Title: A new Twitter based credit rating model methodology
Authors: Raffaella Calabrese - University of Edinburgh (United Kingdom)
Jonathan Crook - University of Edinburgh (United Kingdom)
Leonie Tabea Goldmann - University of Edinburgh (United Kingdom) [presenting]
Abstract: A model is proposed to predict corporate credit ratings using tweets about companies and tweets from the companies themselves. We make three contributions to knowledge. First, we relate tweets from the companies and tweets about the companies to corporate credit ratings. Second, we create different Twitter predictors and compare their performance. More specifically, we compare the performance of the tweet frequency, sentiment scores that have been calculated using different lexicon-based approaches and n-grams that have been created using differential language analysis. Third, we analyze the predictive power of tweets within the four largest industry sections and compare the differences. We analyze data relating to NASDAQ or NYSE listed companies over 2011-2019. We find including information from tweets gives a better predictive performance compared to models that omit them.