Econometrics at the Extreme: From Quantile Regression to QFAVAR 1
这篇综述梳理了分位数建模从理论起源到当前进展，涵盖横截面、时间序列、多元、面板及因子增强模型，为政策制定者提供超越平均值的分布动态分析工具。
ABSTRACT This paper surveys quantile modelling from its theoretical origins to current advances. We organize the literature and present core econometric formulations and estimation methods for: (i) cross‐sectional quantile regression; (ii) quantile time series models and their time series properties; (iii) quantile vector autoregressions for multivariate data; (iv) quantile panel models for longitudinal data; and (v) quantile factor‐augmented models for information compression in data‐rich environments. Each section outlines theoretical foundations and developments, followed by representative empirical applications. Finally, the survey highlights open gaps in quantile modelling. By studying distributional dynamics beyond averages, quantile methods provide policymakers and regulators with tools to design interventions that are robust to risks and effective across the entire spectrum of possible outcomes.