Sato et al. (2013) represents a significant contribution to the literature. The novelty of their paper lies in the construction of a new dataset on industry-specific real effective exchange rates (REER) on a monthly basis from January 2005 to the present. The authors have convincingly demonstrated that REER differs not only between countries, but also across industries, a finding that would have slipped through the net of aggregate data analysis.
The authors have used the industry-wise producer price index (PPI), instead of consumer price index (CPI), not only because the latter includes nontradables, but also because the former tends to vary across industries, reflecting interindustry differences in costs arising from differences in their exposure to exchange rate changes. In this context, industries with large import content are more sensitive to exchange rate changes than those with a large domestic content. It, thus, makes considerable sense to use the industry breakdown PPI data. However, exchange rate changes, in purchasing power parity (PPP) terms, can only be explained by the inflation differential. According to PPP, a currency with a relatively lower inflation rate would appreciate vis-à-vis currency with a higher inflation rate. Seen in these terms, one might still argue that CPI is more relevant for explaining exchange rate appreciations and depreciations, be they in nominal or real terms.
The spotlight on the electric machinery industry is revealing. The results show that the Japanese electric machinery REER exhibits the largest depreciation among Japanese industries. This finding suggests that Japanese electric machinery should be enjoying substantial export price competitiveness. On the contrary, the Japanese electric machinery firms are suffering from worsening business performance and tougher competition in the export market. The explanation given by the authors is that Japan's competitors in the electric machinery industry, especially the Korean firms, enjoy a much larger REER depreciation, and hence stronger export price competitiveness. This is further supported by the results of a factor decomposition analysis that attributes the Korean electric machinery export price competitiveness to a significant drop in domestic producer prices in Korea during the won appreciation period.
While the authors' deduction makes good sense, it does not fully explain the substantial fall in domestic producer prices in Korea in the wake of won appreciation. It may have much to do with the differences in the product mix and the share of imported inputs relative to domestic content in the total cost of production. Export price competiveness gained through REER depreciation can be blunted by higher costs of imported inputs, as an exchange depreciation makes imports more expensive. By the same token, the loss of export price competiveness resulting from REER appreciation can be eroded by lower costs of imported inputs.
Presumably, the Korean electric machinery firms have a much higher import content in their manufactures than their Japanese counterparts. An extension of this analysis into the role of domestic versus import content in manufacturing is needed to better understand the mechanics and dynamics of REER impact on export price competitiveness. Much would, of course, depend on the sources of imports and the strength of the currencies of these source countries.
That the Japanese automobile firms have not lost export competitiveness to their Korean rivals, thanks to the relative decline of domestic production costs, provides more fuel for further inquiry. The decline of domestic production costs in the automobile sector in Japan needs to be probed further. Again, it is unclear if this is attributable mainly to the ratio of domestic versus imported inputs or to the productivity gains at home. A comparison of productivity trends in the automobile industry in Japan and Korea would be helpful, although this transcends exchange rate issues.
While changes in domestic production costs may, thus, represent an important factor in the export price competitiveness/exchange rate equation, there is a need to balance these cost considerations with actual export performance. As is often said, the proof of the pudding is in the eating. It is, therefore, essential to demonstrate that industries with REER depreciations do exhibit better export competiveness than those with REER appreciations. To do this, one has to go beyond an exchange rate analysis into a trade analysis with respect to export destinations, export composition, and the intensity of competition in the export market. A close look at the shifting market shares of Japan, Korea, and China for electric machinery and automobiles in the US market – where the competition is fierce – with an effort to relate them to REER changes would shed more light the issue at hand.
In any case, REER can provide only a partial, albeit dominant, explanation for export price competitiveness. Many other factors can also contribute to export competitiveness. These may include cost of capital, unit labor cost, technological advancement, etc. Be that as it may, an unmistakable sign of companies being hurt by REER appreciation is their decision to relocate their plants to offshore bases. Such evidence will be highly instructive, reinforcing the rigorous REER analysis.