Testing the Relative PPP Hypothesis in CEE States – Does the ‘PPP Puzzle’ Still Keep up?

This paper is focused on testing the relative version of the purchasing power parity (PPP). It tries to shed light on this so called “PPP puzzle” for a set of transition countries – twelve new EU Member States. Since results of similar studies in the literature have been ambiguous, a set of econometrics methods is employed: univariate (standard) tests (URT s), robust versions of URT s including high-powered ones – the non-linear Kapetanios, Sollis (non-linear adjustments) and Bierens’ test (non-linearities in trends). Three definitions of exchange rates (bilateral currency pairs) are utilised: for the euro, the US dollar and the CPI-based real effective exchange rate (REER). Due to limitations as regards to the availability of the data for our set of countries, quarterly data for the period 1995:Q1–2011:Q1 are used. The results of conventional linear do not provide a crystal-clear answer, more robust URT s at least partially do in favour of the PPP hypothesis. In addition, once the source of non-linearities has been controlled for (structural changes), the results show stochastic stationarity of all three series.