Algorithmic Trading Explained

Algorithmic trading, often called algo trading, means using computer programs to decide when, what and how much to trade, and often to send the orders automatically. Algorithms now account for the majority of trading volume in major equity, futures and currency markets. They range from simple rules that buy when a moving average crosses, to execution algorithms that slice large orders, to high frequency systems that react in microseconds. Understanding the landscape helps any trader, whether they build algorithms or simply trade alongside them.

Main types of algorithmic trading#

How much trading is algorithmic?#

Estimates vary by market and definition, but studies and industry reports commonly estimate that algorithms account for well over half of US equity trading volume, and similar or higher shares in futures and spot foreign exchange. Much of this is execution algorithms used by institutions and market makers providing liquidity.

The parts of an algo trading system#

Benefits#

• Discipline: rules are followed without emotion. See Discipline.
• Speed and scale: monitor many markets and react instantly.
• Consistency: the same conditions produce the same actions.
• Backtestable: rules can be tested on history. See Backtesting Methodology.
• Lower execution costs for large orders.

Risks#

• Bugs and errors can cause large losses quickly.
• Overfitting produces strategies that fail live. See Overfitting and Curve Fitting.
• Technology failures: data outages, connectivity problems.
• Crowding and flash events when many algorithms react together.

The 2010 Flash Crash#

On 6 May 2010, US stock indices fell about 9% within minutes and then mostly recovered. Investigations pointed to a large automated sell order in E-mini S&P 500 futures interacting with high frequency traders and thinning liquidity. Regulators later introduced circuit breakers and other safeguards. See The 2010 Flash Crash.

Getting started as an individual#

Retail traders can build simple algorithms using broker APIs, Python and platforms like TradingView alerts or MetaTrader expert advisors. Start with simple rules, paper trade first and add risk controls before risking real money. See Python for Trading, Building Trading Bots and Paper Trading.

Frequently asked questions#

What is algorithmic trading?#

Using computer programs to decide trades and often send orders automatically, based on rules, models or execution logic.

Is algorithmic trading profitable?#

It can be, but most of the profit goes to firms with real edges, fast technology and low costs. Simple retail algorithms can help with discipline but are not automatically profitable.

What are the risks of algorithmic trading?#

Software bugs, overfit strategies, technology failures and sudden market moves when many algorithms react together.

Next, learn the difference between algorithmic and fully automated trading in Automated vs Semi-Automated Trading.