Abstract:
We study the effects of monetary policy on output during financial crises. We use a large panel of advanced and emerging economies to guarantee a sufficiently high number of financial crises episodes. A financial crises dummy, which is constructed based on the narrative approach, is interacted with other key macroeconomic variables in a panel VAR. Theory suggests that monetary policy might be more effective in financial crises if it can ease malfunctioning of financial markets for example by loosening credit constraints or restoring confidence. Alternatively, deleveraging and uncertainty might predominate and make the economy less interest rate responsive and monetary policy less effective in financial crises. Taking a sample from the mid 1980s to today we find that an expansionary monetary policy shock is very effective in raising GDP during the recessionary period of a financial crisis. The effect is stronger than in non-crises times. In contrast, during the recovery period of a financial crisis, monetary policy has a very small effect on GDP. These differences can be explained by a confidence channel. During the joint occurrence of a recession and a financial crisis an expansionary monetary policy shock increases consumer confidence and GDP. During the following recovery monetary policy has no effects on confidence or GDP. Other variables like credit, housing prices and exchange rates can at most partially explain differences in transmission between the different regimes.
