Assessing the investment skill of socially responsible investment (SRI) mutual fund managers can be difficult. One approach is to study the performance of SRI funds after accounting for restrictions that prevent the funds from owning stocks in SRI-prohibited industries, such as defense, alcohol, tobacco, and gambling. This method gives a clearer perspective on manager skill within the SRI fund universe.
The authors aim to determine whether socially responsible investment (SRI) fund managers add value after the impact of social investment constraints has been taken into consideration. They believe that accounting for specific SRI constraints is critical to accurately assess manager skill. The authors evaluate the performance of SRI funds both before and after accounting for restricted investments in alcohol, defense, gambling, and tobacco stocks. After adjusting for these restrictions, they gain a more favorable impression of SRI fund manager investment skill.
How Is This Research Useful to Practitioners?
Prior research in the area of SRI fund performance has focused on the performance of SRI investments relative to non-SRI investments by using standard asset pricing models. Research has estimated that approximately 12% of invested funds in the United States (and somewhat similar magnitude in Europe) follow some type of social responsibility screen(s), so the measurement of such investments is now of considerable interest. The authors’ approach differs because they specifically control for SRI investment restrictions when assessing manager skill. They argue that removing entire industries from the investment set creates systemic changes in the risk factors that are driving returns and reduces the reliability of traditional models to evaluate performance.
When using a standard Fama–French model, the authors find no evidence of skill for any of the SRI funds in the sample. But when they use an SRI-constrained Fama–French model, approximately 20% of the sample has significant alpha, with an average estimated outperformance of approximately 4% a year. It is noteworthy that the assessment of SRI fund performance is so materially affected by removing restricted “sin” industries through the modified model. The study seems to confirm prior research indicating that a performance cost exists when simply excluding traditional sin industries. For example, the authors identify tobacco as the industry that most affected factor returns, followed by defense.
This research will be of particular interest to those involved in the SRI industry and to investors in these strategies. The study will also be of some interest to the broader group of investment professionals engaged in manager performance evaluation and selection.
How Did the Authors Conduct This Research?
The authors evaluate the performance of a group of SRI funds over time. Of the various SRI screens used, this study focuses on product screens. Specifically, the authors examine funds that are restricted from investing in tobacco, alcohol, gambling, and defense stocks. Other screening categories (e.g., environmental) are ignored because the authors believe they are more difficult to measure and quantify than product screening.
They examine the performance of 66 US SRI equity mutual funds. The sample period from 1984 to 2006 is noted as one of the longest in SRI literature. Screening data are collected from the Forum for Sustainable and Responsible Investment, and the Global Industry Classification Standard (GICS) is used to filter prohibited industries. Monthly fund data, index data, and factor models are sourced from the CRSP/Compustat merged database.
To measure performance, the authors use a multifactor Fama–French asset pricing model. Stocks in prohibited industries are removed from the investment set. They then recreate the Fama–French portfolios to generate a new SRI Fama–French model. The adjusted model is used to run a regression of fund performance to determine investment alpha. This approach is superior because it allows the effects of imposed constraints and manager skill to be separated.
One complication arises with the different types of product screens, which can occur in three forms: negative, positive, and best-of-breed. The authors focus primarily on negative screens because they are the clearest. One drawback is that this approach reduces the sample size to 25 funds, whereas larger sample sizes would be preferred for this analysis.
The authors provide a good addition to the body of research on the performance of SRI strategies. The authors’ work creating an SRI version of the Fama–French asset pricing model provides a more accurate view of manager skill and is a good example of why appropriate benchmarking is so important. It would be interesting to determine, for funds experiencing significant alpha, whether any of the alpha can be attributed to the environmental, social, and governance screens not addressed in this study. It would also be interesting to know whether the same conclusions can be made when applied to non-US SRI funds.