Vivi Wong: Statistical Aspects of Sustainability in Optimal Portfolio Theory
Time: Thu 2020-06-11 09.00
Location: Zoom, meeting ID: 901 405 086
Participating: Vivi Wong
Subject area: Mathematical Statistics
Supervisor: Taras Bodnar
The world strives to satisfy sustainability conditions, such as human rights and environmental protection, among others, nowadays. Hence, investors would find it interesting to know whether investing in ethical portfolios will worsen the investment opportunities, compared to investing in portfolios constructed by both sustainable and unsustainable assets. One can choose to analyze this topic based on classical optimal portfolio analysis, such as studying whether the difference between the efficient frontier of sustainable assets and the efficient frontier of those assets and some additional unsustainable assets is statistically significant by applying mean-variance spanning tests. In this thesis, we used monthly and weekly returns, respectively, of 21 stocks in the OMX Stockholm 30 (OMXS30) index over the period 2008-2019, and we performed screenings to obtain stocks that are considered more sustainable. Mean-variance spanning tests were then applied to study whether the differences between the efficient frontier of all the 21 stocks and each of the efficient frontier of the assets obtained after screening are statistically significant or not. The results of the spanning tests showed statistically nonsignificant differences between the efficient frontiers. Hence, our study suggests that an investor would not obtain better investment opportunities when one in addition to the considered more sustainable assets, in the OMXS30 index, adds the stocks that are considered more unsustainable into the investment portfolio.