비물가 요인은 원·달러 환율에 대한 구매력 평가 불균형의 느린 조정을 설명할 수 있는가?

The Purchasing Power Parity (PPP) is a theory that relative price level between countries determines the exchange rate which is based upon an extension of intuitive law of one price. It is doing a role of key linkage to the monetaristic approach of long run exchange rate determination and open macro-economic analysis. Therefore its rigorous test and understanding are critical for modern economic theory evolution. However, according to Rogoff (1996), Obstfeld and Rogoff (2000), the variability of real exchange rate is larger than the level that is expected by the present models explaining the PPP deviation. According to Frankel and Rose (1996), Cheung and Lai (1998, 2000a, 2000b), Papell (1997), Wu and Wu (2001), the half life from PPP deviation to equilibrium adjustment is roughly 3～5 and thus disequilibrium adjustment speed is quite slow. Obviously this adjustment time is too much long than it expected by the speed is price rigidity models; i.e., 1～2 years. Therefore it is regarded as an important puzzle of international economic theory which demands fundamentally different sight to explain the connection of exchange rate with the price. With respect to this issue, remarkable hypothesis/evidences have been suggested to explain slow purchasing power adjustment. Firstly, Engel and Morley (2001), asserts that the root of the PPP puzzle may be resolved by allowing different speeds of convergence for nominal exchange rates and prices. In contrast to usual rational-expectations sticky-price models, which impose equivalent reversion speed for nominal exchange rates and prices, they examine an empirical model that supposes those variables to adjust at different speeds. Empirical evidences from state-space model estimation show that while prices converge relatively fast, nominal exchange ratesconverge slowly. It suggests that the lethargic rate of PPP reversion may come largely from slow nominal exchange rate adjustment rather than from slow price adjustment. Engel and Morley (2001, 24) conjectured that “bubbles or herding might temporarily send the exchange rate off on disequilibrium paths that result in the appearance of slow convergence to the equilibrium.” Cheung et al. (2004), using vector error correction model, estimated the speeds at which the individual variables revert to their long-run values. Their results confirmed those of Engel and Morley (2001) that nominal exchange rates do converge at a much slower rate than prices. Half-lives of nominal exchange rates are estimated to be from 3 to 6 years, while half-lives of prices are found to be just about 1 to 2 years. They also found that about 60~90% of PPP disequilibrium adjustment takes place through nominal exchange rate adjustment. They conclude that if nominal exchange rates converge much more slowly than prices, the PPP reversion speed can be slower than the price convergence speed, as described by the PPP puzzle. Wu et al. (2011) also found that main findings of Cheung et al. (2004) are robust to possible misspecifications in the true data-generating process (DGP). They examined the source of a real exchange-rate adjustment based on the impulse-response function constructed from local projections proposed by Jorda (2005) when the DGP is unknown. They showed that nominal exchange-rate adjustments dominate in the reversion toward PPP regardless of a nominal exchange-rate shock or a price shock. This paper examines these view that not the price but the bubble in the exchange rate mainly explains the slow adjustment of exchange rate to the PPP equilibrium especially focusing on the conjecture of Engel and Morley (2001, 24) through following two fold approaches. Firstly, following Kim (2014), we test whether there is an I(1) stochastic bubble trend, which is independent with an I(1) trend of price in exchange rate. For this, we exploit a transformed error correction model and Beveridge-Nelson de- composition. Secondly, the shock of aforementioned bubble trend may induce a longrun/sustainable response of the PPP disequilibrium error when the cointegration vector is different from it implied by the PPP (possibly due to the barrier/frictions for trade such as tariff, transportation costs which hinders the PPP arbitrage.) We theoretically expect the response of PPP disequilibrium error from which the part induced by the barrier/frictions for trade are deleted may be mainly affected by the ‘sticky’ price as con- ventional approach explains.