The current mainstream approach to monetary policy in the United States is based on the new Keynesian model and is expressed in terms of the federal funds rate. It ignores the role of financial intermediary leverage (or collateral rates). But as the federal funds rate has reached the zero lower bound, the issue is whether there is a useful role of leverage in monetary policy and business cycle analysis. Motivated by these considerations and by recent financial intermediary asset pricing theories in this paper, we investigate the macroeconomic effects of broker–dealer leverage and the interdependence between monetary policy and broker–dealer leverage in the context of a structural vector autoregression model, using quarterly U.S. data over the period from 1967:1 to 2014:3. We address the simultaneity problem of identifying monetary policy and leverage shocks by using a combination of short‐run and long‐run restrictions. We also use the sign restriction approach to the identification of shocks to distinguish between leverage supply and leverage demand shocks, as one would expect the macroeconomic effects of these two types of leverage shocks to be quite different. Our results show that monetary policy and broker–dealer leverage demand shocks produce results that capture reasonable macroeconomic dynamics.