Lumpy Investment and Monetary Policy : A Comparative Analysis in a DSGE New-Keynesian Framework

This paper analyses the theoretical and policy implications of assuming firm-specific lumpy investment behaviour by firms and compares such implications to those occurring when adopting different investment specifications in a new-Keynesian framework. We develop numerical simulations of the lumpy investment model by Sveen and Weinke (2007) and of other five specifications embedding either firm-specific capital or rental market of capital. The assessment is based on impulse responses and moments of economic variables following a productivity and a monetary shock.We preliminarily show that the degree of price stickiness and investment rigidity due to the lumpiness assumption complement each other in allowing for significant effects of monetary policy. Then, the comparison shows that models with firm-specific capital generate more volatility along the business cycle following a monetary shock while dampen the responses to a productivity shock. In such a way, they predict bigger real effects of monetary policy and better replicate empirical data. Within the firm-specific case, a degree of investment lumpiness compatible with the empirical evidence at the micro level makes economic variables more responsive to monetary and productivity shocks compared with the case in which firm-specific capital is associated to convex adjustment cost. Lastly, we show that in such setting a central bank's policy reaction function targeted to interest rate smoothing adds volatility to aggregate variables, compared with a standard Taylor rule