November 20, 2010
Work on Relativistic Econophyics by Hertz Fellow, Alex Wissner-Gross, has been named 2010 Science News of the Year under the category of technology.
From Science News
by Rachel Ehrenberg
November 20, 2010; Vol.178 #11
Researchers find optimal locations for doing business in multiple markets Forget the trading floor—in the future, an empty lot in Uzbekistan or a barge anchored miles off Chile’s southern coast may be the most lucrative spot for playing the market. A new analysis that takes a particular kind of trading to its theoretical limit finds the precise locations between the world’s major securities exchanges for gaming the speed of light.
In today’s markets, computers search for and act on relevant information in a flash, sending orders through fiber optic cables at nearly light speed. By buying or selling shares split seconds ahead of the rest of the market, holding stock for mere moments and then doing it all again, high-frequency traders are turning fractions of pennies into piles of dollars. To trim time in the extreme, firms will even buy space for their computers as close as possible to an exchange’s computers, a practice called “colocating” that cuts data travel time, giving some traders an edge.
But to exploit the 50-odd milliseconds it takes for information to cross the Atlantic, it turns out that the sweet spot isn’t always at the exchange’s door. For some assets sold on more than one market, such as the New York and London stock exchanges, the moneymaking spot is in the middle of the Atlantic Ocean, researchers report in a paper to appear in Physical Review E.
The team figured out primo locations for performing particular trades on the world’s 52 major securities exchanges. The analysis considers the speed-of-light delay between exchanges, and characteristics of the exchanges themselves, such as volume and frequency of trades.
“One surprising feature is that a lot of these optimal positions are in the ocean or other poorly connected areas,” says study coauthor Alexander Wissner-Gross of the MIT Media Laboratory. “Simply owning or having sovereignty over a certain position on the Earth might turn out to be financially interesting.”
But some choice spots, such as Los Angeles, are already well connected, says coauthor and mathematician Cameron Freer of the University of Hawai‘i at Manoa. For trading some stocks sold on both the New York and Tokyo exchanges, the ideal location is probably already wired.
However, even for hot spots with preexisting infrastructure, it’s unlikely anyone will take advantage of this money-making map anytime soon, says computational finance expert Michael Kearns of the University of Pennsylvania in Philadelphia.
If you take high-frequency trading to its theoretical limit, then ultimately the researchers are right—the limiting resource is going to be the speed of light, he says. But right now it’s mostly hopping from router to router that slows data down, so unless you could run an optical fiber cable right into an exchange’s data center, your orders will still be slowed as they pass from one computer to the next.
There are also regulatory issues, notes Kearns. And while everyone’s abuzz with talk of high-frequency trading, recent research by Kearns and others finds that even though it’s vast in terms of volume, the profitability doesn’t match the hype.
“But if I blur my eyes about those things, well, these guys are doing something that is a sensible exercise,” Kearns says. “On Wall Street, never say never.”
CITATIONS & REFERENCES :
A.D. Wissner-Gross and C.E. Freer. Relativistic statistical arbitrage. Physical Review E, in press, 2010.
***Click to View Article