Testing Purchasing Power Parity in Transformed ECM with Nonstationary Disequilibrium Error (in Korean)

In this paper, we show that the wide-spread rejection of the test for the hypothesis of purchasing power parity (PPP) may simply be due to the slow adjustment to the long run equilibrium, rather than the rejection of the hypothesis itself. Our approach is based on the decomposition of the nonstationarity in a transformed error correction model into two components: nonstationarity of the variables included in the model and nonstationarity of the disequilibrium error. We propose statistics that can be used to test whether the disequilibrium error is indeed stationary. The proposed tests have nonstandard limit distributions, but they can be computed using the standard bootstrap procedure. According to our simulation, the tests have relatively good size and power in finite samples. Empirical tests on the yen․dollar and pound․dollar suggest that the slow adjustment to the long run equilibrium is likely to be the reason why the PPP hypothesis is rejected.