Estimating quantile treatment effects for panel data

In this paper, we propose a quantile treatment effect model for panel data to characterize the distributional effects of a treatment. We provide an identification condition to estimate the counterfactual quantile for the treated unit. In addition, we establish the asymptotic properties of the proposed quantile treatment effect estimator, together with an extension to the high-dimensional case, and discuss how to choose the control units and covariates using a LASSO (least absolute shrinkage and selection operator) type method. A simulation study is conducted to illustrate our proposed model and modeling method. Finally, we apply the proposed method to estimate the quantile treatment effects of introducing CSI 300 index futures trading on both the log-return and volatility of the stock market in China.