Gaming the System With High-Frequency Trading 10th Nov 2010
Commercial broadband infrastructure in the US is as poorly tended and updated as the nation’s bridges and highways. Unless you’re a high-frequency stock trader. Then, you get to ride in the fast lane.
“Ultra-fast cables are not built for use by the public,” writes Thomas McCabe, who wrote a recent profile of the
These companies are building new fiberoptic cables over land and under sea to connect major trading centers and shave milliseconds off trading times. Automated trades and super-fast infrastructure lets traders pull off a very simple kind of global arbitrage: buy shares on one exchange and sell them on another, using huge numbers and a momentary informational lag to turn penny differences on individual shares into millions of dollars in profit on a trade.
In the near future, the fundamental limit on high-frequency trading won’t be the cables. It’s the fact that human beings (and their computers) still live in cities on dry land. At an even further extreme, it’s the fact that no information can travel faster than light.
When trading centers are as close as London and Paris, that’s not much of a problem. If they’re as remote as London and Sydney, it places limits on speed — and gives traders who are geographically closer a key advantage.
The solution is outlined by MIT’s Alex Wissner-Gross and the University of Hawaii’s Cameron Freer in a recent theoretical paper titled “
“Historically, technologies for transportation and communication have resulted in the consolidation of financial mar- kets,” Wissner-Gross and Freer write, pointing to the elimination of regional stock markets with the rise of the telegraph. “We have described a type of arbitrage that is just beginning to become relevant, and for which the trend is, surprisingly, in the direction of decentralization.”
Indeed; once you eliminate the need for people to gather together, the effect of traditional economic geography on markets is as minimal as an algebra problem.
Image: Submarine Cable Map 2010. Credit: PriMetrica via Kurzweil AI.