We study the role of capital accumulation for the aggregate dynamics of total factor productivity in a general equilibrium model with rich firm heterogeneity. Using data on the census of incorporated Italian firms, we document that firms that have more recently exhibited large investment episodes, or \textit{spikes}, are more productive than firms that exhibited comparable spikes in a more distant past. Our estimates show that the timing of capital expenditures at the firm level explains about 15 percent of the productivity heterogeneity observed in the sample. Building on Khan and Thomas (2008), we formulate a state-of-the-art model of firm heterogeneity to assess the aggregate relevance of this microeconomic behavior. A non-convex adoption cost prevents firms from adopting the latest technology. In equilibrium, as new and old vintages coexist, the non-degenerate distribution of capital stocks and technologies across firms determines aggregate total factor productivity. After fitting the model to reproduce the distribution of investment rates observed in the data, we show that vintage effects constitute a microeconomic-based amplification mechanism of aggregate shocks relative to a benchmark real business cycle model.
A deterioration in financial conditions or a slowdown in the growth rate of technological progress induce firms to postpone investment expenditures leading to aggregate productivity losses. Through this channel, the shift in the microeconomic distribution contributes to deepening the recession and amplifying the effect of the initial shock.
In an application to Italian data, we show that the model accounts for about one-third of the stagnant productivity growth observed following the 2012 recession.