ABSTRACT
My dissertation sheds light on the relationship between the shadow value of cap-
ital, the market value per capital stock, and investment. When the asset market is
e¢ cient,
rmsmarket value per capital stock, average q, represents the shadow value
of capital, marginal q or fundamental q.
Average q, however, is observed as di¤erent movements from fundamental q be-
cause of i) investment adjustment speci
cations and goods market structure, and ii)
investors speculative behavior. My
rst paper emphasizes the importance of ad-
justment costs and studies the interactions between average q, fundamental q, and
investment. The second paper explores the relationship through the structural vector
autoregressive model at the macro level. The paper allows market speculation and
constructs fundamental q with discounted pro
t rate along with average q.
The
rst paper evaluates the importance of investment installment costs in a
sticky price model by comparing two di¤erent adjustment cost speci
cations; one de-
pends on the investmentto-capital stock ratio, and the other depends on investment
growth. The two adjustment cost speci
cations are considered, since the former has
been adopted in the empirical literature such as such as Hayashi (1982) and the latter
has been adopted in the theoretical literature such as Chirinko and Fazzari (1994).
There is a stronger positive asset price (or average q) response to a positive technol-
ogy shock when the adjustment cost depends on investment growth. In addition, theinvestment growth speci
cation generates a hump shaped response of investment and
a semi-hump shaped response of output.
As indicated in Hayashi (1982), higher fundamental q leads to higher invest-
ment purchases. Higher shadow value of capital means that additional capital stock
creates net pro
ts, enabling
rms to increase investment purchases. An e¢ cient asset
market implies a close positive relation between average q and fundamental q, and
thus higher average q leads to higher investment purchases. Previous literature has
focused on average q and investment at the micro level with a single-equation regres-
sion model, and the result was not satisfactory. I have conducted empirical research
to answer whether investment is sensitive to fundamental q or average q through
comparison of impulse responses to a technology shock. In addition, the extent to
which technology shocks explain average q uctuations is studied through forecast
error variance decomposition.
My empirical paper has applied the structural vector autoregressive model
with the restriction that only technology shocks can alter labor productivity in the
long run. Impulse responses to technology shocks indicate that there exists a positive
interaction between investment and average q. On the other hand, fundamental q
is inuenced by investment but not in an adverse direction; fundamental q follows
investment growth rate. Furthermore, without having average q in the equation, fun-
damental q alone cannot be a signi
cant explanatory variable to predict investment.
Positive technology shocks are expected to raise
rmspro
ts, output, and invest-
ment. The variance decomposition results suggest that technology shocks account for
larger portions of output and investment when average q is used without fundamen-
tal q. When fundamental q is included in the estimation, the portion of investment
uctuations caused by technology shocks shrink signi
cantly, which con
rms that
fundamental q cannot explain investment uctuations.
i