Automatic Trading Linked to News Events

The news can be pretty hard on stock prices.

Announcements about corporate profits (or lack thereof), a change in management, rumors of a merger, even the weather. All these events can cause a company’s share price to move wildly up or down.

Today, the Reuters Group begins selling two trading services that allow subscribers to set up automatic trading orders based on the news. They will give subscribers the ability to mine past and present Reuters news articles in real time and automatically buy, sell or hold a stock based on market-moving events.

The mining and sifting of news take place in computers dedicated to algorithmic trading; that is, automatic buying and selling based on complex mathematical formulas that aim to pick the optimal time to trade a stock.

Although algorithmic trading on news events already exists, Reuters says its system is the most advanced. “If an event breaks somewhere in the world, you want to be able to respond to that, or manage the event risks,” said Richard W. Brown, business manager for Reuters NewsScope, as the new program is being called. “At a very simplistic level it’s about speed.”

The price for the services will vary based on how investors use them, starting at several thousand dollars a month, Reuters said.

Hypothetically, trading with Reuters NewsScope would work something like this:

Investors trading shares of Ford Motor know there are certain events that cause the share price to rise or fall — sudden changes in gasoline prices, a labor strike at a parts supplier, a poor month for automobile sales, for example. Reuters NewsScope will allow investors to fine-tune their computers to scan live news feeds of Reuters coverage and watch for items affecting Ford.

Then, through a separate NewsScope program, traders could further adjust trading parameters by scanning Reuters news archives for articles about past announcements from Ford. The database allows traders to analyze how a company’s stock performed after those announcements and to look at the price and volume of shares traded. Investors would program their computers to buy or sell stock automatically based on historical patterns.

Algorithmic trading, also known as programmed trading, is not a new phenomenon, but it is increasing in popularity. A third of all stock trades in the United States were driven by automatic algorithms last year, some studies estimate.

The Reuters software will give large investors like hedge funds, which use algorithmic trading most frequently, another tool in the hypercompetitive world of institutional investing. John Bates, who developed algorithmic trading software for his company Apama Products, a division of Progress Software, said that as trading technology improves, investors not using the latest software could put their business at risk.

“We’re talking about milliseconds,” Mr. Bates said. “The latency between receiving data in, making a decision that a rule has been met and placing an order, should be milliseconds.”