This study assesses how banking sector integration and especially cross-border lending affect macroeconomic stability. I use a two-country general equilibrium model with heterogeneous banks that are hit by idiosyncratic shocks. According to the concept of granularity, idiosyncratic shocks to large firms (or: banks) do not have to cancel out under a skewed distribution of firm sizes. Given the highly skewed distribution of bank sizes, macroeconomic stability may thus be affected by shocks to large banks. Hence, to grasp the impact of financial liberalization on aggregate fluctuations, the presence of large banks as measured by high concentration in the banking industry has to be accounted for. I study the role of different forms of banking sector integration - i.e. arms-length crossborder lending versus lending via foreign affiliates - for the stability of aggregate lending. I find that banking sector integration decreases the aggregate volatility of lending due to intensifiedcompetition. The model implies that cross-border lending is more stable under lending via foreign affiliates than under arms-length cross-border lending.