composite model exceed the t-statistics of its components. The purpose of
a composite security valuation model is to identify the determinants of se-
curity returns and produce a statistically significant out-of-sample ranking
metric of total returns.
An indication of the relative importance of the eight fundamental
variables and the composite earnings forecasting variables is given by the
time average value of the regression coefficients estimated for each year in
our 1990–2003 study period. They support the low P/E (high earnings
yield) approach to value investing advocated by Graham and Dodd
(1934) and Graham, Dodd, and Cottle (1962) and validated as a cross-
sectional return anomaly by Basu (1977). They also support the Fama and
French (1992, 1995) finding that the book-to-market ratio is an important
variable for explaining the cross section of security returns. However,
while both these variables are significant in explaining returns, the major-
ity of the forecast performance is attributable to other model variables,
namely the relative earnings-to-price, relative cash-to-price, relative sales-
to-price, and earnings forecast variables. The most statistically significant
variable in identifying security returns is the composite earnings forecast
variable. One should use regression modeling of monthly holding period
returns (HPRs) to identify factors influencing returns at particular points
in time.
Further Estimations of a Composite Equity Valuation Model 217
TABLE 8.4 Information Coefficients of the Composite
Security Valuation ModelTechnique Universe IC (t)EP Russell 3000 0.036 (24.33)
BP Russell 3000 0.025 (16.75)
CTEF Russell 3000 0.035 (23.78)
EVL Russell 3000 0.022 (15.00)
EQ9 Russell 3000 0.031 (20.88)
WLRR Russell 3000 0.045 (30.21)EP Japan-Only PACAP 0.046 (21.87)
BP Japan-Only PACAP 0.042 (19.94)
CTEF Japan-Only PACAP 0.020 (8.64)
EVL Japan-Only PACAP 0.042 (19.95)
EQ9 Japan-Only PACAP 0.043 (20.49)
WLRR Japan-Only PACAP 0.053 (24.98)