Last week, Traders reported on the new update to Thomson Reuter Eikon, the trading platform that now includes sentiment analysis for news and Twitter feeds. Since everyone from the buyside to the sellside has tried to use this social media application for trading — a few hedge funds based on Twitter feeds have come and gone – we had a few questions on how helpful Twitter feeds can be for traders who are already drowning in data.
Traders spoke with Philip Brittan, chief technology officer and global head of platform for Financial and Risk, Thomson Reuters, who shed some light on how Twitter can stop gaming and tell the difference between a real tweet and one from a hijacked account.
How does this updated version of Eikon handle gaming via Twitter – the bogus tweets from investors trying to drive up the price of stocks?
Philip Brittan: This is about volume. Not only do we incorporate feeds from both Twitter key influencers and StockTwits but we also include a feed that takes a significant sample of all Twitter volume and this helps avoid distortions from individuals trying to manipulate sentiment. Also the app is designed as a way of spotting activity and trends which we then expect our customers to investigate further – hence we give them the capability to drill down to the individual tweet level and see what is happening behind the scenes.
We are constantly monitoring activity on twitter and we have the ability to blacklist certain Twitter or StockTwits accounts if necessary.
How does this handle hijacked Twitter accounts, like the bogus AP report of an explosion at the White House last year?
Brittan: Whether a stir within the social media universe stemmed from a hacked account or not, a user still wants to know what caused the market to move. Just as bogus news is likely to break first on twitter, the real facts are likely to also surface on twitter before traditional media outlets.
With the situation you mention, the AP tweet was tweeted by one trusted account and then re-tweeted widely by the general Twitter population – anyone looking closely at the data would notice that the news was not picked up by other trusted news outlets, setting alarm bells ringing.
Again this app is designed as a way for our customers to spot trends which they can then investigate further by drilling down to the underlying data. Our main objective is to alert the user that something is going on – its then up to the user to decide if this is the right opportunity for them to execute a trade. At this moment in time we dont think that Twitter sentiment is suitable for machine-based trading but this will likely evolve.
Does Eikon have a snark filter? How does it handle sarcasm, negativity and all of the bile that can flow on social media?
Brittan: Partly this is countered by the weighting system we employ that gives more credence to trusted sources and influencers. It is also countered by volume – when you are looking at data in volume, sentiment that is going counter to the main is flattened out.
We are constantly monitoring activity on twitter and we have the ability to blacklist certain Twitter or StockTwits accounts if necessary.
Can traders automate this – can they execute trades based on the sentiment automatically? Like complex event processing?
Brittan: At this moment in time we dont think that Twitter sentiment is suitable for machine-based trading but this will likely evolve. The human filter and analysis is important to understand what is happening behind the scenes. On the machine side we do provide text analytics for long-form news and social media posts as there is more text to analyze sentiment on content that is validated by trusted news sources and agencies.
Were traders asking for this ability?
Brittan: Yes, our customers are really interested in finding out how they can harness this data. Social media and its influence on the markets continues to be an area of increasing interest. Historically, it has been difficult for human traders to keep pace due to the sheer volume and detail of data and the need to interpret it and spot trends immediately, with this new tool were delivering one way for traders to more easily sift through the noise on twitter.
There have been a few Twitter sentiment solutions and even hedge funds out there that have come and gone. How will Eikon avoid this fate?
Brittan: Eikon is our flagship financial markets desktop bringing together content and capabilities from across Thomson Reuters. The addition of this social media sentiment tool is another step for us in our mission to combine our unique content and insight with innovative visualization and analytics tools and its a really interesting area but its by no means the only area that we are turning our attention to. This app is designed to provide an additional level of analysis for our users. Acting as one component of their overall decision making process, the Social Media Monitorcan help our users make sense of the high volume of unstructured social content.
