Some Stylized Facts and Empirical Results

The large appreciation of the yuan occurred after PBC switched from the peg to the dollar to a basket of major currencies in July 2005. Before the change, the yuan's nominal exchange rates against the dollar were virtually kept constant. The policy shift opened the door to the yuan's exchange rate variations following market forces. Figure 1 [ PDF 88.4KB | 1 page ] shows monthly yuan exchange rates against the dollar from 2004 to 2008, along with prices of the US imports from PRC. Both are measured in indexes with the first month of 2004 as the base month. Prior to July 2005, the exchange rates of the yuan to the dollar remained constant. Then, the yuan appreciated steadily until July 2008. The cumulative appreciation added up to 21%. During this period, the price of the US imports from PRC, however, initially decreased. The decline was reversed in May 2007, after that the import prices rose about 5% by the end of 2008. The pattern of the import price movements appears less responsive to the yuan's appreciation.

Figure 2 [ PDF 26.7KB | 1 page ] illustrates movements of the bilateral exchange rates between the yuan and the yen and aggregated price indexes of all imports as well as prices indexes of manufacturing imports from the PRC. Swings of the bilateral exchange rates mainly reflect movements of the yen's exchange rate against the dollar before July 2005. The volatility of the exchange rates was very high. From August 1998 to December 1999, the yuan depreciated against the yen and the yen/yuan exchange rate fell to 12.4 from 15.6; then, the yuan started to appreciate against the yen and the yen/yuan exchange rate reached the level of 16.1 in February 2002. After that, the yuan trended downward again until January 2005, when the yen/yuan exchange rate was 12.5. A new appreciation cycle ended when the global financial crisis erupted in September 2008 and the PRC government reverted to the rigid peg to the dollar policy. Despite the high volatility, prices of Japanese imports seemed to track closely movements of the bilateral exchange rates between the yuan and yen. The correlation coefficients are surprisingly high, 0.56 between the aggregated import price and the exchange rate, and 0.71 between manufacturing import price and the exchange rate. The high correlations suggest that, Japanese import prices had been adjusted to reflect the exchange rate variation.

In the estimations, nominal exchange rates between the dollar and the yuan were defined as dollar/yuan; nominal exchange rates between the yen and the yuan were defined as yen/yuan. Therefore, higher values of exchange rates mean that the yuan appreciates against the dollar or the yen. The estimated coefficients of exchange rates Δe_t and its lags are expected to be positive and significant should there exist any pass-through. The regression models with different lagged independent variables were estimated. I applied the Akaike information criterion (AIC) to determine optimal numbers of lags. I started with zero lags and then added lagged variables. Once the value of AIC started to rise, I stopped estimating new models with additional lagged variables. The ordinary least squares method was employed to estimate the regression model.

For the US case, the estimates with one- and two-period lagged variables of Δe_t and one period-lagged variable of Δmc_t were selected. Table 1 [ PDF 21.5KB | 1 page ] summarizes the estimates. The estimated coefficient of Δe_t is 0.23 and statistically significant at 1%, suggesting that, if the yuan appreciates by 1% against the dollar, import prices would be expected to rise 0.23%; the estimated coefficient of Δe_t-1 is 0.02, but this is insignificant; the estimated coefficient of Δe_t-2 is 0.24, implying that 24% of the appreciation would be passed on to import prices after two periods. The impact of production cost changes on import prices was very small. There is no price adjustment in the current period for production cost changes as the estimated coefficient of Δmc_t is 0.03 and insignificant. The adjustment occurred one period later, but was only 0.06% of price increase for a 1% increase in marginal cost as suggested by the estimated coefficient of Δmc_t-1, which is significant at 5%.

In summary, the pass-through effect of the yuan's appreciation on prices of the US imports from PRC is about 23% in the short run and 47% in the long run. The low pass-through effect of the yuan's appreciation implies that moderate appreciation of the yuan would have very limited impact on the bilateral trade imbalance between the PRC and the US. Usually export demand elasticity is lower than one. The impact of the yuan's appreciation on PRC exports is determined by the joint impact of the pass-through and export demand elasticity. For example, if the export demand elasticity is 0.5, using the result of this research, we can conclude that a 10% nominal appreciation of the yuan against the dollar, would lead merely 1.15% decrease in PRC exports to the US in the short run, and 2.35% decrease in the long run; a 20% appreciation could only reduce the PRC's exports to the US by 4.7% in the long run, which is negligible compared with the more than US$260 billion trade deficit the US had with PRC in 2008. The relatively small sample size does undermine the accuracy of the estimated pass-through effects. It is worthwhile to be cautious when one interprets the policy implication of the result.

In the case of Japan, the sample size is much larger compared with the US. We use the monthly data from 1998 to 2008, 11 years including 132 observations. In addition, the exercises were done for not only all imports as a whole but also for imports in five commodity categories. Compared with the case of the US, Japanese import prices were apparently more responsive to variations of the bilateral exchange rates between the yen and the yuan and degrees of pass-through are much higher. Table 2 [ PDF 27KB | 1 page ] reports the estimates for all Japanese imports as a whole and imports in food, raw materials, apparel, manufacturing, and machinery. The model with only one-period lag of Δe_t was employed for the estimations. As a reference, we also reported estimates with one-period lag of Δmc_t and two-period lag of Δe_t side by side in the table.

For all Japanese imports from the PRC, the estimated coefficient of Δe_t is 0.55 and statistically significant at 1% and that of Δe_t-1 is 0.44, also significant at 1%. The estimates indicate that, for a 1% change in yen/yuan exchange rates, 0.54% would be expected to be transmitted into import prices in the current period and 0.44% in the second period. In other words, the pass-through effect of yen/yuan exchange rates on Japanese import prices is 55% in the short run and 99%, an almost complete pass-through in the long run. Moreover, changes in production costs also affected import prices significantly. The estimated coefficient of Δmc_t is 0.42 and significant at 1%, suggesting that 42% of marginal cost increase in PRC exports would be expected to be passed on to Japanese importers.

The substantially high pass-through in the short run and a near complete pass-through in the long run were also found in sectoral analysis. In manufacturing, the short run elasticity of import prices to yen/yuan exchange rates was 0.50 while the long run was 0.99; in food, 0.38 in the short run and 0.83 in the long run; in machinery, 0.36 in the short run and 0.71 in the long run; in apparel, 0.50 in the short run and 1.10 in the long; and in raw materials, 0.62 in the short run and 0.87 in the long run. All of these estimates are statistically significant at 1%. The empirical results imply a very high pass-through effect of the yuan's appreciation on prices of Japanese imports from the PRC. However, since the yuan has been pegged to the dollar, which is also used as a major invoicing currency, the role of the dollar should be an important factor in determining the pass-through effects. It is imperative to control the impact of the dollar and the currency invoicing factor when evaluating actual pass-through effects of the yuan's appreciation on prices of Japanese imports.

Download this Paper [ PDF 217.3KB| 18 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.