract pegged to the Standard & Poor’s 500-stock index, triggering a decrease in prices. Hunsader, monitoring transactions from his offices in Winnetka, Ill., saw the move happen — “It’s so large, it sticks out like a sore thumb,” he said — though at the time, he didn’t know what would happen to it.
It left the market vulnerable to other big movements: a large, apparently not-fraudulent sale a few minutes later.
Around 2:42 p.m., the falling prices on Sarao’s futures contract spread to other markets. Most high-frequency trading algorithms have a built-in mechanism that shuts them down when the market gets too turbulent, the computerized version of the Wall Street adage “when in doubt stay out.” These programs were wary of the unexplained decline in prices sparked by Sarao’s “spoof,” and were thus unwilling to buy the lower-priced shares.
But with hardly any traders in the market to generate demand, prices started to fall even further. Soon shares of valuable companies like Accenture were selling for pennies — an opportunity a human trader would have jumped at, if humans were running the stock market and not computers.
“Where a human would say this is ridiculous, with algorithms the goals are so short term they’re not really seeing a bigger picture,” Hunsader said. “And the information is traveling so fast that human participants are essentially locked out of the marketplace.”
About 10 minutes after the start of the decline a circuit breaker tripped, stopping trading on futures contracts. The pause lasted only five seconds, but it was enough for the market to reset itself. The Dow Jones soon regained the 600 points and things were back to normal — sort of.
“Flash crashes” like the one in 2010 have happened again. So have examples of “spoofing,” which Hunsader says he sees on a daily basis.
“These are all because of these algorithms that operate at speeds of no common sense,” he said. “For the market to function properly it should have a diversity of opinions and participants who could take advantage of these prices. … Placing and canceling an order so fast that somebody 20 miles away could not physically take advantage of that information — that’s just silly.”
Hunsader doesn’t think that high-frequency trading needs to be regulated more — he said existing rules are enough to deal with the issue, so long as officials enforce them. And the arrest of Sarao might put other market manipulators on notice.
Not everyone agrees. Michael Lewis, the nonfiction writer behind “Moneyball” and the closest high-frequency trading has to a celebrity critic, argued that regulators need to do a better job leveling the playing field for ordinary investors in his 2014 book “Flash Boys.”
“The stock market is rigged. It’s rigged for the benefit of really a handful of insiders. It’s rigged to sort of maximize the take of Wall Street, of banks, the exchanges and the high-frequency traders at the expense of ordinary investors,” he told NPR last year.
But proponents of high-frequency trading say that it creates more opportunities to execute trades, increasing the amount of cash available to businesses that need it. And a recent study from Columbia Law School found that the practice does not inherently create unfairness in the market.
Hunsader’s solution? “I simply think you need a grown up in the room who says ‘Well these are the rules, we’re not going to haphazardly apply them,'” he said.
Based on the criminal complaint against Sarao — he allegedly named his firm “Nav Sarao Milking Markets Ltd.” and told the Chicago Mercantile Exchange to “kiss my ass” — the call for a “grown up” seems apt.