台灣自1952年開始引進外人直接投資，成為台灣整體經濟活動中重要的一環，而觀察僑外來台投資的趨勢，發現在1980年帶短時間內有巨幅的增長，因此本研究欲探究在此時期影響來台投資的主要原因為何。且由於台灣自1979年起開始為機動匯率時期，因此本研究將針對實質匯率對直接投資的影響作一深入的探討。 本文以 Chen et al. (2006) 的模型為基礎，檢驗選取之九產業1982-2008年間的長期追蹤資料，結合產業來台直接投資動機間的差異，深入瞭解實質匯率是否對其投資決策行為有不同的影響，並以最大概似估計法 (Maximum Likelihood Estimator, MLE) 估計的 Tobit 迴歸分析模型做為分析工具。 實證結果顯示，當來台投資產業為市場導向時，台幣實質匯率的升值與預期其升值，都將使廠商增加來台投資；但預期台幣實質匯率的升值也會使得成本導向的廠商增加來台投資量。台灣相對來台投資國的工資率提高，將阻卻來台投資的決策行為。在台產業的市場大小，也會影響廠商的投資意願，當產業在台灣的市場較大時，會提高僑外資投資的意願。而匯率波動所造成的不確定性，可能由於以來台直接投資替代出口商品到台灣，使來台直接投資量上升。

中國經濟的崛起，改寫了世界貿易的版圖。但近年來中國的貿易收支失衡問題，特別是中國對於美國的持續大幅貿易順差以及對日本的持續大幅貿易逆差，一方面導致美國對中國匯率政策的強烈批評，另一方面也開始引起學者對於中國能否透過人民幣匯率的調整解決其貿易收支問題的研究興趣。惟最近一些實證研究發現，由於全球生產網路之發展，導致實質匯率與中國對外貿易之關係中出現一些傳統貿易理論難以解釋之現象，對於中國能否透過匯率政策之調整以解決其貿易收支問題也仍未有一致的結論。 本論文之目的在於研究中國近年來對外貿易之主要決定因素，尤其是深入探討實質匯率以及外人直接投資在中國對外貿易發展中所扮演的角色。由於美國以及日本為中國最主要的貿易夥伴，本論文的研究內容聚焦於中美與中日貿易決定因素之比較研究。本論文之主要創新在於實證模型之建構。本論文的實證模型與過去文獻的主要差異有數點：1. 本論文建立聯立方程體系以分析出口供需之相互影響。2. 在供給面本論文考慮了外人直接投資以及進口中間財對中國出口供給或其進口產業之國內生產的可能影響。3. 在進出口需求函數中，除了雙邊實質匯率，本論文也考慮了第三國匯率以及外人直接投資可能產生的間接影響。4. 在中國進口需求函數中本論文將實質所得拆解成國內需求與加工出口兩部分，以比較中國國內需求與加工出口在進口傾向上的差異。 本論文實證結果顯示，在實質匯率的影響方面，當人民幣對美元實質升值，確實能降低中國對美出口。惟因為出口下降導致加工出口所需之進口中間財與資本財減少，造成中國自美之進口金額不增反降。由於進口減少的幅度更甚於出口，因此人民幣升值對於降低中美貿易順差的效果難以顯現。此外，當人民幣對日圓實質貶值時，雖然刺激中國對日本的出口，但因應加工出口所需而進口的中間財與資本財將因此增加，進而使中國自日本的總進口上揚，如此一來將抵銷匯率貶值改善貿易收支逆差的效果。至於第三國匯率的波動，對中美或中日貿易收支也有一些間接影響，例如經由影響中國自這些地區進口中間投入之成本或透過與中國在美日最終財市場的競爭或互補關係。由此可知，在全球生產網路盛行的今天，中國似乎不易單獨透過匯率政策之調整以解決其貿易收支失衡的問題。 在外人直接投資的影響方面，本論文實證結果顯示，日本外資企業在中國的投資，一方面雖提升中國對日本的出口，但另一方面因為在中國生產外資企業需要大量進口零組件與加工產品，造成中國來自日本的進口大幅增加，結果造成中國對日本的貿易逆差情況雪上加霜。相反地，美國對中國的直接投資增加對中國對美出口並未有所助益，反而有提升自美進口的作用，有助於改善中美貿易失衡。但是，來自其他國家企業在中國的投資，一方面將加工生產完成的產品大量輸往美國，另一方面，這些企業在中國生產應付當地市場的產品取代來自美國的進口產品，導致中國對美國貿易順差的擴大。由此可見，外人直接投資在中國對外貿易收支失衡問題上亦扮演十分重要的角色。 至於實質所得對進口的影響方面，本論文實證結果顯示，中國的國內需求提升雖然會誘使來自美國或日本的進口增加，但是相對幅度有限，而因中國的加工出口所引發的進口增加卻十分顯著。此外，美國實質所得提高而誘發美國自中國進口的彈性相當可觀，日本實質所得提高而誘發日本自中國進口的彈性則相對有限，而中國國內需求提升對自美進口的效果更是微不足道。此結果顯示，中國與其主要貿易夥伴在消費者偏好上的差異，也對中國對外貿易收支失衡有一定的影響力。 / The emergence of China’s economy has changed territorial boundaries on the global trade map. China’s recent external trade imbalance, especially its huge and persistent trade surplus with the U.S. and its deficits towards Japan, has provoked serious critique from the U.S. on China’s exchange rate policy and at the same time has attracted research interest in the issue: Can a Renminbi (RMB) revaluation improve China’s problem over its trade imbalance? Because of the development of the global production network in recent years, some studies have found some puzzles in the relation between the RMB’s real exchange rate and China’s foreign trade, which are inexplicable from the viewpoint of traditional trade theory. There is no consensus on the effectiveness of a RMB revaluation in adjusting China’s trade imbalance either. The objective of this thesis is to study the determinants of China’s external trade, particularly the roles played by the real exchange rate of the RMB and foreign direct investment (FDI) into China. Because U.S. and Japan are two major trade partners of China, this dissertation focuses on comparing the determinants of Sino-U.S. trade and Sino-Japan trade. The main contribution of this dissertation is its construction of novel empirical models, which differ from most previous studies in the following aspects: 1. We construct