3rd Birthday of Flash Crash: Is the Stock Market Safer?

Today is the three-year anniversary of the Flash Crash when traders and investors saw the high-speed stock market fall 700 or so points in 20 minutes and quickly rebound.

But the gut-wrenching move struck fear into retail investors, who saw blue chip stocks like Proctor & Gamble and Accenture, trade at a penny, as it exposed flaws in the electronic market structure of the U.S. stock market.

While we haven’t had another event as extreme as the Flash Crash, a string of technical snafus in the past two years have amplified the some of the structural imbalances in the wired market. Two weeks ago, a fake Tweet read by algorithms resulted in the Dow plummeting 145 points in seconds, and individuals stocks (even Google) routinely experience mini-flash crashes.

Many pundits are commenting today, debating whether some of the new structural mechanisms put in place by regulators and exchanges can prevent a repeat of the Flash Crash.

Some of the experts on CNBC point to the single stock market wide circuit breakers and the elimination of so-called stub quotes – bids and offers that were placeholders for market market’s quotes, that were triggered that day, as signs of progress that will support the markets if another prevent were to precipitate a flash crash.

Though we haven’t had another episode of this magnitude since May 6, 2010, market participants are concerned that some event will trigger another wild reaction and that a Flash Crash 2.0 is inevitable. Going back to the day, CNBC Reporter Bob Pisani cites the confusion about the orders and prices. “That caused people to shut down their trading systems and that was a loss of liquidity. That was a major exacerbating factor in the flash crash.” And there was rioting in Athens, something that people had not seen before, he notes. Computer-driven structural safeguards are in place now, like the limit up/limit down (LULD) rules. Under LULD, Pisani explains, if a stock trading at $10 falls 5 percent to 9:50, that stock would halt trading for five minutes.

Pisani said the circuit breakers/LULD are going to be more effective than people think.

Three years ago, there was no single circuit breaker in place on all the 13 exchanges, and now there is a uniform safety belt, he points out. The severity of the decline would be mitigated by the elimination of stub quotes and single stock circuit breakers.

But Art Cashin, head of NYSE floor operations for UBS Financial Services, is clearly not satisfied. Cashin tells CNBC that he doesn’t think that regulators have fully explored or arrived at the real cause of the Flash Crash. “They haven’t drilled down,” said Cashin. He is also skeptical of the safe guards.

“You always worry whether the life boats are real or are they painted on the side of the ship to make you feel better.”

While Pisani puts his faith in electronic mechanisms, Cashin points out that humans are the ones he trusts. “Nothing traded for a penny at the NYSE. Human intervenes and said that doesn’t look right. But at other places, where trading was entirely automated, computers were instructed to send their orders to certain venues for favorable treatment, he said.

Cashin emphasizes it was the humans that spotted the problem with Knight Capital last August when software flooded the market with unintentional orders. And human traders also began to worry about the Twitter hoax not being real, whereas automated trading algorithms reading the fake tweet reacted to it and sent the Dow down 145 points within seconds.

On CNBC, Pisani noted that exchanges are moving toward a kill switch, which will give them another mechanism to shut down renegade systems efficiently.

One of the concerns about whether a flash crash will happen again is that in a crisis, HFT firms can pull their liquidity during a catastrophic event. And the creation of a surveillance system for the SEC called the Continuous Audit Trail or CAT seems to be perennially delayed. In light of this inertia, the potential for another flash crash is out there, but there are mitigating factors, like Pisani said, and most institutional algorithms don't trade off of Twitter, at least not yet.