Greenium or myth : do green labels affect bond yields in the Nordic markets?

Abstract

Estimates show that there is a significant need for more investments into sustainable projects to meet the climate goals set in the Paris Climate Agreement. Green bonds are a relatively new type of financial security developed to channelize capital into more environmental-friendly projects. “Green” financial securities, like green bonds, creates an opportunity for investors to contribute to the green shift. However, it is unclear whether investors sacrifice some of their returns when investing in environmental-friendly securities instead of conventional securities. The objective of our study is to investigate whether investors receive lower yields on their green bond investments, compared to what they would have earned on identical conventional bonds. More precise, our study examines if investors receive lower yields from labeled green bonds compared to what they earn on similar conventional bonds in the Nordic secondary markets. As previous research on this topic is limited, our study will contribute with new insight into the Nordic bond markets, which will be valuable for both investors and issuers. A matching method is used to examine whether investors receive lower yields on Nordic green bonds compared to conventional bonds. In this method, each green bond is matched with a similar conventional synthetic bond, which is composed of two conventional bonds. Further, a fixed effects regression is conducted to investigate whether there is a difference in the secondary market yield between the green bond and the matched synthetic bond, both for the whole sample and for various subsamples. A total of 77 Nordic green bonds are matched and analyzed in this research. Additionally, the estimated greeniums is regressed on different bond characteristics to capture potential determinants of the green bond premium. The findings show no statistically significant difference in the yield between green and conventional bonds when the full sample is analyzed. Hence, there cannot be stated that there is a greenium for the full sample. However, when the full sample is divided into subsamples, three of the subsamples have statistically significant greeniums. These subsamples are bonds issued in SEK, Investment grade bonds, and bonds with an issue amount between 251-500 million SEK. The greeniums found for the respective subsamples are 0.64 bp, 0.60 bp, and 1.2 bp.