Green bonds have become a new financing instrument enabling the funding of projects with environmental benefits. Although the green bond market represents only 1.5% of the fixed-income market, there was a rising interest from both institutional and corporate investors and increased issuance of green bonds in the past five years. There is an ongoing debate in the literature about the existence of the green bond premium - the difference investor is willing to pay for the green feature of the bond. The first part of the thesis follows Zerbib’s (2019) methodology for constructing green bond premium and further analyses its relationship with bond-specific characteristics. In the second part of the thesis, we expand on Zerbib’s methodology by using machine learning techniques. We do this by using greenwashing allegations data on European companies issuing green bonds to investigate its impact on green bond yield. Since the green bonds can be seen as a ”green promise” a company makes, the greenwashing concern can be seen as a broken covenant. The results of the Random Forest Regression show that greenwashing news is currently not priced in, and the most important variables influencing the green bond yields are the issue size, maturity and currency.     «

Green bonds have become a new financing instrument enabling the funding of projects with environmental benefits. Although the green bond market represents only 1.5% of the fixed-income market, there was a rising interest from both institutional and corporate investors and increased issuance of green bonds in the past five years. There is an ongoing debate in the literature about the existence of the green bond premium - the difference investor is willing to pay for the green feature of the bond....     »