Algorithmic FX Trading : News Analytics - Capturing market opinion for high performance FX

First Published in e-Forex Magazine July 2012

Nicholas Pratt

As the use of machine readable news becomes more accepted among FX traders, Nicholas Pratt examines the challenges facing the providers of such services from sentiment to speed to social media.

Share |

This magazine first wrote about machine readable news in 2008. Back then it was a relatively nascent technology within the FX market. The main data providers - Dow Jones, Thomson Reuters and Bloomberg - had all developed elementised news services so that their news-based data could be fed into traders' algorithmic trading engines. Similarly the trading engine providers had adapted their architecture to be able to receive news feeds rather than market data feeds and then incorporate these feeds into new or existing trading strategies.

What was clear at this time was just how much more development was possible. These same offerings are now being extended to provide more digestible analytics which can be easily exploited by high performance FX trading firms. The next generation of algorithms are also being designed with machine readable news in mind and the ability to effectively exploit the increasing amount of news data available in the market.

More has also now been achieved in terms of automatically tracking sentiment, political bias, credibility and other more nuanced and qualitative features. And a number of providers have stepped into the MRN market to offer complementary services designed to help with the dissemination and interpretation of news or to bridge the gap between the news providers and the trading engine developers.