Title: Quantile vector-autoregressions
Authors: Tomas Krehlik - Charles University in Prague (Czech Republic) [presenting]
Jozef Barunik - UTIA AV CR vvi (Czech Republic)
Abstract: Quantile vector autoregressive processes are introduced for modeling rich dependence structures in economic and financial time series. A novel simulated Whittle-like minimum distance estimator of a general process is devised for estimation of parameters. An overview of construction of the estimator is provided accompanied by Monte Carlo simulations and theoretical results that underscore its properties. Moreover, we discuss methodological issues with the estimate, various potential applications of the methodology on both financial data and macroeconomic data that highlight usefulness of the model for an economist.