Lumpy Investment, Fluctuations in Volatility and Monetary Policy

58 Pages Posted: 16 Mar 2020 Last revised: 26 Oct 2021

See all articles by Min Fang

Min Fang

HEC Lausanne; Geneva Finance Research Institute

Date Written: February 14, 2020


I argue that monetary policy is less effective at stimulating investment during periods of elevated volatility in firm-level productivity. Empirically, I document that high volatility weakens investment responses to monetary stimulus. I then develop a heterogeneous firm New Keynesian model with lumpy investment to interpret these findings. In the model, non-convex capital adjustment costs create a sizable extensive margin of investment which is more sensitive to changes in both the interest rate and volatility than the intensive margin. When volatility is high, firms tend to stay inactive at the extensive margin, so monetary stimulus motivates less investment at the extensive margin. I find that the quantitative implications of the model are primarily shaped by the specifications of the capital adjustment costs. Unlike much of the prior literature, I use the dynamic moments of investment to identify this key model element. Based on this parameterization, high volatility reduces the effectiveness of monetary stimulus for investment by 30%. This reduction is about half of what I find in the data. Therefore, the effect of monetary policy depends on both the lumpy nature of firm-level investment and fluctuations in volatility.

Keywords: Lumpy investment; Ss model; irreversibility; volatility; uncertainty; firm heterogeneity; monetary policy;

JEL Classification: E52, E32, E22

Suggested Citation

Fang, Min, Lumpy Investment, Fluctuations in Volatility and Monetary Policy (February 14, 2020). Available at SSRN: or

Min Fang (Contact Author)

HEC Lausanne ( email )

Lausanne, 1015


Geneva Finance Research Institute ( email )

Bd du Pont d'Arve 42
Geneva, Geneva 1211


Do you have a job opening that you would like to promote on SSRN?

Paper statistics

Abstract Views
PlumX Metrics