The following plots show the estimates of interactions between bank dummies and quarter dummies obtained from regressions of the form
\[r_{ikt}=\alpha+\sum_m\beta_mq_m+\sum_b\gamma_b bank_b+\sum_g\beta_g (bank\times quarter)_g+u_{ikt}\]
where \(r_{ikt}\) is the demeaned interest rate firm \(i\) pays on average to bank \(k\) in quarter \(t\) (alternatively it will be the total debt firm \(i\) owes bank \(k\) at the time). We are interested in the \(\beta_g\)s.
Although we estimate one big regression with all the banks and all the quarters, we present the estimates in the following plots focusing on each pair of banks involved in the merger.
Panels 1,2 and 3 show the estimates of the interactions for 3 of the mergers with their 95% CI. The 4th panel shows the estimates of the quarter fixed effects (should approximate what’s going on with the rivals).