Innovation, R&D, and Stockholder Wealth
It is well known that there is substantial underinvestment in R&D in the
United States economy, primarily because social benefits exceed private
rates of return from innovative activities (Mansfield, Rappaport, Romeo,
Wagner, and Beardsley 1977). Furthermore, underinvestment in R&D di-
minishes potential stockholder wealth because R&D activities have been
associated with the market value of major industrial firms (Ben-Zion 1984;
Guerard, Bean, and Stone 1990; and Guerard and Mark (2003). It is help-
ful to model the R&D budgetary process of major industrial companies,
since empirical evidence supports the hypothesis that decisions on R&D
expenditures are made simultaneously with a firm’s other financial deci-
sions (Switzer 1984; Guerard and McCabe 1992), as discussed in Chapter
- A firm’s decision to increase its R&D expenditures impacts its decisions
on dividends, investments, and new debt issuance. Thus, there is a need to
understand how corporate decisions on expenditure reallocation may af-
fect a firm’s stock price.
As discussed earlier, in Chapter 6, the empirical research on the per-
fect markets hypothesis has yielded mixed results since its formulation
by Miller and Modigliani (1961), although the majority of studies pub-
lished in the 1970s support the existence of imperfect markets. A review
of this research led Guerard, Bean, and Stone (1990) to formulate a re-
vised model of the stock price valuation process that allows interdepen-
dencies that were precluded by the perfect markets hypothesis. One of
the goals of this study was to revive the Mueller (1967) hypothesis that
research and development expenditures should be included in tests of
the perfect markets hypothesis. Guerard, Bean, and Stone (1990) used
data obtained from Compustat and U.S. Patent Office tapes to estimate
such a model.
In July 1984, the National Science Foundation and the U.S. Bureau
of the Census announced the availability of a longitudinal database
that would enable researchers to explore the relationship between R&D
and other economic variables on an enterprise basis. This chapter com-
pares the results obtained from the NSF data with those derived from
the Compustat/Patent Office data. The model used in the Guerard and
McCabe (1992) study assumes interdependence between several finan-
cial decisions. It employs investment, dividends, and new capital financ-
ing equations to describe the firm’s budget constraint. The manager
may use available funds for research and development (R&D), capital
investment (CE), or dividends (DIV), or to increase net working capi-
tal (LIQ). The sources of funds are represented by net income (NI), de-