a system of equations to examine the demand-supply relationship; 2. On the supply side, the effects of FDI and intermediate imports on China’s export are considered; 3. On the demand side, in additional to bilateral real exchange rates, the indirect effects of a third-country’s exchange rates and FDI on China’s exports and imports are also incorporated into the models; 4. In China’s import demand function, we decompose real income into domestic demand and export demand, with a view to comparing the effects of China’s domestic demand and export processing due to their differences in import propensities. This thesis’ empirical results indicate that the real appreciation of the RMB against the US dollar can reduce China’s exports to the U.S., however, the decrease in exports causes a larger decline in China’s imports of intermediate goods from the U.S. for export processing purposes. Hence, the real appreciation of the RMB seems ineffective in improving China-U.S. trade balance. Moreover, the real depreciation of the RMB against the Yen has similar counterbalancing effects: it stimulates China’s exports to Japan and at the same time increases China’s intermediate goods and capital goods imports from there. As a result, China’s trade deficits might not be improved significantly from the real depreciation of its currency. The empirical evidence of this thesis reveals that changes in third-country’s exchange rates can also bring about indirect effects on China’s trade imbalance either through their impact on the import costs of intermediate goods from those countries or their substitutive or complementary relationships in the final good markets in the U.S. and Japan. Therefore, China seems unable to adjust its trade imbalance unilaterally through an exchange rate policy. As for the impact of FDI into China, we show that Japanese FDI towards China stimulates China’s exports to Japan, but also raises its import demand from Japan, because of the processing production of Japanese enterprises in China. As a result, Japan’s FDI makes the Sino-Japan trade imbalance even worse. On the contrary, FDI from the U.S. into China does not stimulate American exports to China, but does increase China’s imports from the U.S., which can improve Sino-U.S. trade balance. Third-country FDI into China significantly enhances China’s exports to the U.S. on the one hand, while its products become substitutes for U.S. exports into China’s market on the other. Hence, it is another driving force for the aggravation of Sino-U.S. trade surplus. In a nutshell, FDI has played a very important role in China’s trade imbalance problem. Finally, empirical results from the effects of real income on import demand indicate that China’s import demand stimulated by its export processing activity is very significant. In addition, while the relationship between China’s domestic demand and its imports from the U.S. or Japan is positive, the magnitude is rather limited. In contrast, the income elasticity of U.S. import demand from China is very high, but the income elasticity of Japan’s import demand from China is relatively small. To sum up, China’s trade imbalance problem seems attributable to the differences in preference among consumers in China and its major trade partners to some extent.

