With the publication of Michael Lewis’ new book, Flash Boys, there’s been a lot in the news lately about high frequency (or high speed) trading. (For those who aren’t familiar with Lewis, he is also the author of a couple of very good sports-related books – Moneyball and The Blind Side – that were made into a couple of very good sports-related movies – as well as of Liar’s Poker, an earlier book about Wall Street.)
I’m not going to get into Lewis’ argument – that high frequency trading, which exploits a 1-2 millisecond – and even fraction of a millisecond – advantage in transmission speeds from exchange to exchange, basically rigs the game.
What’s really interesting about this, from my point of view, is what’s going on behind the scenes with the systems that enable the trading, which often rely on FPGAs.
PC Magazine had an article by Mitchell Hall on this last fall that did a good job explaining things:
On a basic level high-frequency traders use a combination of hardware and software to see how much someone else is willing to buy or sell a given security for fractions of a second before their competition does. They can then trade accordingly. It’s almost like being able to bet on a horse race from the future; you already know who’s crossed the finish line first.
While software algorithms play into things, the real advantages come on the hardware side and the drive for zero latency:
High-frequency trading is at the forefront of hardware acceleration. This means using hardware-accelerated network stacks and NPUs (network processing units), or custom-designed FPGAs. NASDAQ offers its ITCH data feed processed by an FPGA instead of a software stack. “Some people have put market data processing into an FPGA,” says [former high-frequency trader David] Lauer, “some people have put trade logic into an FPGA, risk controls into an FPGA; some people are using GPUs to do massively parallel analysis on the fly. So there’s some pretty interesting stuff going on in hardware acceleration that I think is more cutting edge than most industries.”
FPGAs have been around for about 30 years. Interesting that they still remain such a workhorse, isn’t it?