FT reports that new trading systems will be driven by news feeds, as well as historical data. Systems would pick out places where a company is mentioned, and predict what might happen to the stock price as a result, and use that information as the basis for buy-or-sell decisions. The hope would be to avoid some of the volatility that recent events have shown are caused by automatic trading systems (at least half of stock trades are now generated algorithmically, not by human decisions on the spot).

Unfortunately, as the article notes, events such as the precipitous drop in United Airlines stock price last fall were caused by reactions to incorrect news items. So we had better hope that the news analysis algorithms will be smarter than human judgment at figuring out which news items should be regarded as trustworthy. (Or perhaps, in the UAL case, how long after selling on the basis of a false rumor to wait before buying back in again, just before everyone else figures out the rumor was false.)

This entry was posted on Friday, January 16th, 2009 at 5:27 am and is filed under The explosion, The Internet and the Web. You can follow any responses to this entry through the RSS 2.0 feed. Both comments and pings are currently closed.