High-Frequency Trading

High-frequency trading (HFT) is the practice of executing orders at sub-second time scales, using algorithms that exploit small price discrepancies, liquidity imbalances, and the mechanical features of market infrastructure. It is not a trading strategy in the traditional sense; it is a latency arms race in which the competitive advantage is measured in microseconds, and the primary strategy is being faster than the next algorithm.

The emergence of HFT transformed market microstructure from a field of academic interest into a domain of active engineering competition. HFT algorithms do not forecast fundamental value. They forecast the order flow of other participants — detecting large orders as they are fragmented across venues, anticipating the replenishment of liquidity, and positioning to capture the spread before slower competitors. The result is a market whose price dynamics at short time scales are dominated by strategic interaction among algorithms rather than by information about fundamentals.

The regulatory and ethical debate about HFT is muddled by the confusion of speed with value. Speed is not a social good; it is a private good that is extracted from the market's collective latency. The question is not whether HFT provides liquidity — it does, but it also withdraws it precisely when it is most needed. The question is whether a market whose microstructure is designed to reward speed is a market whose prices are better, or merely a market whose profits are more concentrated.