Intangible Capital and the Excess Volatility of Aggregate Profits

Aggregate profits measured from NIPA data are over six times more volatile than output. We use recent estimates of the return on physical capital to decompose NIPA profits into the part that can be explained by capital income and a residual called net profits. We find that capital income explains only a small fraction of the overall volatility of NIPA profits, the net-profit series is over seven times more volatile than output with a contemporaneous correlation of .55. Most dynamic general equilibrium models of the business cycle cannot deliver this excess volatility of net-profits as they rely mainly on capital income to explain NIPA profits. We develop a model of the U.S. economy in which firms expend resources to create intangible capital (IC), which is an additional input in their production technology. The model is estimated using aggregate data on output growth and labor productivity. Simulations using two versions of the estimated model deliver pro-cyclical net profits that are over three times more volatile than output. IC investments are large and pro-cyclical and act to propagate shocks over time. Overall, the model explains aggregate US business cycles well on traditional metrics. In particular it does a good job of explaining observed movements in hours, productivity, output and the labor wedge without relying on preference shocks. As an external validation exercise, we show that the model is capable of delivering movements in IC investment that resemble recent estimates of R&D investment in the US. Moreover it delivers counter-cyclical markups along with a pro-cyclical profit share, another feature of the data. We emphasize that we do not use profit data to estimate the model so that any movements in profit over and above those of output are solely generated within the model as firms optimally reallocate resources between goods production and intangible capital creation.