本篇論文以 Granger and Teräsvirta (1993)所提出的非線性STAR模型，來進行東亞四個國家之實質有效匯率的實證研究。實證對象為中國大陸、日本、南韓與台灣四個國家，研究期間為1997年1月至2007年12月，樣本外資料期間為2008年1月至2009年12月。 實證結果發現： 1. 四個國家的實質有效匯率均拒絕線性假設，表示四個國家的匯率資料皆可以配適非線性STAR模型。 2. 模型決策檢定結果顯示，四個國家的實質有效匯率資料皆適合使用轉換函數為Logistic形式的STAR模型，即LSTAR。 3. 比較非線性STAR模型與傳統線性AR模型的預測能力，四個國家在STAR模型的預測能力表現較佳。然而AR模型與STAR模型的預測情況皆不甚理想。

中國在世界貿易舞台上所扮演的角色日益重要，人民幣的均衡水準成為十分熱門的議題。本文為貿易財與非貿易財部門的相對生產力找尋新的代理變數來檢視Balassa-Samuelson效果，以解決中國物價指數因受政府控制而難以反映實際生產力的問題。此外，亦將對人民幣匯率的預期引進實證模型中，以量化分析來探討人民幣是否具有自我實現的行為。本文實證結果顯示人民幣均衡匯率與相對生產力、貿易開放程度、熱錢數目間有長期關係。貿易財與非貿易財相對生產力與熱錢數目的上升可能帶動人民幣升值；貿易開放程度則與均衡匯率呈反向關係。 / As China plays an important role in the world economy, the research on the real exchange rate in China has received a lot of attention. This study chiefly contributes to adopt a more reliable proxy different from previous studies to capture the Balassa-Samuelson effect; in addition, we formerly introduce the expectation of Renminbi into the empirical model. This paper empirically investigates the behavior of the real exchange rate in China during the period from 1980 to 2005. Particularly, we provide a single-equation in an empirical framework which takes the specific features and characteristics in China economy into account. The empirical evidence shows that the degree of openness, the volume of hot money, and the relative productivity of traded to nontraded sectors are key factors to the determination of the long-run equilibrium real exchange rate in China. Both the increases in the relative productivity and an anticipated appreciation in the real exchange rate eventually may lead to a real appreciation in the currency. In contrast, the influence for opening up to the world economy is that the real exchange rate may depreciate.

China has been in a state that the currency appreciation is needed to restore the external balance. However, it appears that Chinese government worries about that the Renminbi appreciation may decrease the output. This purpose of this paper is to empirically investigate whether the contractionary hypothesis prevailing in developing countries holds for China with quarterly data over the period from 1995Q1 to 2006Q2. We apply VAR models by means of impulse response functions and variance decompositions. The empirical evidences indicate that even taking the spurious correlation into account, the real appreciation of Renminbi leads to a fall in China’s output. Thus, our findings do not support the contractionary devaluation hypothesis. Moreover, the impact of the exchange rate on output is not through the inflation rate. In the short run, real exchange rate shocks have much power in explaining the output’s variation while the U.S interest rate and government spending shocks are determinants to the variation in output in the long run. Particularly, the monetary policy has relatively weak effect on the output and the real exchange rate.