Abstract
This research investigates the effect of Twitter sentiment on corporate bond returns. Previous work on this topic subjects that there is a connection between twitter sentiment and corporate bond returns. However, what is not confirmed is the magnitude of this relationship. Hence, my research attempts to contribute to this literature by applying a new dataset and estimation approach to estimate the correlation between aggregate twitter sentiment and corporate bond returns for the year period 2018. Using the prospect theory which states that people prefer more certain gains than the prospect of larger gains, I hypothesize that sentiments triggered by tweets can affect people's judgement of current versus future investment gains, hence there investment behavior. A negative sentiment about some corporate bond can lower the demand for that bond and consequently lead to price and yield drops for those bonds, whereas a positive sentiment can produce the opposite effect. I run linear regression models to test my hypothesis and I find that aggregate sentiment is not a statistically significant predictor of bond returns. The best model achieves nearly 30% R-squared value. I fail to reject the null hypothesis. The other predictors in the model: bond rating and maturity are more significant in affecting bond returns. Bond rating is the best performing variable in the model and is significant at the 1% level.
Advisor
Luri, Moses
Department
Business Economics; Statistical and Data Sciences
Recommended Citation
Jani, Fungai, "Investigating the Impact of Twitter Sentiment on Corporate Bond Returns" (2024). Senior Independent Study Theses. Paper 11107.
https://openworks.wooster.edu/independentstudy/11107
Disciplines
Business
Keywords
Twitter, sentiment, bonds, linear regression, sentiment analysis, NLP, market
Publication Date
2024
Degree Granted
Bachelor of Arts
Document Type
Senior Independent Study Thesis
© Copyright 2024 Fungai Jani
