How large is volatility due to large firms? We answer this question through both reduced-form analysis and a calibration exercise. First, we exploit time and spatial variation across German cities and show that i) higher concentration is associated with more persistent local business cycles, ii) local concentration Granger-causes local employment volatility. From a business cycle perspective, we find evidence in favor of granularity-driven recessions only. Next, we calibrate a structural model along the lines of Carvalho and Grassi (2019) and find that the more fat-tailed productivity distribution in bigger cities crucially depends also on the higher probability for firms to grow.