The rise of the machines

by. Melanie Friedrichs

A couple of weeks ago I was having dinner with a friend, and we got to talking about algorithmic trading, also known as high-frequency trading. Wikipedia defines algorithmic trading as the “use of electronic platforms for entering trading orders with an algorithm.” Algorithmic traders make money by identifying trades that will predictably yield an insignificant profit, and then using computers and leverage to execute those trades at a speed and volume high enough to make a significant profit.

Algorithmic trading is a relatively popular topic these days, primarily because it’s making a some people a lot of money, but also because algorithmic traders make money by leeching off the inefficiencies of the system, and are therefore everyone’s favorite scapegoat. Even Mr. Potter probably could have shook his finger at algorithmic traders without feeling like a hypocrite. However, defendants argue that by improving the price signal and reducing volatility, algorithmic traders do provide value, and I’m inclined to agree.

Despite its questionable morality, algorithmic trading has grown considerably over the last 20 years. From 2008 to 2011 about 2/3 of all US stock trades were executed by algorithmic trading firms, although this ratio has fallen over the last couple years as existing dealers, investment banks, and even individuals, have started to adopt the techniques of once-boutique algorithmic trading firms. And even though profits have also fallen as the industry gets more competitive, algorithmic trading is increasingly becoming a prerequisite to survive; when all your peers are trading with computers, you’re dead if you’re not trading with computers too.

At dinner my friend referenced a paper published in Nature in 2012 by a group of physics professors from the University of Miami called… (this title deserves a drumroll)…

continue reading »