On The Calibration of Competitive
Industry Dynamics Models
By Gian Luca Clementi and Dino Palazzo
Abstract
We show that, in
the context of the neoclassical model of investment with mean--reverting and
log--normally distributed productivity shocks, information on the asymptotic
distribution of the investment rate does not identify the parameters of the
stochastic process. This likely explains why a variety of recent models with
firm--level heterogeneity - both in macroeconomics and in finance - are able to
generate sensible cross--sectional investment rate moments in spite of assuming
radically different values for the persistence and volatility of the shocks.
Information on investment rates does entail a restriction on the two
parameters, which in turn implies that -- contingent on the curvature of the
production function -- not all recent estimates of the parameters will be
consistent with empirically plausible cross--sectional investment rate moments.