In general terms "algorithmic trading" means the use of a predefined set of rules written into computer code to automate the process of buying or selling. By this general definition, the term algorithmic trading could apply to any computer generated market operation. However, the term is most widely accepted to refer to order execution, irrespective of whether the origin of the actual underlying trading decision is computer generated or not; whereas the trading process where the computer entirely replaces the human trader, taking both the trading decision and (often, but not always) managing execution is refered to as "automated trading", "systematic trading", or sometimes "black-box trading".
Execution algorithms are used primarily to achive the best possible execution price (Best Execution) or minimize the market impact of sometimes very large orders, identify and exploit liquidity across multiple trading venues, and to prevent information leakage. As algorithmic trading has proliferated, an incredibly wide range of algorithms have been developed to address a very broad spectrum of requirements and achieve very specific objectives. Typical objectives might include:
- Disguise large orders by trickling volume into the market in small clips under optimal circumstances.
- Identify venues with the best liquidity for a specific instrument at any given time of day.
- Manage the execution of an illiquid instrument
- Identify potential hidden pools of liquidity.
- Identify the price levels of blocks of stop orders.
Whilst price and volume data are the most common inputs to an algorithmic trading strategy, many algorithms now take data inputs including sentiment indicators, fundamental data, and inputs from other instruments and asset classes; whilst the behaviour of many algorithms can be altered 'on the fly' with discressionary inputs as the trader's view of the market changes.
One of the key drivers of the proliferation of algorithmic trading in equity markets has been the regulatory environment. As regulators have sought to create more efficient markets, improve transparency and competition, directives such as Regulation NMS in the United States and MiFID in Europe have resulted in a proliferation of equity trading venues, meaning that a considerable amount of liquidity has moved away from the traditional national exchanges to alternative trading venues. The result has been that with liquidy fragmented across multiple venues, both buy-side and sell-side firms have been forced to embrace algorithmic trading in order to remain competitive, with the latter also having some best execution obligations imposed by regulation.