Is news sentiment well suited to predict fluctuations in U.S corporate green bond returns? A comparative study of news and market-based sentiment indicators.

Abstract

This thesis explores the relationship between news-based and market-based investor sentiment and the returns on corporate green bonds, focusing specifically on the S&P 500 firms. The study tests two hypotheses: (1) Market and news-based sentiment indicators positively predict fluctuations in corporate green bond returns; and (2) Market-based sentiment more accurately predicts these fluctuations compared to news-based sentiment. Using the S&P 500 Green Bond Index as the benchmark, three sentiment indicators are evaluated for their effectiveness in forecasting green bond returns. Despite mainstream literature indicating otherwise, the employed Ordinary Least Squares (OLS) regression models indicate a statistically insignificant relationship between green bond returns and all of the sentiment indicators.Newsbased sentiment indicators provided a marginally more consistent and stable prediction over the studied period from 2014 to 2024. This research contributes to the temporal understanding of how investor sentiment impacts green bond returns, among else highlighting the importance of robust and representative data.