معکوس واردات، سرمایه گذاری مستقیم خارجی و نرخ ارز

This paper investigates linkages among “reverse imports”, foreign direct investment and exchange rates. As an example, we have in mind the competition in the Japanese market of a Japanese multinational firm and a Chinese domestic firm. Products are differentiated based on Japanese consumers’ brand name recognition. The model shows that yen appreciation leads to an increase in Japanese production in China and “reverse imports” and a decrease in Japanese domestic production. Due to the barriers in brand name, the exports of the Chinese firm could fall, because the increase of reverse imports may erode the market share of the Chinese firm, even though total exports from China increase. Further, we find that yen appreciation may improve the profits of the Japanese firm and welfare in Japan under reverse imports, against conventional wisdom. The predictions of the model fit well with the actual numbers and shed light on the current debate on the Chinese currency.

China's exports have experienced two-digit growth rates in the past decade. “Made in China” is available in almost every corner of the global market. Many point to the relatively low labor cost as the ‘secret of success’. While we do not wish to down play the importance of labor costs, we note that labor in China was even cheaper 10 years ago compared with that of today. Why was “Made in China” not popular then? A quick reality check reveals that the devaluation of the Chinese yuan since the late 1980s played an important role in fostering China's export growth in the last two decades, especially through the indirect effect—the influx of export oriented foreign direct investment (FDI). In fact, this indirect effect has been far more important in increasing exports than the direct effect, which is on the exports of Chinese domestic firms. For instance, from 1992 to 2005, China's exports more than tripled from US$ 84.94 billion to US$ 762.00 billion (China Statistics Yearbook, various years). But such dramatic increase is mainly the contribution of foreign multinationals, i.e., the subsidiaries of global multinational enterprises (MNEs) in China, because the exports of foreign invested firms surged from US$ 1.74 billion to US$ 444.20 billion over the same period. China has become the largest destination country for FDI in the developing world. Multinationals flock to China for its cheap labor, land and other inputs (which Krugman, 1998, calls “firesale FDI”), to conduct export-processing operations. These have caused cries of “deindustrialization” and “hollowing out” in the home countries of the MNEs.2 The present paper sets out to investigate the linkages among “reverse imports”, FDI and exchange rates. We especially have in mind Japanese subsidiaries producing in China and then importing back to Japan for consumption, i.e., the so-called “reverse imports”.3 One Chinese and one Japanese firm compete in the Japanese market. The Japanese firm can also produce in China. Products are differentiated due to brand name recognition. We find that exchange rate changes, wage differentials and barriers in brand name recognition contribute to increases in Japanese outward FDI and reverse imports. Specifically, the appreciation of the yen leads to an increase in Japanese FDI in China and “reverse imports” and a decrease in Japanese domestic production. Due to the barriers in brand name, the exports of the Chinese firm could fall, because the increase of reverse imports may erode the market share of the Chinese firm, even though total exports from China increase. In addition, yen appreciation may improve the profits of the Japanese firm and welfare in Japan under reverse imports, which is contrary to conventional wisdom. This implies that outsourcing has made the Chinese economy and those of the MNEs’ home countries interdependent, as is recently recognized by not only business groups but also the Japanese government and the media that a major contribution for the Japanese GDP growth in 2003 and 2004 was “the China factor”. It is counter intuitive that the yuan's depreciation may erode the Chinese firm's market share. This arises under the possibility of reverse imports: yen appreciation helps the Japanese firm to gain an edge on the Chinese firm in acquiring cheaper Chinese inputs, leading to more Japanese FDI and reverse imports. The relatively low cost of reverse imports together with high consumer recognition enhances the competitiveness of the Japanese MNE in the Japanese domestic market. If the substitutability between the products of the Chinese firm and the Japanese affiliates in China are high, then these increased reverse imports will replace goods made by the Chinese firm. The model sheds light on the current China debate, because the predictions of the model fit well with the actual figures of China. The sudden popularity of “Made in China” can be ascribed to China's devaluation in the early 1990s, in addition to other factors. The devaluation reduced the input cost and improved the relative wealth of foreign investors, leading to more FDI inflows and eventually higher reverse imports to other countries. In 2001, more than 50 percent of China's exports are produced by foreign firms. In other words, it is the influx of export-oriented FDI, which contributed to China's export boom. There are numerous studies on China's FDI boom, for instance, Lardy (1995), Henley et al. (1999), Cheng and Kwan (2000), Fung and Lau (2001) and Zhang (2001) identify potential market size, low labor cost, preferential policies (e.g., tax credits), openness, geographic proximity, good infrastructure and political stability as primary factors attracting FDI inflows to China; Branstetter and Feenstra (2002) show that FDI inflows reflect political openness and state ownership in China; Feenstra and Hanson (2003) examine the organization of export processing operations of foreign MNEs in China and test the property rights model. However, they abstract from analyzing the role of exchange rate fluctuations. In a recent paper, Greaney (2003) generates reverse imports in a model of trade and FDI networks. Ekholm et al. (2003) and Yeaple (2003) analyze the MNE's location choices in three-country models, but focusing on trade and transportation costs and production technology. In the present paper, we focus on the effects of exchange rate movements, based on the Sino-Japanese trade and FDI. Essentially, there are two channels through which currency devaluation impacts FDI inflows: a wealth effect and a relative production cost effect, both benefiting the foreign investor and leading to more FDI inflow. Theoretical models in this strand include Kohlhagen (1977), Cushman, 1985 and Cushman, 1988 and Froot and Stein (1991) and a few empirical studies provide evidence supporting the theoretical arguments, see for instance, Klein and Rosengren (1994), Goldberg and Kolstad (1995) and Kiyota and Urata (2004). Blonigen (1997) argues that Japanese FDI into the US during 1985–1990 were motivated by the desire to acquire the knowledge assets of US firms, in addition to the low value of the US dollar (see also Feenstra, 1999, for an excellent survey). As a complement to this literature, the present paper shows that reverse imports are another means through which exchange rate fluctuations can affect FDI flows. In addition, Zhao and Xing (2006) model the production allocation choices of a multinational enterprise in a three-country framework—a northern country and two southern countries. The paper finds that a currency appreciation in the southern country (X) producing the lowest-quality good with the lowest cost may reduce production (employment) in the north, while an appreciation in the other southern currency (Y) always raises production in the north. A northern depreciation against both southern currencies may increase production in country X, but always reduces that in country Y. While Zhao and Xing (2006) mainly analyze production shifting and competition between two southern countries, in the present paper we focus on whether exchange rate changes can benefit both a northern and a southern country under reverse imports. The rest of the paper is organized as follows. Section 2 presents some stylized facts on reverse imports, exchange rates and Japanese FDI in China; Section 3 sets up the basic model, Section 4 derives the equilibrium and its properties, Section 5 looks into the conditions for FDI and reverse imports to arise, Section 6 investigates the impacts of exchange rate changes on profits and welfare and Section 7 provides concluding remarks.

