Profitability of socially responsible investing – Performance of SRI equity indices during market downturn

Abstract

Socially responsible investing (SRI) has transformed from minor niche to mainstream investing philosophy in the past decade. Threat of climate change, growth of ethical consumerism and growing awareness of corporate social responsibility have shifted investors to demand their assets to be managed in a more sustainable way. This has lead to drastical increase in the assets managed by SRI strategies, alone in the US 8,72 trillion US dollars are managed by SRI strategies.

As amounts of assets managed by SRI strategies have exploded, the academia have not been able to achieve consensus of whether implementing non-financial environmental, social and governance (ESG) information into financial analysis is a value increasing factor. As most of the studies find SRI to have no disadvantages against conventional investments, there is also plenty of evidence of superior performance of both SRI and conventional investments.

The purpose of this study is to examine the performance of SRI equity indices against conventional indices in the US and Europe from 2003 to 2016. As previous literature has found SRI to outperform conventional investments during times of market crises, the sample is split into three sub-periods to analyse whether SRI indices outperformed conventional indices during financial crisis in 2007-2009. Performance is evaluated by Sharpe and Treynor ratios, CAPM, Fama-French 3-factor model and Fama-French 5-factor models in contrast to benchmark indices.

Statistically significant results from the regression analysis showed that in most of the cases SRI indices were outperformed by their benchmarks. Slight evidence of the outperformance of SRI indices during financial crisis was found from the US markets, but the European SRI indices underperformed also during financial crisis. Surprisingly evidence of the overperformance of European SRI indices in the post-crisis period was also found. However, the split sample results were not statistically significant and thus should be considered with caution.