Bank size and macroeconomic shock transmission: Does the credit channel operate through large or small banks?

In this paper, I use U.S. call report data to construct a larger panel dataset with bank-level observations. I find that larger banks' lending is considerably more sensitive to the strength of their borrowers' and their own balance sheets compared to smaller banks and that the sensitivities to borrower balance sheets are larger in magnitude compared to lender balance sheets. When I incorporate various macroeconomic shocks (identified by an estimated DSGE model) into the empirical model, I similarly find that the transmission of shocks to the real economy operates mostly through large bank lending and borrower balance sheets.