Goldman’s Equity Algo Army Goes Offline After ‘Mistake’ Triggers Trading Limits

At its height back in 2000, the U.S. cash equities trading desk at Goldman Sachs’s New York headquarters employed 600 traders, buying and selling stock on the orders of the investment bank’s large clients. Today there are just two equity traders left.

That’s from a pretty memorable February 2017 piece in the MIT Technology Review.

Essentially, it describes an environment where carbon-based traders are being replaced by algos, although to be fair, those algos are “supported” by dozens of computer engineers, so it’s not a complete wipeout of humanity.

This mirrors an epochal shift that’s been underway for years and has become so ubiquitous that it’s now the subject of market memes utilizing the first scene from Terminator 2 to visually describe what the future will look like.


Of course at the top of everyone’s list of concerns when it comes to this brave, new world governed as it is by hair-trigger, headline-chasing algos, is the possibility that the machines occasionally go “rogue” for whatever reason.

Well on Wednesday morning, Goldman’s algorithms for executing cash equity trades (so that would be the algos that replaced the people mentioned in the MIT article cited above) “went offline briefly after a mistaken order triggered their trading limits,” a person familiar with the matter told Bloomberg.

A spokesman for Goldman did confirm the report, adding that trading was merely “paused after trading limits were breached.” In other words, this wasn’t, according to the bank anyway, “a technical fault with the algorithms.”

Notably, Goldman says this has nothing to do with today being the first trading day of MiFID II.

One certainly imagines we’ll get all manner of these type of algo “halts” once Goldman and the rest of the blue chips get in on cryptocurrency market making.

For now, everyone can relax because apparently, Goldman’s equity algos are now working again.

Smile… and hope they don’t become self-aware…



Speak On It