KLD Research & Analytics, Inc. -- Socially Responsible Investing and Research

There's no doubt there is an inherent contradiction between conservatism and unfettered capitalism. Conservatives ought to be worshiping at a higher altar than the bottom line on a balance sheet. What in heaven's name is it that we conservatives want to conserve if not social stability and family unity?

The greatest area of controversy about screened portfolios centers on their performance. Proponents of a mission-based investment policy must address performance first generally and then specifically. In the context of the institution's asset allocation decisions, its manager searches, and its monitoring of its portfolio managers.

Any discussion of a change in investment strategy must include an examination of the relative financial performance of the vehicles under consideration. Mission-based investing is no exception to that rule. Indeed, socially responsible investing (SRI) has endured some bad press on performance, and proponents must anticipate questions based on it.

The following sections present brief responses to performance questions. More detailed ones will be found in the articles listed in the Resource Guide in the Appendix, below.

The Modern Portfolio Theory holds that, in general, diversification reduces risk and maximizes long-term returns. Therefore, anything that limits an investor's ability to diversify increases investment risks unnecessarily. To eliminate, say, tobacco company securities limits a manager's ability to diversify into an industry that may outperform the rest of the stock market.

This argument ignores the fact that one hires a manager because s/he is good at narrowing the universe of investable options. And, Modern Portfolio Theory is just that: a theory. Indeed, as discussed below, some studies indicate that particular screens do not impair portfolio performance. Citations to these and others will be found in the Resources appendix.

One benchmark, the Domini 400 Social Index (DSI), now exists for socially screened portfolios made up of U.S. equities. This 400 company index modeled on the Standard & Poor's 500 (S&P 500) applies social screens across nine areas. Launched on May 1, 1990, it has outperformed the Standard & Poor's 500 from its inception through the end of April 1998.

One should not draw conclusions about future performance based on past performance. In addition, the Domini 400 numbers do not represent data from a full market cycle -- from bull to bear to bull. The data only cover a part of the post-1982 bull market.

Experience with portfolios only dates to 1970, the year a group of Methodist clergy and laity launched the Pax World Fund, a balanced fund. The first pure equity fund to feature multiple screens was the Dreyfus Third Century Fund, launched in 1975.

Antedating Pax World and Dreyfus Third Century were some mutual funds that eliminated "sin stocks." The Pioneer Fund has excluded tobacco and alcohol companies for 50 years. An illustration supplied by the Fund shows that $1,000 invested, when the fund was launched on March 1, 1928, would have been worth $5,780,117 on December 31, 1997. A similar investment in the S&P 500 would have yielded $1,190,147. The Fund's average annual return for the period was 13.2 percent, while the S&P 500's was 10.67 percent.^*

An interesting study by the WM Company, a leading U.K. consultant to charities and public authorities, compared the performance of constrained (screened) funds to that of its total U.K. Charity Fund Universe. The constrained universe consisted of 20 funds with a market value of 822 million Pounds at the end of 1995. The constraints were, by U.S. standards, quite limited -- alcohol, tobacco, and armaments. However, constrained funds comprised of U.K. equities performed identically to unconstrained funds over the four years ending with 1995; constrained "overseas equities" underperformed unconstrained by 0.1 percent. WM concluded:

The policy of excluding certain ethically unacceptable sectors/stocks from portfolios appears to have had little impact over the period 1992-95. The performance of Charity Funds following such an exclusion appears little different, in aggregate, from that of the Unconstrained Charity Universe. Other constraints, in the form of income requirements or asset mix, are more important in performance terms.^*

	DSI 400	S&P 500	S&P MidCap
12 months ending 8/31/98	11.42%	8.11%	-9.45%
24 months ending 8/31/98	26.23%	23.29%	11.44%
36 months ending 8/31/98	23.75%	21.76%	11.58%
60 months ending 8/31/98	19.33%	18.23%	11.85%

Are there studies on screening's effect on the performance of U.S. equities?

Yes. Some studies have indicated that investors applying social criteria need not expect to lose anything vis a vis the broad market indexes. Appendix D, Resources, provides starting points for a literature review. The citation for the articles discussed below will also be found there.