This paper assumed a structure in which a Japanese MNE produces in both Japan and China and engages in reverse imports from its Chinese affiliate, and a Chinese firm competes with the Japanese firm. We investigated the conditions for reverse imports to arise, and found that exchange rate changes, wage and capital cost differentials and barriers in brand name recognition contribute to increases in Japanese overseas production and reverse imports. The model could shed light on the current China debate. One of the most important markets for Japanese affiliates in China is Japan itself. Yen appreciation raises the relative production cost in Japan, driving production to low cost countries through FDI and importing back for serving the Japanese domestic market. The model showed that due to the barriers in brand name recognition, it is uncertain whether the exports of the Chinese firm rise or not when the yuan depreciates against the yen. It is likely that the growth of reverse imports erodes the market share of Chinese firm. In other words, Chinese firms could be a loser of yen appreciation, due to the strong competition of reverse imports by Japanese MNEs. The predictions of the model fit well with the actual data. Depending on product differentiability, yen appreciation may improve profits of the Japanese firm and welfare in Japan under reverse imports. One might argue that even though Japanese welfare increases under reverse imports, Japanese employment may fall because the MNE shifts production to China. We conjecture that the gain from reverse imports may be smaller taking consideration of unemployment assistance and training. We have assumed that firms compete in quantity with differentiated products. They could also compete in prices. It is well known that prices are lower and outputs higher under price competition than under quantity competition (see for instance, Cheng, 1985). Also, a quasi-linear utility function (which generates linear demand functions) has been assumed. More complicated demand and utility functions will give more variations. In addition, sales in China were assumed to be zero for simplicity. If we allow positive sales in China, as long as there is market segmentation, our analysis on reverse imports, FDI and the exchange rate remains robust. We have investigated the impact of exchange rate changes. The mechanism is the same if wages or other production costs increase in China. Further, commercial policy such as tariffs can be incorporated into the model. Increases in import tariffs cause similar effects as currency depreciation in Japan. In particular, in the presence of reverse imports, an increase in the import tariff might hurt domestic firms with overseas subsidiaries. This implies that the optimal tariff is lower than in the absence of reverse imports. Finally, production in the present model consists of one stage only. The analysis becomes more complicated in a structure with both intermediate and final productions. This remains a fruitful avenue for further research.