摘要：Using unit root test, hypothesis of purchasing power parity in Croatia is tested with the data spanning from 1952 to 2003. Although “the power problem” suggests that at least 75 annual observations of the real exchange rate is required to reject the null hypothesis with 50% probability, unit root test technique with only 51 annual observation for Croatia has rejected unit root hypothesis. Furthermore, using simple, autoregressive models estimated on the data, we show that univariate equations explain 20-55 percent of the in-sample variation in real exchange rates, although the degree of short-run persistence was high in certain periods. The econometric estimates imply a half-life of shocks to the real exchange rate of about 0.9 years for mark-kuna, 2.2 years for dollar-kuna and 1.2 years for lira-kuna.
关键词：real exchange rate; purchasing power parity; the power problem; unit root; long span test.
