Current news does affect stock prices and is something that investors always keep an eye on especially when seeking investment. This will enable them to find out whether certain stocks will be affected in some way or another and the degree to which certain news may have an impact. Any form of sentiment triggered by the general public towards a particular news item / the current state of affairs, being a scandal, natural disaster, introduction of a new product, decline in sales numbers, or positive financial results of a company, usually affects stock prices.
Other forms of news that usually impacts stock markets are: crude oil prices, inflation, unemployment, government policies, political unrest, bonus dividends and stock buy backs, inclusion or exclusion from indexes, change or death of top officials, changes in demand and supply and terrorist attacks amongst other factors. It has been proven that stock prices tend to react to negative news quicker than for positive news. This shows that news have a direct impact on the market [3].
A very good example is the latest Volkswagen emissions scandal [1] who have seen their stock price go down around 30% in five days from when the Environmental Protection Agency publicly announced that the automaker manipulated software to hide the emissions its cars produce [2] on 18th September 2015. Over a month from this scandal, the Volkswagen AG stock price is still at $109.30 (October 30th) from the $167.80 price that it held on 17thSeptember i.e. 34.9% decrease, before this scandal was announced.
Does social media have a role to play in all of this?
Well …, social media platforms help in spreading the word about current state of affairs or important news items. When news is spread on such platforms by a certain amount of users, this creates what is called a “trend” in Twitter. A trend on Twitter refers to a hashtag-driven topic that is immediately popular at a particular time. A hashtag is a keyword or phrase that is preceded with a hash (#) sign, for example with #ChampionsLeague or #UKElections [4]. Usually, a certain amount of tweets about a particular topic are made in order for a topic to trend and make it into one of Twitter’s top ten trending topics. The number of unique users tweeting and the amount of retweets of a particular message is taken into consideration when classifying trending topics, which are refreshed periodically in Twitter [5].
Given the number of posts that are made every day – currently around 500 million [6] – Twitter contains a lot of relevant data that contains several patterns hidden in it. Through the discovery of such patterns i.e. the relationship between what people are tweeting regarding publicly traded companies and what’s happening with the companies’ shares, we are able to predict trends in stock prices [7]. Work of this sort has been on-going for at least the past four years, where the sentiment of tweets is analysed in order to try and beat the stock market. In 2011, London-based firm Derwent Capital Markets were analysing emotions expresses across 10% of approximately 100 million daily tweets by using certain algorithms devised by Johan Bollen, in order to predict changes in the stock market. This led to Derwent’s 25 million fund in finishing its first month of trading in July with a return of 1.85% [8]. Currently, several companies are partnering with Twitter in order to further invest in Big Data Sentiment Analysis around stocks, such as Bloomberg (September 16, 2015) [9]. A recent study by academics at Johns Hopkins University has found out that crowd-sourced company earnings estimates and sentiment data generated by tweets may be both more accurate than Wall Street’s offerings and help generate trading profits [10].
In SSIX we are building on top of the existing state-of-the-art research and development sentiment analysis techniques in order to extract the relevant and valuable economic signals in a cross-lingual fashion for several social media platforms, such as Twitter, StockTwits, LinkedIn, and public media outlets, such as Bloomberg, Financial Times and CNBC. This will help in generating custom X-Scores powered indices for a given sentiment target or aspect i.e. company or financial product.
A paper about “Analysis of Cyberbullying Tweets in Trending World Events” written by Keith Cortis and Siegfried Handschuh from the University of Passau, has recently been presented at the 15th International Conference on Knowledge Technologies and Data-Driven Business (i-KNOW 2015). This conference aims at advancing research at the intersection of disciplines such as Knowledge Discovery, Semantics, Information Visualization, Visual Analytics, Social (Semantic) and Ubiquitous Computing. Even though, this research was focused on tackling the problem of cyberbullying via a novel approach that analyses online posts in trending world events, a similar approach and concept will be adopted for stock markets in SSIX. In this case stock market tweets that take place during certain trending world events (which will probably result in a trending topic on Twitter) will be further analysed to see if they can be of any relevance for the SSIX indexes being generated.
This blog post was written by the SSIX partners at the University of Passau.
For more information on SSIX, visit our website ssix-project.eu.
For the latest update, like us on Facebook, follow us on Twitter and join us onLinkedIn.
References:
[1]: http://www.bbc.com/news/business-34324772
[2]: http://fortune.com/2015/09/23/volkswagen-stock-drop/
[3]: http://www.sharemarketschool.com/how-does-news-affect-stock-prices/
[4]: https://www.hashtags.org/platforms/twitter/what-do-twitter-trends-mean/
[5]: https://totalsocialmedia.wordpress.com/2009/07/26/trending-topic-on-twitter-how-many-tweets-to-reach-the-top/
[6]: http://www.internetlivestats.com/twitter-statistics/
[7]: http://www.forbes.com/sites/jeffbercovici/2012/03/22/twitter-chatter-can-predict-stock-prices-study-finds/
[8]: https://www.newscientist.com/article/mg21128295-900-using-twitter-to-follow-trends-beats-the-stock-market/
[9]: http://www.bloomberg.com/company/announcements/bloomberg-and-twitter-sign-data-licensing-agreement/
[10]: http://www.reuters.com/article/2015/09/02/us-market-saft-idUSKCN0R229E20150902