investment, and new debt issues are estimated annually, cross-sectionally
over the entire period.
The estimated system equations following the Guerard, Bean, and An-
drews (1987) four-equation system, equations (6.2) through (6.5), are esti-
mated annually for the U.S. firms on the WRDS database and are shown in
Table 6.10 through Table 6.21. Here we find stronger evidence of the inter-
dependencies of financial decisions in the larger universe than in the origi-
nal Guerard, Bean, and Andrews (1987) study. In the estimated capital
expenditures investment equation, dividends are an alternative use of funds
(large negative and statistically significant coefficient), whereas investments
are positively associated with increasing research activities and net effective
debt financing at the 10 percent level of significance. Net income and de-
preciation have positive and statistically significant coefficients in the in-
vestment equation. The change in sales does not positively influence
investment, the accelerator position, as was the case in the study presented
in Chapter 4 using the WRDS database and traditionally issued long-term
debt. The change in sales is a positive and statistically significant determi-
nant of investment when the largest firms in the United States are analyzed.
In the estimated dividend equation, dividends are negatively associated
with capital expenditures and positively associated with R&D and net in-
come. We do not find a positive coefficient on the effective debt variable as
one would expect according to the imperfect markets hypothesis. New ef-
fective debt financing is significantly associated with higher capital expen-
ditures, dividends, and R&D variables; the larger coefficient is found on
the capital investment variable, as was the case in Guerard, Bean, and An-
drews (1987). R&D is associated with higher effective debt financing and
negatively associated with capital expenditures and dividends. There are
significant violations of the independence (perfect markets) hypothesis in
the capital expenditures, dividends, new debt issues, and research activities
equation estimations.
The statistically significant and positive coefficient on the external
funds issued variable is convincing in the investment equation and comple-
ments the work of McCabe (1979), Peterson and Benesh (1983), and Guer-
ard and McCabe (1992). Dhrymes and Kurz (1967) and Switzer (1984) did
not always find a significantly positive relationship between new debt and
investments. Mueller (1967) found an inverse relationship between invest-
ments and research in his earlier investigation and no relationship between
the variables in his later work with Grabowski (1972). Switzer found no
significant association between R&D and investment. Decreases in net liq-
uidity are associated with rising investment and dividends. Net income and
depreciation positively affect investment and negatively affect new debt fi-
nancing. Dividends are positively associated with rising net income and
backadmin
(backadmin)
#1