The purpose of this paper is to revisit the evidence for purchasing power parity (PPP) using long, low‐frequency data (over 100 years) for 23 organization for economic co‐operation and development (OECD) countries against each of four different base currencies – the Deutsch mark, the Japanese yen, the British pound, and the US dollar.
The paper uses standard unit root tests and level and trend stationarity tests, and also investigates the robustness of the results to alternative testing methodologies from statistical physics, such as Lo's modified rescaled range statistic and the Hurst exponent.
The results indicate that the theory of PPP does not hold.
Motivated by the mixed results from previous research on the validity of the theory of PPP, the robustness of standard unit root and stationarity tests to alternative testing methodologies are investigated. In particular, the paper uses two tests from statistical physics – Lo's modified R/S statistic and the Hurst exponent.
