Document Type

Conference Paper


Available under a Creative Commons Attribution Non-Commercial Share Alike 4.0 International Licence


Computer Sciences

Publication Details

In Proceedings of WI '19: IEEE/WIC/ACM International Conference on Web Intelligence


In June 2016, the British voted by 52 per cent to leave the EU, a club the UK joined in 1973. This paper examines Twitter public and political party discourse surrounding the BREXIT withdrawal agreement. In particular, we focus on tweets from four different BREXIT exit strategies known as “Norway”, “Article 50”, the “Backstop” and “No Deal” and their effect on the pound and FTSE 100 index from the period of December 10th 2018 to February 24th 2019. Our approach focuses on using a Naive Bayes classification algorithm to assess political party and public Twitter sentiment. A Granger causality analysis is then introduced to investigate the hypothesis that BREXIT public sentiment, as measured by the twitter sentiment time series, is indicative of changes in the GBP/EUR Fx and FTSE 100 Index. Our results from the Twitter public sentiment indicate that the accuracy of the “Article 50” scenario had the single biggest effect on short run dynamics on the FTSE 100 index, additionally the “Norway” BREXIT strategy has a marginal effect on the FTSE 100 index whilst there was no significant causation to the GBP/EUR Fx. The BREXIT Political party sentiment for the “No Deal” was indicative of short term dynamics on the GBP/EUR Fx at a marginal rate. Our test concluded that there was no causality on the FTSE 100.