In 1993, S. Hamilton, et al., reported in Financial Analysts Journal on their studies of socially screened mutual funds. They concluded: "Investors can expect to lose nothing by investing in socially responsible mutual funds [when compared to a benchmark of randomly selected mutual funds]."

In 1995, M.A. Cohen, et al., of Vanderbilt University reported that between 1987 and 1989 Green investors, who typically have quite stringent screening criteria, did not seem "to pay a premium for their convictions" as compared to the S&P 500.

In a 1997 study in The Journal of Forecasting, John B. Guerard, Jr., then Director of Quantitative Research for Vantage Global Advisers, reached a similar conclusion. Guerard compared the performance of Vantage's 1,300 company universe against a subset of 950 that passed four major screens: military; nuclear power; product exclusion (alcohol, tobacco, and gambling); and environment. Mr. Guerard found:

The Vantage Global Advisers' unscreened . . . universe produced a 1.068 percent monthly average return during the January 1987-December 1994 period, such that a $1.00 investment grew to $2.77. A corresponding investment in the socially-screened universe would have grown to $2.74, representing a 1.057 percent average monthly return. There is no statistically significant difference in the respective return series. More importantly, there is no economically meaningful difference between the return differential. (Emphasis added.)^*

Articles collected in the Winter 1997 Journal of Investing, including one by Mr. Guerard, confirm his conclusion. In "Additional Evidence on the Cost of Being Socially Responsible," Mr. Guerard concluded that the use of a broad range of social screens produces a more efficient portfolio than one that is only divested of tobacco stocks. He found that the use of environmental, product (alcohol, tobacco and gambling), military, and nuclear power screens produced portfolios with higher excess returns than unscreened portfolios. Of these screens in isolation, only the military screen significantly diminished returns between 1992 and 1997.

Yes.^** In April 1998, Stephen Williams of the WM Company submitted a masters thesis to the University of Edinburgh illustrating various types of "ethical indexes" modeled on existing Financial Times broad market gages. These studies indicate performance similar to the Domini 400 against the Standard & Poor's 500.

Frank J. Travers of Oppenheimer & Co. studied 23 screened portfolios which he described as non-U.S. His Winter 1997 article in The Journal of Investing, "Socially Responsible Investing on a Global Basis: Mixing Money and Morality Outside the U.S.," concludes that over the periods he studied, these portfolios outperformed Morgan Stanley's EAFE Index and produced competitive returns to a universe of unrestricted portfolios.

To date, the only study on screened bond portfolios is one by Louis D'Antonio, Tommi Johnsen, and Bruce Hutton of the University of Denver. Their article, "Expanding Socially Screened Portfolios: An Attribution Analysis of Bond Performance" appeared in the Winter 1997 issue of The Journal of Investing.

They posed the question: could one develop a bond index with screens identical to the Domini 400's and, if so, how would it perform? They concluded that an analogue to the Lehman Corporate Bond Index could be constructed based on the Domini 400's constituent companies. As to performance, they concluded:

A portfolio of bonds comprising issues from firms represented in the Domini 400 performs comparably in terms of risk-return to the Lehman Brothers Corporate Bond Index. The SRI and LCB indexes are identical in the exposures to term structures. There is no significant difference between the portfolios . . ..

While the active returns for the SRI portfolio indicate that it outperformed the LCB index on a simple return basis, the premium appears to be related to differences in credit risk. In general, our research indicates that . . . there should be no penalty [to socially responsible investors] for following their social values after adjusting for credit risk.^*

The performance of the Domini 400 and the results of the studies discussed above shift the burden of proof to critics to show that there are indeed "costs" to social screening. Thus far there appears to be little evidence that social screening necessarily results in lost return. The affirmative case for screened portfolios has not yet been proven. Time and a full market cycle will determine the "cost" question. But, the ultimate questions remain unresolved:

Late in 1992, BARRA published an analysis of the performance of the Domini 400 Social Index versus that of the S&P 500. Its conclusions, in part, were that:

[The] bulk of the outperformance [of the DSI versus the S&P] arises not from the factor biases inherent in the introduction of social screens, but from the specific asset return itself. There is a specific return premium to the index over this time period which is presumably related to the social screens KLD has developed, though the 95% confidence level is not quite statistically significant.^* [Emphasis added.]

Whether this and similar preliminary analyses will hold true remains to be seen.