Academic journal article Multinational Business Review

Testing for Purchasing Power Parity: A Data Matching Problem

Academic journal article Multinational Business Review

Testing for Purchasing Power Parity: A Data Matching Problem

Article excerpt


The purchasing power parity (PPP) theory of exchange rate was popularized by Cassell in the 1920's. The idea advanced by Cassell is that international rates of exchanges are determined by the purchasing power of the national currencies so that exchange rates are merely the reflection of the law of one price. Thus, purchasing power parity in its strongest form, assumes that any differences in the goods market should be reflected in the currency market. The idea of PPP has been extensively tested and studied. Nevertheless, the empirical validity of this concept is still unclear. The continued interest and research done in the area attest to the importance associated with the theory, which is one of the bases for models of exchange rate behavior. The use of this and other exchange rate theories is of immence importance (due to its implications for setting prices and hedging) to multinational firms (MNFs). Thus it is very important to establish whether this theory of exchange rate determination is accurate or even applicable.

One of the simplest tests for purchasing power parity was conducted in the 1970's by Isard, [10] where he questioned the validity of parity by showing that changes in exchange rates are paralleled closely by movements in price ratios, thus negating the law of one price. Several studies have also suggested that empirical evidence validating the long-accepted parity is weak. Studies by Frenkel [5], Junge [11]), Krugman [13] all find evidence against PPP. More recently, Kim [12], who used the cointegration approach, found evidence in favor of PPP for some currencies. But studies by Ender [3] and Corabe and Quliaris [1], which also used cointegration, found little evidence in favor of PPP.

Given the mixed evidence, it is only logical to wonder if the differing conclusions for or against the concept of PPP are a result of flawed data or flawed methodology utilized to conduct the studies. In order to resolve this controversy, it would be worthwhile to address the obvious difficulties in the previous studies.

One possible area involving problems with testing purchasing power parity which has not received close empirical scrutiny is the potential problem created by the structure of published data that are used in these studies. This problem was studied by Magee [14] and his approach suggested lagging the exchange rate and price data to match the time period of the commodities representing the contracts. This approach delves into "data contamination" that may result from the fact that currency data available during a certain period may not reflect the time period at which the contracts originated.

Another problem in testing for PPP is the issue that parity should not be considered as a theory of exchange rate determination. Rather, it is a long-run equilibrium phenomenon. Thus, it is not surprising to observe deviations from PPP when tests are done using a short-run context. To address the issue that PPP is a long-run phenomenon, we utilize the cointegration technique.

On the empirical side, to ensure robustness of the results, both the consumer and the wholesale price indices were used to proxy the relative prices of the commodities where CPI is more representative of the consumer basket of goods while the WPI concentrates more heavily on goods, which by their nature, are traded more often across nations. In addition, both the absolute and relative versions of PPP were tested and nine developed countries were taken into consideration to show how parity holds across various countries.

The rest of the paper is organized as follows. The first section describes the data used in this study. The second section deals with the question of data mismatching, and the third section deals with the cointegration approach to testing PPP. And finally, the last section summarizes the paper.


In this study we use exchange rate data from the post-Bretton Wood era. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed


An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.