Algorithmic trading uses computer programs to automatically submit and manage orders according to predefined rules or strategies. In modern markets, it covers a wide range of activities, from simple rule-based routing to complex strategies that determine when, where, and how much to trade. Programs can execute in milliseconds or microseconds, route orders across multiple venues, adjust orders in response to price and liquidity, and report trades to venues and brokers. Because it interacts with market microstructure, it is closely tied to latency, order types, and the capacity of venues to match or reject orders.
Traders implement algorithms to achieve execution objectives such as reducing market impact, improving execution quality, or exploiting short-lived price opportunities. Common elements include routing logic (which venue to send to), timing rules (when to send, modify, or cancel), and risk controls (limits on size, price tolerance, and maximum drawdown). Typical strategies include time-based approaches (e.g., splitting a large order over time) and liquidity-seeking approaches (targeting venues with available depth). Many programs also use awareness of tick sizes, venue fees, and the current market state to optimize routing and execution.
The practice affects liquidity provision and price formation and is subject to market rules and regulation. Firms rely on governance, monitoring, and risk controls to maintain orderly operation and to address latency, routing errors, or unintended market impact.
In a typical setting, an institution's algorithmic system submits a large order across multiple venues following predefined rules to manage execution over time.
High-Frequency Trading · Latency · Order Types · Market Microstructure · Execution Venues