ESG-scores som Investeringsparameter: Anvendelsen af ESG-scores og sektor i investeringsstrategier på det amerikanske aktiemarked i perioden 2011-2020

Abstrakt

This academic paper is investigating the effects of ESG-scores, sector and ESG-provider in factor investments and how the evolvement of ESG as a concept has affected its validity. The term social responsibility has been around for ages and has been discussed since the creation of the bible. The focus on ESG is however fairly new within investments, which means the current academic literature is still evolving. Research has concluded that there are arguments for and against using ESG as an investment criterion. Some believes that ESG can be used to adjust the risk on a portfolio, while other argues that high rated ESG-stocks are overvalued. This paper will try to contribute to covering this gap in the literature, by reviewing existing literature and deriving what is the best practice when using ESG as an investment criterion and by devolving its own investment strategy. The first analysis of the portfolios constructs high and low portfolios using only ESG, E, S and Gscores and indicates that it is not possible to generate a significant alpha by investing in stocks with e.g., the 10% highest and lowest scores. Furthermore, it indicates that by creating a long/short strategy between the two portfolios, in three out of four cases it generates a beta neutral portfolio, but not a positive alpha, hence it is not possible to use it as a risk factor in the Fama French multifactormodel. The second part of the analysis confirms what the literature review concluded - that sector has an effect on the generated alpha in the portfolios. The third part of the analysis concludes that when using sector in combination with ESG-scores it is possible to optimize the alpha and overall performance on all portfolios except ESG(L). The fourth part of the analysis indicates that there are no statistical differences between the portfolios using different ESG-data providers. They are however not completely the same but show the same overall trends. The overall conclusion of this paper is split in three, i) L-portfolios are outperforming H-portfolios on both alpha and geometric return. ii) H-portfolios have a lower standard deviation resulting in a higher Sharp Ratio; hence ESG-scores can be used to modify the risk. iii) There are no statistically significant differences between the two ESG-data providers.