| Popis: |
In our age, where a customer-oriented approach is dominant, financial enterprises have started to focus on social media platforms in order to create an active communication network with their investors. In this study, encompassing two different time periods, Pearson correlation analysis and Granger causality analysis are conducted in order to identify the relation between the investors' varying sentimental polarity values in their posts they share on Twitter, which is an effective social media platform, and daily changes in Bist30 index returns. As a result, significant correlations are found between the polarity values of tweets and the daily changes in Bist30 index returns. In addition to this result, especially in the period of unusual events, it is seen that there is a Granger causality from the polarity values of tweets towards the daily changes in Bist30 index returns. This study that works with the financial Turkish tweets and proposes a new method while conducting Doc2Vec algorithm at the level of sentiment analysis, is one of the most comprehensive studies since it not only contains correlation analysis and Granger causality analysis steps, but also applies all the statistical and econometric tests, which are necessary to carry out analysis. |
