The authors’ research on socially responsible investing (SRI) focuses on a defining characteristic of an increasingly popular investment strategy whose merits are nonetheless debatable given prior researchers’ mixed findings. Adoption of SRI by asset managers is accelerating globally. SRI is typically implemented by screening out controversial investments that fail to satisfy investor-designated noneconomic beliefs and values. Historical assessments are inconclusive about whether such self-imposed constraints reduce returns.

The authors strive to minimize such uncertainty with a methodologically novel investigation that expands on previously reviewed time periods, geographic regions, and criteria for applying SRI. The US SRI space has recently been estimated to include over 150 funds with approximately $150 billion of assets. Money managers whose clients either currently adhere to SRI standards or are considering doing so could find useful guidance for their decisions in this article.

The authors’ noteworthy findings include the following:

• Negative screening can substantially reduce the choice set to the extent of imposing a statistically significant (at the 1% level) negative alpha of –0.25%.
• This reduction is attributed to the opportunity cost of forgone investments in controversial firms, which varies with the exclusions imposed by specific screens.
• These excluded firms can be particularly beneficial in mitigating downside risk. During the Great Recession (December 2007–December 2012), controversial portfolios had an average economically significant relative annualized outperformance of 14.1%.
• Across all economic cycles, the greatest outperformance is more likely to cluster around firms in such industries as alcohol, gambling, and tobacco (the classic “triumvirate of sin,” or TS).
• Broadly defined, controversial stocks compose approximately 7% of world market capitalization and 12% of the S&P 500 Index, led by contraceptives (4.8%), animal testing (2.7%), and nuclear power (2.4%). Controversial stocks can be found in 94 countries, with only 42% in the United States, Australia, Japan, and Canada.
• Different investors may find different stocks acceptable or objectionable; for example, fur or meat may not have the same stigma as TS for some investors. The views of fiduciaries—including pension managers—may differ from those of their beneficiaries.

The authors’ precision may have been enhanced by their use of distinctive design features. Instead of relying on broad industry classifications to screen out entire industries as nonconforming with SRI, they conduct their analysis at the firm level. The authors combine industry classifications with keyword searches on 14 controversial issues, ranging from abortion and genetic engineering to nuclear power and weaponry.

Each company in the final sample is qualified manually by matching it with the study’s general, history, and activity descriptions. The final sample—culled from the ORBIS company database and combined with available return and market value data from Datastream—contains 1,634 stocks in 94 countries in both developed and emerging markets, fairly evenly balanced across North America, Asia, and Europe. The sample covers January 1991–December 2012.

They apply a comparative mean–variance analysis to the 14 controversial issues. Regressions are calculated via a Carhart (Journal of Finance 1997) model that compares risk-adjusted, value-weighted returns on the controversial stocks (for each controversial issue and for such combined clusters as TS) with those of the market and negatively screened portfolios. Because the data are non-normally distributed, a median regression using a least absolute deviation (LAD) estimator (a type of quantile regression, or QR) replaces the conventional OLS technique. Increasingly popular with financial researchers, QR methodology is more robust for fat-tailed and skewed data. The authors perform sensitivity analyses using OLS estimation as well as tests with a different market index, starting points for the index composition, and equal weighting. Only the last check shows contrasting results, with reductions in abnormal performance by controversial stocks.

The authors are persuasive in assessing the real tradeoffs associated with negative screening and SRI, contributing a clarity that has eluded previous investigators. Marshaling extensive evidence, they confirm the intuition that a reduction in economic choices, even for laudable ethical purposes, offers no free lunch and that “wages-of-sin” controversial stocks tempt investors with higher potential returns. In fact, higher expected returns may be necessary to entice investors to purchase these stocks.

What could be described as the opposite of the authors’ focus—buying stocks to actively intervene and improve corporate performance in SRI terms (e.g., impact investing and shareholder activism/engagement)—is missing from the study. Similarly, certain issues relevant to SRI (e.g., environmental sustainability, fair-labor practices, and human rights) are unexamined, with no explanation for their absence.

Reputational concerns that have applicability to SRI are also absent from the study. For a company with deficiencies on some SRI dimension, do changed practices affect perceptions over time? The authors note that they used no revenue/ownership threshold in classifying firms as controversial, apparently deeming any SRI failing significant. That approach affords scant margin for ethical error, fostering both a disincentive to adapt and a bias toward expanding the population of controversial firms.