A0179
Title: Dynamic models for multiple quantiles
Authors: Alessandra Luati - Imperial College London (United Kingdom) [presenting]
Abstract: Recent developments in models for dynamic multiple quantiles are discussed. The baseline semiparametric model introduced recently, based on quantile spacings and score-type updates, is reviewed and extended to account for: heterogeneous tail behaviour, cross-tail effects, and exogenous variables. The extensions result in a flexible class of models ensuring that quantiles do not cross in finite samples and that extreme quantiles are estimated based on information coming from all the regions of the underlying conditional distribution. M-estimation is carried out, and the asymptotic properties of the estimators are discussed. Open problems and illustrations conclude the tutorial.
