High-frequency trading, or HFT, is a subset of algorithmic trading in which speed and rapid order activity are central to the approach. Firms may use specialized software, fast data feeds, and infrastructure close to trading venues. Definitions differ across research and regulatory contexts; there is no universal trade count that makes every strategy HFT.
What the systems do
Strategies can include market making and exploiting short-lived price differences. A market-making system may frequently update buy and sell quotes as prices and its inventory change. Many submitted messages are cancellations or modifications, so message volume is not the same as completed trading volume.
HFT often involves short holding periods, but it is not synonymous with all automated trading, direct market access, or market manipulation. The specific behavior matters when assessing an activity.
Costs and market effects
Small expected gains can be erased by fees, infrastructure costs, unfavorable price moves after execution, or inventory exposure. Software failures can generate losses quickly, making testing, position limits, and monitoring important.
HFT can supply or consume liquidity. Its effects on spreads, price discovery, and market stability depend on the strategy and conditions; high speed alone guarantees neither profits nor continuously available liquidity.