In this paper we use a long memory framework to examine the validity of the Purchasing Power Parity (PPP) hypothesis using both monthly and quarterly data for a panel of 47 countries over a fifty year period (1957 to 2009). The analysis focusses on the long memory parameter d that allows us to obtain different convergence classifications depending on its value. Our analysis allows for the presence of smooth structural breaks and it does not rely on the use of a benchmark. Overall the evidence strongly points to the presence of a long memory process, where 0:5 ≤ d < 1. The implication of our results is that we find long memory mean reverting convergence, something that is also consistent with Pesaran et al (2009). In explaining the speed of convergence as captured by the estimated long memory parameter d we find impediments to trade such as distance between neighboring countries and sticky prices to be mainly responsible for the slow adjustment of real exchange rates to PPP rather than nominal rates for all country groups but Asia, where the opposite is true.
