A recent Wharton School study that finds underperformance by socially responsible investment mutual funds is based on an unsound research design, according to critics.
A recent study from the Wharton Schooll of the University of Pennsylvania purports that portfolios containing socially responsible investment (SRI) mutual funds underperform portfolios representing a broader fund universe. The paper, entitled Investing in Socially Responsible Mutual Funds, considers fund managers’ stock-picking skills, and employs such sophisticated analytical techniques as the Sharpe ratio and the Capital Asset Pricing Model (CAPM). Critics of the study point to the study’s potential design flaws, as well as to other current studies being added to the already-large body of empirical studies that demonstrate the competitive performance of SRI funds.
"This study makes several elementary mistakes," said Richard Torgerson of the Financial West Group and a registered principal in the Progressive Asset Management (PAM) network. "First and foremost, the universe selected to compare SRI funds against actually undercuts their entire premise."
The study, authored by Wharton professors Christopher C. Geczy and Robert F. Stambaugh and graduate student David Levin, uses 46 different non-SRI mutual funds from a universe of 894 equity mutual funds. The authors then constructed 36 different portfolios of varying allocations to reflect various decision-making models.
"Fully 50 percent of the 46 non-SRI funds making up the portfolios compared against SRI portfolios are completely unavailable to the average mutual fund investor, either by having investment minimums above $100,000 or by being closed to new investors at the time the study was done," said Mr. Torgerson. "None of the 36 portfolios constructed by Geczy, Stambaugh and Levin were constructed with any less than 44 percent of the portfolio consisting of unavailable funds."
"Five of the 36 portfolios were made up 100 percent of unavailable funds," Mr. Torgerson added. "So in effect, the Wharton study urged investors to eschew SRI funds in favor of other funds that they can’t invest in."
None of the three Wharton authors responded to SocialFunds.com’s requests for their commentary.
Interestingly, Innovest Strategic Value Advisors criticized the study on a related but distinct aspect.
"At one point, the Wharton study compares the performance of broadly-based SRI funds with a group of 28 mainstream equity funds, of which fully 17 are real estate funds!" said Bijan Foroodian, director of quantitative analysis at Innovest. "Given the enormous differentials in the risk/return characteristics of these two different types of investments, this comparison is, at best, disingenuous."
Mr. Foroodian authored a recent Innovest report that examined SRI simulations of a U.S. public pension fund. Five out of six simulations outperformed the actual portfolios throughout 2002 by an average of 100 basis points, or one percent. The simulations were different only in that they were tilted according to Innovest’s EcoValue’21 ratings, which identify environmental best practice. The sole underperforming simulation did so only by .04 percent.
The pension fund that served as the model employs the EcoValue’21 overlay on an actual $150 million portfolio. That portfolio outperformed its S&P 500 benchmark by 1.5 percent in the year since its inception in March 2002.
Recent performance also may undermine the study’s own conclusions regarding its comparisons. For example, one scenario compares an SRI portfolio comprised of the California Mid Cap Index (ticker: SPMIX), the Citizens Core Growth (WAIDX), and the Domini Social Equity (DSEFX) funds to a non-SRI portfolio totally allocated to the Vanguard Total Market Index Institutional (VITNX) fund. The study finds that the latter portfolio outperformed the former by five basis points per month up through December 31, 2001.
After pointing out the apples-to-oranges comparison of an institutional fund to three retail funds, Mr. Torgerson discussed his analysis of the portfolios’ performance from the end of 2001 through June 30, 2003.
"The portfolio that was supposedly given a 5 bps advantage by Sharpe ratio measures actually underperformed the SRI portfolio by 3.16 percent," Mr. Torgerson stated. "Their own portfolio selection this instance shows about a 16 bps monthly advantage to the SRI portfolio instead of the 5 bps penalty predicted by the December 31, 2001 Sharpe ratio."