A1467
Title: Detecting structural changes in time varying parameters of panel models
Authors: Padma Sharma - Federal Reserve Bank of Kansas City (United States) [presenting]
Abstract: A hierarchical Bayesian procedure is developed to study the dynamics of bank stock returns to changes in their capacity to provide liquidity as well as the strength of their capital position and identify structural changes in this relationship over the last 30 years. The hierarchical model relies on a dynamic extension of the spike-and-slab prior that identifies change points in the relationship between bank stock prices and their liquidity buffers as well as their capital ratios. The proposed framework detects distinct structural changes across different covariates, which remain undetected by existing methods that only detect a set of common change points across all covariates. The analysis uncovers previously overlooked instances of structural changes in the relationship between bank liquidity and equity returns. Structural changes occur less frequently in the relationship between capital ratio and equity returns and are limited to periods of severe market distress, such as the global financial crisis and the onset of COVID-19. The method highlights the importance of a flexible method for detecting structural changes that allow for different instances of change points across covariates.
