We build a DGE model that ads a banking sector to the standard RBC model. We look at the response of the real interest rate to innovations in the banks' technology and in the nonbank firms' technology. While technological innovations in the nonbanking sector put upward pressure on the interest rate, technological innovations in banks exert downward pressure on the interest rate. This implies that, if the technological innovations in banks are strong enough, stochastic simulation experiments generate negative correlations between the real interest rate and current and future values of real output. This is especially significant because negative correlations between the interest rate and output are a key post-war U.S. business cycle fact difficult to replicate in benchmark dynamic models