Within the context of a stochastic growth economy, the shocks to techn
ology are modelled as a four-state Markov process. The parameters of t
his process are chosen so that the implied conditional distributions f
or the marginal product of capital can be ordered in terms of first- a
nd second-order stochastic dominance. This characterization of time-va
rying uncertainty is then used to analyse numerically the implications
for bond and equity prices. Our primary finding is that the response
of asset price to an increase in technological uncertainty may differ
substantially in this production economy relative to that observed in
an exchange setting (as recently studied by Barsky 1989 and Abel 1988)
. This difference is due to the endogenous behaviour of consumption an
d capital gains and depends critically on technological shocks exhibit
ing positive autocorrelation.