Twitter, despite its reputation for occasionally spreading fake stories, remains one of the best ways to keep tabs on the latest news, financial or otherwise. And this morning, a handful of tweets from “HFTAlert” on the latest high-frequency trading shenanigans caught our attention, retweeted via Stocktwits:
“@HFTAlert: spamming algo hitting $BAC again… this time 830,000 quotes in 34 seconds.”
“@HFTAlert: now they’re blasting $KEY, now 1.4 million quotes. $BAC”
“@HFTAlert: $KEY algo finally stopped… ran for 5 minutes. 1.6+ million quote changes…”
A more detailed breakdown of this morning’s quote-stuffing activity (along with some charts for perspective) can be found here. It’s likely to just be another blip in the radar, but these blips are starting to add up. We’ve seen calls for regulatory intervention from Congress, industry folks, and regulators, and there are even TED talks out about it. These blips might not be much to look at right now, but the TED talk, in particular, makes a good point towards the end of the video.
The speaker, Yan Ohayan, talks about Watson, the Jeopardy-playing computer program built by IBM. He points out that after Watson defeated the past human champions, they claimed that Watson’s advantage wasn’t that the computer knew more answers than the humans – it was that Watson could simply buzz in faster than they could.
In part, the same idea is what gives high-frequency traders their advantage – the machines aren’t actually “better” traders; they’re just faster traders. As long as the exchanges continue to allow or encourage practices like co-location and issuing tens of thousands of cancelled quotes per second, it will remain an advantage to trade milliseconds faster than other market participants.
Of course, high-frequency traders have other advantages that Watson lacks. They can “cheat” by putting in fake quotes to make someone move their price up or down, then cancel those quotes and take the other side of the trade. Sort of like if Watson could hit the buzzer 1000 times per second in Jeopardy to confuse the other players, and then hear their answers and jump ahead of them in line.
Ted talks are sometimes derided as an intellectual snob-fest, but Ohayan certainly makes us wonder – how many blips will it take before we see a real response to the HFT algos? Or will it take another flash crash to propel changes?