| This paper studies the sources of fluctuations and its dynamic evolution in 15 countries with over 50 years' long-term data. We find that: The contribution of non-tradable RER increases in countries which have undergone a dramatic economic growth. In floating exchange rate period, the changes in relative price of tradable RER is more important in accounting for fluctuations in real exchange rate. In sum, it's still very hard to find any satisfactory signal to confirm purchasing power parity. This article further explores the shock sources of the RER. By dividing the impacts on RER into supply,demand,fiscal and monetary factors, the structural vector autoregressive model indicates that money supply and demand factors being the major shocks, both the proportion of occupied maximum near 80%. This conclusion directly explains why the non-tradable RER dominate the evolution of RER - the fluctuations in nominal exchange rate, which is usually driven by money supply, results in the change in RER.Finally, we test whether the Balassa-Samuelson hypothesis could explain the evolution of long-term RER. What we found is controversial with the common sense-the BS hypothesis plays an important role in the evolution of non-tradable RER, but this effect is nearly muted by greater fluctuations in tradable RER.Because of commodity market segmentation around global, as well as large swings in the nominal exchange rate, purchasing power parity is much more difficult to hold than excepted, resulting in the persistence of tradable goods price differential. Therefore, a country in determining its exchange rate policy can be more in accordance with the needs of the country's economic requirements. This conclusion will bring particularly practical value to the exchange rate reform in China, which is right now undergoing a comprehensive adjustment in price system. |