A recent paper by two scholars affiliated with the the University of Delhi discusses the performance of socially responsible stock portfolios both during periods of boom and of bust within the Indian economy.
The scholars, Vanita Tripathi and Varun Bhandari, also helpfully reference two recent studies by Noel Amenc and Veronique Le Sourd that ask the same questions on the same issue, using a different nation-state as data base. The two pairs of scholars reach diametrically opposed conclusions.
Without further ado, I hereby admit my own confusion on the whole big subject of SRI. As a portfolio strategy, it shouldn’t work, although perhaps it does. If so, we are confronted with a bumblebee (an aeronautically impossible instance of observable flight). By the end of this post, I hope at least to make the nature of my confusion more clear, in the hope that our readers will comment in their usual helpful way.
Three Hypotheses
Tripathi and Bhandari look into three hypotheses: (1) that there is no significant difference in the returns of socially responsible stocks portfolio, general stocks portfolio and market portfolio; (2) that performance of socially responsible stocks is similar to that of the general stocks and the market stocks portfolio using various risk adjusted metrics; and (3) that there is no difference between the SRI portfolio and the general portfolio on the basis of net selectivity return. They believe their database falsifies each of these.
For consideration of the second of those hypotheses, they look at the modified Sharpe ratio, double Sharpe ratio, M2, the three factor Fama-French model, and Fama's decomposition measure. Fans of the classic movie Casablanca might call those "the usual suspects."
Their conclusion is that despite the higher risk, SRI portfolios generated "significantly higher returns and hence outperformed other portfolios on the basis of all risk-adjusted measures as well as net selectivity returns during both recession and boom periods." They used the period 1996-2013 as a database, and they note that it includes three boom periods (November 1997 to December 1998; February 2002 to October 2007; April 2009 to March 2011) and four busts (January 1996 to October 1997; January '00 to January '03; November '07 - March '09; April '11 to December '13)
The CNX 500 Equity Index, which (in 2015) represents about 95.77% of the free float market capitalization of the stocks listed on the NSE, served as a proxy for the market portfolio for this study.
The study’s results support the view that SRI is good for investors in India. This is important as a policy matter, given the mandates of the Indian Companies Act of 2013. The article concludes with the two scholars calling on "regulators, policy makers and mutual funds [who] should construct and make available various socially responsible investment products to initiate the movement of socially responsible investing in India."
But What About…
This is not likely to prove a definitive debate-ending paper. The literature review in the Tripathi-Bhandari paper itself shows that many other scholars have come to quite different conclusions from other databases. For example, Amenc & Sourd (2008) established that SRI funds in France did not produce positive and statistically significant alpha over the six year period that ended in December 2007, with reference to the Fama-French three-factor model.
Further, two years later Amenc & Sourd (2010) updated these results, considering performance amidst the global financial crisis, and again found that SRI funds provide no protection from market downturns during this period as the risks of these funds were quite high.
So, either something is askew in the empirical data or methods of such studies, or France is different from India. More generally, it is logically possible that some national economies, and their exchanges, are more favorable to the use of SRI strategies than others. But it seems more plausible that there is some evidentiary problem with one or another set of the conflicting studies on this question that has yet to be sorted out.
In general, though, the puzzles that pervade all arguments for the bottom-line claims of SRI are what they always have been. Whatever are the specific qualities that make some ‘good’ companies better for investors than others: aren’t those qualities observable? If they are observable, then why haven’t markets arbitraged away the value of investing there? If they aren’t observable, then how is any index of the SRI-worthy companies ever reliable?