The rise of machine readable news?

Machine readable news has been making slow gains in the financial services over the past few years, though it still strikes many as more of a novelty. But you don't want to suggest as much to Thomson Reuters, which has been investing in its NewsScope product since 2007, when it bought ClearForest. The idea is to aggregate and systematize the flow of news in ways that are useful to traders and portfolio managers, generating buy and sell indicators essentially on the fly. The only way to make it work of course is to deliver the service fast, before the market has reacted to the information. 

Thomson Reuters has recently launched NewsScope Direct, a machine-readable news service aimed at high-frequency trading outfits. The service will be located in London and Chicago and aims to give clients a way to "integrate news and events into a variety of trading strategies." The goal is to give high frequency traders an edge when it comes to the foreign exchange and futures markets. 

To some, all this might seem fantastic. The concept conjures images of computers "reading" the news, applying artificial intelligence and making trading decisions. But it's not really all that far-fetched to tag and order news items in something close to real time and then deliver the results to the end user in time to make profitable decisions. One application would be to use machine-readable news as a circuit breaker of sorts. For example, if you are about to make a huge trade in GE and a rush of GE news hits the airwaves and Internet, the service may throw up an alert, giving you time to determine if the news has any relevance for you. There could be myriad ways to tag the news logically--company and event identifiers, topics, stage of the story, and other valuable metadata--so that clients can easily combine these factors to generate trading signals for everything from volatility based strategies to those incorporating signals for returns. 

Thomson Reuters is not the only firm developing machine-readable solutions. Dow Jones has been in the hunt, as have a host of smaller firms. But Thomson Reuters for now anyway has created the most buzz. 

It's fair to say that this really hasn't become a must-use technology. Not yet anyway. But the industry is in constant need of an edge. And some are more than willing to explore the benefits. 

Consider CA Cheuvreux, Credit Agricole's European broker that trades in 60 countries. It has been experimenting with machine readable news. More to the point, the firm is using "machine-readable, XML-tagged news feeds for better execution." The first step is to "to identify automatically very strange, highly volatile days, and react immediately, faster than traders could, to news items that will lead us to believe the stock will move quickly. The algorithm will then choose pre-set rules based on those market conditions. The second phase is to derive directional signals, whether the market is going up or down in the short term." These signals would be delivered quickly to traders, who could then act on them. 

We'll likely see more experimentation. - Jim