Capital reallocation between firms is procyclical and leads to variations in measured aggregate productivity. In this paper, we ask how much of the cyclical variation in measured productivity is the consequence of capital reallocation. We build a heterogeneous‐firm model to study the effects of exogenous shocks to total factor productivity (TFP) and to the costs of reallocation. These shocks cause an endogenous cyclicality of measured aggregate productivity. Only a model driven by exogenous TFP shocks is able to generate both data‐consistent cyclical movements in reallocation and sizeable variations in measured aggregate productivity. We find that capital reallocation does not play a major role in amplifying aggregate productivity variations over the business cycle.