Testing absolute PPP hypothesis for twenty countries through the skeleton from a SETAR model- some new evidence

The long run PPP hypothesis was tested considering real effective exchange rate dataset for twenty countries which it was provided by the International Monetary Fund (IMF). By focusing on a nonlinear approach, the study tests IMF monthly dataset for general and specific nonlinearities. Additionally, the study presents a method to estimate the value that real exchange rate may converge in the long run. Linear and nonlinear cases were distinguished by the Tsay and Hansen?s test. The number of regimes was determined by the Hansen?s test. The Self-Exciting Threshold Autoregressive (SETAR) model was applied to estimate potential thresholds to indicate the states turning points of the countries competitiveness. Results suggest that real exchange rate for thirteen countries are highly nonlinear and subjected to regime switching. The asymptotic stability analysis guarantees the data stationarity behavior. Absolute PPP hypothesis was supported in five out of thirteen cases. The real exchange rate generally converges to a stable equilibrium not far from the value predicted by the PPP hypothesis asserts.