Panel data methods that account for cross-sectional dependence in investigating the long-run theory of purchasing power parity for 16 developed countries

Abstract

The purchasing power parity theory (PPP), one of the widely investigated areas in International Finance, proposes that prices of goods ought to be the same in both local and foreign economies or internationally. This paper examined the PPP theory for a group of 16 developed countries using powerful statistical panel data analyses that put into consideration dependence on cross-sections in panel data. The paper used the Pesaran panel unit root test, the cointegration test of Westerlund, the linear Augmented Mean Group (AMG) and Common Correlated Effect Mean Group (CCEMG) estimators. In addition, this study extended the linear AMG and CCEMG estimators to nonlinear AMG and CCEMG estimators in examining the long-run PPP theory. The test of unit root revealed that all the variables are integrated of order one. Since all variables were I(1), tests of cointegration were conducted to see whether there are long-run relationships amongst the variables. The tests of cointegration depicted the appearance of long-run relationships and hence, the occurrence of PPP in this group of countries. Furthermore, the long-run relationship was evaluated using both linear (AMG and CCEMG) and the extended nonlinear (AMG and CCEMG) estimators. It was seen that the PPP theory is valid in the long-run and there is nominal exchange rate appreciation and depreciation.