| This dissertation develops the theory of extremal (near-extreme and extreme) quantile regression: estimation, inference, and applications in economics. In particular, the limit laws for extremal quantile regression statistics are obtained under conditions reflecting the inherently rare nature of the conditional extremal events, thus supplying a regression analogue of the classical extreme value theory. The relevant inference framework is provided. The results apply to cross-section as well as time series data. Examples, ranging from decision making under extreme uncertainty to price search models, provide motivation for and empirical applications of the theory developed. |
