Recently, I visited the New York Stock Exchange (NYSE) to hear how their new hybrid model that combines human intuition with computer speed is better than all the other 12 exchanges in the marketplace. NYSE is the ONLY human trading floor left in existence, and it is more of a backdrop for financial news networks than anything else. Robots have taken over the financial system, and Michael Lewis is not happy.
In “Flash Boys: A Wall Street Revolt,” Lewis tells the story of how a group of traders banded together to create an exchange safe from high-frequency trading robots that manipulate price and nickel-and-dime investors. To do that, those HFT firms have to send tons of messages to exchanges with their algorithms. To see Lewis debate in detail watch Video 2 above. I am more interested in the technology behind the curtain.
Since the Flash Crash in 2010, software experts, regulators and investors have been talking about whether or not the harrowing, instantaneous 9% decline in the stock market was caused by high-frequency trading firms overloading exchange software systems.
Think of it like driving so fast down the highway that your car spins out of control and takes a bunch of vehicles down with it.
“This is true across industries based on technology,” said Lev Lesokhin of CAST, a company that visualizes inherent risks in financial software systems. “They don’t really … have good oversight of the structure of their systems … Look under the hood of these companies … there’s not enough attention being paid to the programs they’re using, and these programs are a large part of what’s causing the problem.”
Some say that this communication overload is what caused the Flash Crash. Think of it like having tons and tons of browsers open on your computer streaming videos and music. Even if you’re only looking at one, all that information could cause your computer to freeze up and crash.
The faster the HFT firms get, the more complex the algorithms they send to exchanges become. Stock exchanges, in turn, have to evolve to deal with that speed, as well, and write their own programs. More algorithms, more code, more programs, more problems. According to Lesokhin, all of these systems are being built on top of each other.
“The systems are getting too complex for any of the brilliant developers they have building them to manage,” according to Lesokhin.
And those systems are moving at the speed of light. So when the SEC suggests things like automated ‘kill switches’ — programs that would stop trades dead in their tracks if certain triggers were hit — to stop software gone wrong, it’s talking about an afterthought.
Take what happened to Knight Capital back in 2012 as an example. The company lost $400 million in no time at all because new trading code that it introduced to its system woke up old code written into the system. It was Frankenstein stuff. Together, the old code and new code started selling low and buying high at breakneck speed.
Yeah, you can kill switch that, but only after the damage has been done. Until then, it could hurt everyone in the market. This could be why Goldman Sachs has come out in favor of HFT-agnostic exchange IEX, the subject of Lewis’ book.
Lewis wrote that investment banks, exchanges and HFT firms are all colluding to scam investors, but this software issue is one that hurts actors on every side of a trade — even firms like Goldman (and they are very worried).
Image Credit: CC by Chris Isherwood