Measuring Arbitrage in Milliseconds (WSJ Blog)

요즘 HFT에 대한 기사가 이곳 저곳에서 나온다.
심지어 "Latency arbitrage"라는 단어까지 나온다.

WSJ Blog에 올라온 기사이다.

일단 최근까지 Citadel이나 르네상스 테크와 같은 Hedge Funds들이

have paid for “proximity hosting,” the installation of their servers in the same data centers as the exchanges, to shave more “microseconds,” or millionths of a second, off that latency.

이유는 속도경쟁이지만, 전략은 "Latency arbitrage"이다. 즉 제일 빠른 사람이 승자가 되는 것이다.
실행과정은 이렇다.

the first step is to maximize the speed of market-data software and exchange connections, and the second is to find the “golden egg” - a venue or counterparty whose information is slower.

Many quantitative funds, which rely on computer models for trading, have the ability to get in and out of large stock positions in milliseconds, which means buying a stock at current prices at an exchange and selling it in a dark pool after an uptick you know is coming is a cinch



기사원문이다.
http://blogs.wsj.com/marketbeat/2009/03/09/measuring-arbitrage-in-milliseconds