Quantile unit root inference for panel data with common shocks

Abstract

This paper proposes new quantile unit root tests for panel data with common shocks, whose critical values can be simulated based on our asymptotic theory. The Monte Carlo simulation results indicate that our tests perform well in finite sample. An application for real effective exchange rates (REERs) suggests that the Purchasing Power Parity (PPP) hypothesis holds for higher quantiles but does not hold for lower ones, and the asymmetric dynamics of REERs provide a plausible explanation for the PPP puzzle. • We extend quantile unit root test to panel data with common shocks. • A simple and practical procedure for the tests is proposed, and the critical values can be obtained based on our asymptotic theory. • Monte Carlo simulations suggest the tests perform well in finite sample. • An application for real effective exchange rates (REERs) indicate that there are asymmetric dynamics in REER series, which provides a plausible explanation for the PPP puzzle.