A0907
Title: Quantile forward regression in high-dimensional distributional counterfactual analysis
Authors: Hongqi Chen - Hunan University (China) [presenting]
Abstract: The focus is on introducing a novel quantile forward regression approach for constructing distributional artificial counterfactuals. In the context of counterfactual analysis, where the number of control units frequently surpasses the pre-treatment time dimension, the quantile forward regression provides an approach to mitigate this challenge. The methodology involves the step-wise selection of control units from a candidate set. The theoretical properties of quantile forward regression are established, encompassing a bound on its weak submodularity ratio and asymptotic convergence results, and its asymptotic efficacy is assessed. Through extensive Monte Carlo simulations, the superior finite sample performance of the quantile forward regression approach is showcased compared to the $l_1$-penalization approach. The evaluation focuses on counterfactual prediction accuracy and the selection of control units. Finally, the application of quantile forward regression is demonstrated in an empirical study, analyzing the impact of an anti-corruption campaign on luxury watch importation.
