Testing Purchasing Power Parity Theorem for Interdependent Cross-Sections: An Application of Second-Generation Panel Cointegration Estimators

Abstract

The aim of recent study is to re-investigate the long-run Purchasing Power Parity (PPP) hypothesis for 27 European states from a new perspective. It is based on its conventional form which assumes perfectly competitive markets and the absence of trade frictions. The analysis uses second panel data for the years 2000 to 2022 and employs both first and second-generation cointegration techniques (PMG, MG, DFE, CS-ECM, CS-ARDL and CCE). No substantial support is obtained in favor of the valid existence of PPP for our subject economies based on initial first-generation findings. Among all employed estimator, only PMG is the one that shows a statistically significant long-run relationship between the exchange rate and the price gap, in line with theoretical expectations. However, cross-sectional dependence tests show strong economic interlinkages between countries, prompting the use of second-generation techniques. All in all, results are conflicting: CS-CEM provide slight supports for PP under contemporaneous averages, whilst CS-ARDL and CCE yield contradictory evidence when lags are included. The study reveals that there is little strong and consistent evidence that PPP is valid over the long term in Europe. Further, the study suggests that exchange rate modeling and economic convergence strategies in Europe should be cautious in relying on PPP assumptions. This prudence is most justified in the absence of adequate consideration of cross-country interconnection and market imperfection.