Socially Responsible Investments: Performance analysis between a SRI and a conventional index

In 2018, the total Socially Responsible Investments SRI assets under management (AUM) amounted $30,7 trillion. In the world, the top 400 asset management companies represent approximately $65,7 trillion. This means that the SRI AUM counts for approximately half of the total assets under management in the world. Due to the huge increase of this investment philosophy, the following question about the performance of those investments could be raised: Does Socially Responsible Investments show a significant difference in performance compared to conventional investments? In order to answer this research question, the following study has been conducted. The FTSE4Good index family was compared to its correspondent benchmark, indices of the FTSE index family. Four geographic zones were selected, namely the US, the UK, the EU zone and the World as a global zone. In order to measure the comparative performance, the Sharpe ratio, a linear regression of the CAPM and the Fama and French three-factor model have been be used. Using the Sharpe ratio, all the results were negative (for both SRI indices and the benchmarks). This means that the risk-adjusted excess returns were negative. But the SRI Sharpe ratios presented in the four regions higher values than the one of the benchmark indices. The CAPM regression presented a positive alpha for US and the world, and a negative one for the UK and EU. However, the four alpha values, that represent the outperformance of the investment compared to its benchmark, were not statistically significant. For all regions except UK, the betas had values close to one, which means an almost similar level of risk for both indices. The Fama and French three-factor model presented a SMB value negative and close to zero and a HML value close to zero, except for the World. As a conclusion, it can be said that SRI has the potential to present outperformance to conventional investments, although the results obtained in our analysis are not particularly significant.