CMStatistics 2018: Start Registration
View Submission - CFE
Title: Structural estimation of time-varying spillovers: An application to international market liquidity Authors:  Lukas Boeckelmann - Paris School of Economics (France) [presenting]
Arthur Stalla-Bourdillon - Dauphine University - Banque de France (France)
Abstract: A structural version of the popular Diebold-Yilmaz spillover framework based on a single comprehensive empirical approach is proposed. Key to our approach is a SVAR-GARCH model that is identified by heteroskedasticity and allows for the construction of time-varying up-to-date forecast error variance decompositions. Building on these advances, we analyze the degree, time variation, direction and determinants of market liquidity spillovers on international equity markets. We find that liquidity spillovers (i) have a significant though time-varying impact (ii) are stronger between stock markets with more correlated fundamentals and (iii) increase in times of high risk aversion.