Hamilton, Jo, and Statman (1993). The relative monthly outperformance of 7
basis points was not statistically different from zero. It is not obvious what cri-
teria were used to determine the socially responsible universe in the Hamilton,
Jo, and Statman study. Recent studies by Diltz (1995a,b) found no statistically
significant difference in returns for 28 stock portfolios generated from a uni-
verse of 159 securities during the 1989–1991 period. Diltz found that only the
environmental and military business screens were statistically significant at the
5 percent level during the 1989–1991 period.
- J. Rothchild, “Why I Invest with Sinners,” Fortune(May 1996).
- Morningstar, Principia for Mutual Funds, March 31, 1996.
- The CREF Social Choice Account was a $1.174 billion account as of Decem-
ber 31, 1995, consisting of 61.49 percent socially screened equities, 37.67 per-
cent bonds, and 1.72 percent short-term commercial paper. The CREF Social
Choice Account uses screens for environmental, weapons, nuclear power, alco-
hol, tobacco, and gambling products, as well as MacBride Principles (a code of
fair employment by U.S. firms in Northern Ireland to prevent religious discrim-
ination). The CREF Social Choice Account has matched its performance
benchmark for the past one- and five-year periods ending March 31, 1996,
producing total returns of 23.56 percent and 12.70 percent versus its bench-
marks of 24.00 percent and 12.32 percent, respectively. The recent CREF So-
cial Choice equity component is important because CREF underperformed in
its unscreened equity fund during the past one year. The CREF Bond Market
account has been a market performer in its bond investment for the one- and
five-year periods ending March 31, 1996, producing bond returns of 10.52
percent and 8.51 percent versus the Lehman Aggregate Bond Yield index of
10.79 percent and 8.49 percent, respectively. The CREF Stock Account earned
one- and five-year returns of 28.81 percent and 13.55 percent versus the S&P
returns of 32.10 percent and 14.66 percent, respectively. The CREF Social
Choice Account has produced total returns consistent with its balanced perfor-
mance benchmark and has not substantially underperformed on its equity
component. The reader is referred to the College Retirement Equities Fund
prospectus, “Individual, Group, and Tax-Deferred Variable Annuities,” April
1, 1995, for a description of the CREF Social Choice Account.
- Fama and French (1995) actually tested whether higher book-to-price stocks
outperformed the lower book-to-price stocks. It can be confusing when one
thinks of the low P/E approach of Graham, Dodd, and Cottle (1962) in which
an investor purchases low price-to-earnings stocks (i.e., one should not pur-
chase a stock that has a price-earnings multiple exceeding 1.5 times the aver-
age price-earnings multiple of the market) and the higher earnings yield, or
earnings-to-price (EP), approach tested in the academic literature. The two
earnings formulations yield roughly the same result when applied to low-P/E
or high-EP decisions; see Guerard and Takano (1992). Wall Street persons tra-
ditionally think of the low-P/E and low-PB models, whereas academicians pre-
fer the conventional EP and BP models because the conventional formulations
are not plagued by small negative and positive denominators, such as with very
268 NOTES
