By: Michael Feldman

Mexico City’s suburbs don't exactly conjure up images of New York’s financial district, but one young trader is looking to close that gap. Alberto Alonso, a 32-year-old entrepreneur, has created a supercomputing system that does the type of algorithmic trading more commonly associated with those on Wall Street. A story on Bloomberg’s news site describes how Alonso came to build his machine, and why he believes the algorithm he developed will be able “mint money” once the system goes into full production.

Alonso, who dresses in jeans and polo shirts, looks more like a Silicon Valley techie than a button-down financial trader. His interest in developing the machine, known as Breogan (a mythical Galacian king), came about when he realized that about a quarter of the local trades on the Mexican Stock Exchange (Bolsa Mexicana de Valores) are international transactions that involve currency conversions. The conversion opens up arbitrage opportunities when small discrepancies in stock prices at different currencies can be identified by a computer fast enough to identify the gap.

That’s where the Breogan comes in. The system is able track currency exchange rates in real-time to find those price differentials and execute a trade in “less than 70 milliseconds.” According to Alonso, the operation has brought in about $475,000 (8.6 million pesos) since February, while working on about one-sixth of the international trade volume performed in Mexico.

Although the Bloomberg story characterizes Breogan as a “supercomputer,” from the provided images, the $350,000 system looks to be a single rack of servers that is small enough to be housed in Alonso’s office. He has plans to scale up the machine’s processing power by investing an additional $50,000.