What Is Algorithmic Trading?
Algorithmic trading is typically automated and is commonly referred to as automated trading. It allows investors to process vast amounts of data—usually focusing on time, price, and volume—across different markets with speed and efficiency. Trades that can be executed at a fraction of a second using algorithms are what’s known as high-frequency trading, which is a subcategory of algorithmic trading.
Institutional investment firms—including registered investment firms, insurance companies, mutual funds, pension funds, and hedge funds—typically have the resources and expertise to design and manage the algorithms required for algorithmic trading. Institutional firms that don’t have such resources can purchase or use firm-specific algorithmic trading programs from brokers and other third-party providers.
Algorithmic trading and electronic trading platforms have largely replaced the traditional form of trading performed by humans on the trading floors of exchanges such as the New York Stock Exchange. Nowadays, algorithmic trading accounts for the majority of equity trading on all exchanges in the U.S. Exchanges usually work with trading firms to execute trades, and some charge a flat fee to facilitate orders.
What Is an Algorithm?
An algorithm is a program with a set of instructions to perform functions with an end result. In its simplest form, for example, a computer programmer can write the code for a program to purchase a certain number of shares at a certain price and time on a specific exchange and then later sell those shares at another price to produce a profitable trade. A more complex algorithm could include orders to sell different stocks on separate exchanges at different times of the day or simultaneously.
A Brief History of Algorithmic Trading
Algorithmic trading can trace its origins to as early as the 1970s when the Nasdaq stock market began operations with automated trading. In the 1980s, a broker named Thomas Peterffy programmed a personal computer with an electronic feed from a stock data terminal called Quotron to search for options relative to fair value that would calculate the best spreads to buy or sell. He pioneered trading options on the S&P 500 Index and eventually started an electronic brokerage called Interactive Brokers, which operates the largest electronic trading platform in the U.S.
In 1998, the Securities and Exchange Commission relaxed regulations on alternative trading systems that paved the way for the widespread use of computerized high-frequency trading, and algorithmic trading gained momentum by the late 2000s. In 2010, an algorithmic trading program went awry, with erroneous trades that sent the stock market plunging momentarily before recovering quickly in an event known as the Flash Crash.
How Does Algorithmic Trading Work?
An algorithmic trading strategy known as price action might look at high and low prices of a stock and analyze where the price might be in the future. Another strategy might rely on technical analysis, such as moving averages and relative strength index, to determine how much a stock’s price will rise or fall. Another strategy might combine price action and technical analysis, in which technical analysis could be used to confirm a price action strategy.
Common Types of Algorithmic Trading Strategies
Algorithmic trading accounts for the majority of trading on all exchanges in the U.S. Below are a few of the most common strategies used by algorithmic traders.
High-Frequency Trading Algorithm
Algorithmic trading tends to account for a large portion of high-frequency trading, which executes trades at very high speed to take advantage of price discrepancies that exist for a short period of time.
Market-Making Algorithm
This algorithm provides liquidity to the market and ensures that there are enough buyers and sellers in the market by placing buy and sell orders at slightly different prices from the current market price.
Trend-Following Algorithm
This algorithm identifies trends of a stock’s upward or downward movement and seeks to capitalize on the trend’s continued action.
Arbitrage Algorithm
This algorithm tries to take advantage of price differences of a stock trading in separate markets and exchanges. Just like arbitrage trading, the program seeks to buy a stock on an exchange at a lower price, and sell the same stock at another exchange at a higher price to make a profit.
Volume-Weighted Average Price Algorithm
Institutional investors that seek to minimize the impact of their trades at large volumes of stock would conduct trades at the volume-weighted average price (a stock’s average price for a period based on the weighted volume of trading) over a particular period.
Does Algorithmic Trading Improve Trading?
Market-making algorithmic trading can narrow bid-ask spreads. It can also improve price efficiency and reduce the time necessary for prices to adjust to newly released information. Automated trading, including algorithmic trading, has largely replaced humans who used to conduct trades on the floors of exchanges. Automated trading accounts for almost all of the trading volume on U.S. exchanges.
What are the Limitations of Algorithmic trading?
Algorithmic trading can go wrong when incorrect information in the programs leads to bad trades, as the Flash Crash in 2010 demonstrated. During periods of unusually high volatility or market stress, such as the 2020 market crash during the COVID-19 pandemic, when there were no market makers on the floor of the NYSE, the use of algorithms could create large up and down price swings.
