Daily interbank rate determination and volatility in a banking crisis

The paper studies daily interbank rate determination and volatility in the Dominican Republic during a major banking crisis. The investigation uses a novel, automatic, general-to-specific technology (PcGets) to reduce a baseline (mean) specification linking interbank rates and aggregate banking system excess reserves. This specification is subsequently embedded in a GARCH model. Recursive coefficient analysis reveals that in times of financial stability positive or negative shocks have similar effects. In contrast, during banking crisis negative impacts (e.g. a decrease in excess reserves) generate larger volatility of interbank rates than positive ones.