“If Wall Street algorithms trade stocks in milliseconds, does the literal physical distance of a server to the stock exchange dictate who gets rich?”
Michael Lewis would tell you “Yes,” but that’s an outdated answer. At some point it doesn’t make sense to play the speed game (e.g. digging holes through mountains to be 1us faster) because it’s just not scalable. There’s been a shift to AI/ML in the past decade, that’s why you rarely hear about HFT in the news anymore. But smarter strategies are slower by nature (complex things take longer to compute). The top quant firms are still really fast, but they’re not trying to be *the fastest* if that makes any sense. As in, they’re not playing purely a dumb speed game anymore.
For more context, speed has always been a thing in trading, even before the markets were electronic. The Pony Express, the telegraph, etc. People hired faster runners to place trades before everyone else.
Modern technology made things so close to the speed of light, to the point where being a microsecond faster just doesn’t yield the same ROI anymore. Thus the focus on smarter strategies. Also many of the fastest routes are commoditized today (vendor-owned).
When I worked in HFT, the most unfair things I experienced were from brokers and lawyers and accountants. Not from access to speed (colocation). There’s a lot of misinformation on Reddit lol