Up until now, machine-readable news has been largely championed by latency-sensitive traders looking to give their automated trading strategies some extra edge.
But as machine-readable news evolves, it will not be just high-frequency and quant traders that benefit from turning news into quantifiable, actionable data, according to Rich Brown, global business manager, machine-readable news at data vendor and analytics provider Thomson Reuters.
Machine-readable news applications, such as Thomson Reuters’ NewsScope Direct, analyse and quantify news stories such as economic data releases so that they can act as automatic triggers for trading decisions. Having news interpreted and analysed automatically at a low latency gives traders the chance to act fractions of a second before it has the chance to move the markets.
Recent technology-led service upgrades have increased demand, says Brown. Since Thomson Reuters expanded its network to include co-location sites in Chicago and London, in addition to New York, more proprietary trading houses and hedge funds have used the service to power foreign exchange and futures trading strategies. Despite its restrictions on high-frequency trading, interest has spread to Brazil, where more than 40 local firms recently attended two quant trading forums in which Thomson Reuters participated.
Now, the uses for machine-readable news are starting to evolve beyond the high-speed world.
“We are increasingly looking to bring the product suite to the mainstream,” says Brown. “The acceptance of news analytics for both trading and investment strategies has increased dramatically in the last year. As we continue to demonstrate the value to longer-term investors, portfolio managers and traders will be able to work more closely together to analyse alpha generation and scrutinise investment decisions.”
For instance, using Thomson Reuters’ NewsScope Analytics tool, traders can use a sentiment engine to track and anticipate the impact of news on market trends. This works by weighting news items according to factors such as tone – i.e. how positive or negative it is – its projected impact on stock prices and how new or repetitive the news is. Long-only traders and their portfolio managers can then use this analysis to inform trading and investment decisions.
“Clients can create a customised index based on the parameters that are most relevant to their trading strategies,” says Brown. “They can look at the impact of IBM on the tech sector, or the tech sector compared to other sectors, for example, and adjust strategies based on news items of interest.”
The NewsScope Archive, which has news content dating back to 1987, can be used to back-test trading algorithms and analyse how news events have affected markets historically. It also supplies another source of transaction cost analysis by allowing users to track the flow of news against market movements in the event of adverse algorithmic performance. And because data is added to NewsScope on a T+2 hours basis, algorithms can be reprogrammed with updated inputs between the market openings of different regions.
Other potential applications include helping regulators identify market abuse as they struggle to police markets under the weight of multiplying market data volumes. Using NewsScope, market surveillance teams can look for unusual trading patterns prior to formal company announcements, with a view on how the price of a stock would have behaved in normal circumstances.
Printer-friendly version
