Program Trading: Meaning, Purpose, Example

What Is Program Trading?

Program trading refers to the use of computer-generated algorithms to trade a basket of stocks in large volumes and sometimes with great frequency. The algorithms are programmed to run and are monitored by humans—although once running, the programs generate the trades, not humans. However, humans can activate or deactivate the program as needed.

Understanding Program Trading

Program trading is defined by the New York Stock Exchange (NYSE) as the purchase or sale of a group of 15 or more stocks that have a total market value of $1 million or more, and are part of a coordinated trading strategy. This type of trading may also be referred to as portfolio trading or basket trading.

Orders are placed directly in the market and executed according to a set of predetermined instructions. A trading algorithm might buy, for example, a portfolio of 50 stocks over the first hour of the day. Institutional investors, such as hedge fund managers or mutual fund traders, use program trading to execute large-volume trades. Executing orders in this way helps reduce risk by placing orders simultaneously, and can maximize returns by taking advantage of market inefficiencies. Placing such a large number of orders by hand (by a human) would not be as efficient.

Program trading accounted for 50% to 60% of all stock market trades placed during a typical trading day in 2018. As of 2021, program trading is estimated to account for 70% to 80% of all U.S. stock market trades placed during a typical trading day, with that number rising to above 90% during periods of extreme volatility.

Program trading has been greatly facilitated by certain realizations in the field of investing, Including:

• The notion that trading a diversified portfolio of securities reduces the inherent risks of investing.
• The fact that institutions hold, and trade, a higher fraction of equity than ever before and program trading allows them to execute their diversified strategies more efficiently.
• Technological advances have reduced trading costs, making program trading more efficient and worthwhile.

Firms may have program trading strategies that execute thousands of trades a day or that only execute trades every few months. Indeed, the volume and frequency of program trading varies greatly by firm, and by the strategy the program is based on. A day trading program will be far more active than an investing program designed to only periodically rebalance a portfolio.

Many market participants blamed program trading for causing extreme volatility that contributed to significant market crashes in the 1980s and 90s. As a result, the NYSE introduced rules that prevent program trades from being executed during certain times to minimize volatility. Program trading restrictions are known as trading curbs or circuit breakers.

Program Trading Purpose

There are several reasons for program trading. These include principal, agency, and basis trading.

• Principal trading: A brokerage firm may use program trading to buy a portfolio of stocks under their own account that they believe will increase in value. To generate additional revenue, they might then "onsell" these stocks to their customers to receive a commission. The success of this strategy largely depends on how successful the brokerage firm’s analysts are at selecting winning stocks.
• Agency trading: Investment management firms that trade exclusively for clients may use program trading to buy stocks that are in the firm’s model portfolio. Shares then get allocated to customer accounts after being purchased. Fund managers may also use program trading for rebalancing purposes. A fund might use program trading, for example, to buy and sell stocks to rebalance a portfolio back to its target allocations.
• Basis trading: Program trading can be used to exploit the mispricing of similar securities. Investment managers use program trading to buy stocks they believe are undervalued and short stocks that are overpriced. A manager could short a group of semiconductor stocks that are thought to be overvalued, for instance, and purchase a basket of hardware stocks that appear undervalued. Profits result when the prices of the two groups of securities converge.

Program Trading Example

Assume that a hedge fund holds 20 stocks in a portfolio and allocates 5% of the portfolio to each stock. At the end of each month, they rebalance the portfolio so that each stock once again represents 5%. They do this by selling stocks that have a higher than 5% allocation, or buying stocks that have a lower than 5% allocation. Some stocks may be dropped from the portfolio, and others added. Any new stocks that are added will be allocated 5% of the portfolio.

Over the course of time, some stocks will rise and some will fall, resulting in a change to the overall portfolio value, as well as a change to the percentage allocation that each one of those stocks represents.

If the portfolio is $10 million, for example, a 5% stake is $500,000. Assume the hedge fund bought Apple Inc. (AAPL) when it was trading at $100, and now it is trading at $200. Assuming all other stocks didn't move (not likely to actually happen, but for demonstration purposes), the position is now worth $1 million, the rest of the portfolio is worth $9.5 million, so the total portfolio is $10.5 million. APPL represents 9.5% of the portfolio ($1 million divided by $10.5 million). A 9.5% allocation is much more than 5%, so shares would be sold to reduce the allocation back to 5%, which is $525,000 (5% of $10.5 million).

Now, imagine that all 20 stocks are moving every day, and at the end of each month some will be worth 5.5% or 6%, and others will be worth 4% of the portfolio. A program trading algorithm can look at the portfolio equity and quickly execute all the trades at once, buying the stocks that are under-allocated and selling the ones that are over-allocated to rebalance the portfolio in seconds. Manually doing this would be much harder and more time-consuming.