Hedge funds, proprietary trading desks in i-banks, etc. have been using automated trading strategies for decades. But on the the agency equity trading side (the

Hedge funds, proprietary trading desks in i-banks, etc. have been using automated trading strategies for decades. But on the the agency equity trading side (the guys who move institutional investors in/out of larger long-term stock positions --- and the guys who are the driving force behind all the dark pools), it's still at least somewhat new. Institutional clients were reluctant let a computer trade their big order. They wanted a human to be deciding how much to quote, when to trade a little fast/slow in response to market conditions, etc. And up to the early 2000s, agency "trading strategies" were still shockingly primitive. Marketing literature claimed they "intelligently managed market impact risk vs. volatility risk". Fancy way to describe a glorified time slicer. Human traders (not to mention prop algo strategies) earned easy money gaming these dumb things. Eventually, the "human advantage" faded away because the humans in question were being expected to handle more and more volume (to earn more and more commissions) --- they just didn't have enough time to pay lots of attention to each client order and the prices being quoted all over the place on all the new alternative venues. And so the banks finally had an incentive to pour money into making their algorithmic trading systems less primitive.