In a bid to push further into computer-driven stock trading, Reuters is launching a new market data product that scans company news articles, scores how positive or negative they are and then, based on the information, triggers orders in algorithmic trading applications.
The machine-readable service, which is based on linguistics technology developed by Corpora, scans and analyses news across thousands of companies. Reuters says the service will enable computer-driven trading systems to react to market moving news in milliseconds.
The system works by assigning numerical "sentiment scores" to words or phrases which are then processed to give an overall positive, neutral or negative score to the company in the news article. These scores are added together to calculate the prevailing sentiment for a company, a sector, an index or to assess global market sentiment.
Peter Moss, global head of enterprise solutions at Reuters, says the system allows machines to interpret the sentiment of news stories as they are published, enabling them to react in an informed way to market events at ultra-high speed.
"Imagine a machine scanning hundreds of stories on companies' results, measuring the sentiment around them and incorporating that into algorithmic trading strategies," says Moss. "This powerful capability can have a meaningful impact on a client's portfolio performance."
Reuters says the product will be targeted at the high-growth area of algorithmic trading and will be connected to its Reuters Market Data System (RMDS) platform.
Last December Reuters launched two products that allow its streaming news and data output to be used for the purposes of automated and algorithmic trading. The products are designed to enable computers to recognise and process key items of data in Reuters news stories, allowing customers to incorporate the information into automated trading strategies.
The UK vendor has also recently expanded its exchange data service, RDF-Direct, to include more European and Asian venues in response to customer demand for full tick data that can be used for algorithmic trading applications.
Meanwhile earlier this year Dow Jones launched a data feed that delivers news items in XML format which can be rapidly processed by electronic trading programs
In a separate move Reuters says it is acquiring all of the outstanding shares of Massachusetts-based ClearForest, which makes software used to search archives of news, Web pages and documents.
ClearForest's text analytics software is used to tag and categorise unstructured information to extract specific facts about people, organisations, places or other details from news articles and other documents.
Gerry Campbell, global head of search and strategic development, says: "This technology has the potential to impact the user experience across all Reuters products. ClearForest has developed a superior, commercially proven technology in the field of Text Analytics that will make it immensely easier for our customers to find, understand and navigate our content."
Terms of the deal have not been disclosed. Reuters says ClearForest has received sufficient shareholder approval to complete the transaction, which is expected to close in 30 days.
The UK vendor says it plans to retain and continue to work with ClearForest's existing management team and workforces in the US and Israel.