隨著浮動匯率的普遍採用，目前台幣主要釘住限元，因此雖然台幣業美元的匯率不 變，但只要其他幣別對美元之匯率變動，新台幣業外價值事實上也跟著變動，本文 主要目的在於利用各種模型計算台幣的有效匯率，並加以比較。 有夜率一般作為目標變據（例如貿易量）最後變動的部分指標，因此可用為本國產 品在世界市場競爭力的衡量，故本文交導入韓國有效匯率的計算以與台灣在此構而 競爭力的比較。本文架構為： 第一章、導論 第二章、有效匯率的理論模型 第一節、有效匯率的起源 第二節、理想權數的有效匯率模型 第三節、適合開發中國家或小國的有效匯率模型 第三章、台灣有效匯率的計算 第一節、雙邊貿易權數所計算出的EER及其解釋 第二節、多邊貿易權數所計算出的EER及其解釋 第三節、適合台灣模型所計算出的EER及與前兩節的EER之比較 第四章、韓國有效匯率之計算 第一節、適合韓國的模型所計算出的EER及其解釋 第二節、以有效匯率為構面，分析台、韓競爭能力 第五章、結論 #2810649

這篇論文主要探討的是現任央行總裁彭淮南上任後，央行對於匯率管理上的政策行為。 我們觀察從19991年到2009年這段期間的名目有效匯率發現，雖然匯率波動於一個區間之內， 但名目有效匯率在大部份的期間都是被低估的，尤其是近幾年來更是相當的偏低。 所以我們懷疑臺灣央行所管制的匯率目標區他們較希望能維持新臺幣在被低估的水準，當新臺幣匯率受到高估時，央行則希望能將其下修。 為了探討這個情況，我們透過Krugman(1991)所發表的匯率目標區模型來試著得到新臺幣的最適匯率行為。又因為臺灣的特殊經濟背景， 我們融合了Chen，Funke和Glanemann(2009)的寬鬆邊界匯率目標區模型和Torres(2000)的隨機邊界內的央行干預模式來修正傳統的Krugman模型。 然而，傳統的Krugman模型我們得到最重要的結論是，當匯率越接近上下邊界時會有越強烈的蜜月效果，也就是說匯率的波動越平緩。 但是當我們融入上面兩種修正在配合新臺幣的名目有效匯率實際資料時，我們發現蜜月效果僅僅存在於下界，就算是央行所保護的區間有變動時， 在下界還是相對於上界穩定相當多。這就是因為在下界存在著強烈的蜜月效果，所以匯率在越接近下界時會越穩定，甚至有很大機率維持在下界附近。 這個結論跟我們觀察實際資料所發現的現象是相當符合的，所以能夠解釋為什麼新臺幣的名目有效匯率會總是被低估。 / This paper discusses the policy rule used by Central Bank of the Republic of China(Taiwan) with target zone to the exchange rate dynamics. We focus on the recent phenomena on the exchange rate of NTD, and try to figure why the NEER of NTD is always underestimated. Due to the regime of Central Bank of the Republic of China(Taiwan), we combine two extensions into the basic Krugman(1991) target zone model which are Chen, Funke and Glanemann's soft edge target zone model(2009) and Torres's stochastic intra-marginal intervention pattern. And we estimate the parameters with simulated method of moments(SMM). By this two extensions, we conclude that there is strong honeymoon eect to the exchange rates at the lower bound, but the honeymoon eect is very weak at the upper bound. This conclusion is matched with the empirical data, and explains why the NEER of NTD is always underestimated.

本文採用1990年至2004年10月之月資料，建立人民幣出口、進口及雙邊貿易之三項實質有效匯率指數。考量中國大陸特殊政、經情勢，以及加入WTO後，進出口貿易更加自由化的前提下，運用中國大陸進出口值、物價水準，建立人民幣實質有效匯率指數，所依循理論較無疑慮，而援引之數據亦無太大爭議。 為貼近中共當局宣稱建立「符合市場經濟的靈活的匯率制度」，本文選擇與中國前十二大貿易夥伴組成一籃子貨幣，用以編纂人民幣實質有效匯率指數，將現行釘住美元之匯率改為釘住「一籃子」貨幣，能有效反映人民幣價值，亦可衡量中國外貿競爭力。 實證結果發現，相對於基期(2000年)，1990∼1993年指數多低於100，幣值高估；1994年匯率併軌，匯價發生結構性改變，幣值過度低估，後指數逐步下跌，1997年趨近均衡匯率。1997年亞洲金融風暴至2002年，甚至出現與均衡匯率並無太大偏離現象，即便有所失調亦能在短期內適切調整。2002年下半年至2004年年底，匯價低估趨勢確立，但偏離均衡匯率僅5％上下，幅度不太。 中國宣稱要「和平崛起」，自應承擔更多國際義務，人民幣升值是無可違逆的趨勢。為避免干擾經濟發展，勢必採行積極之管理浮動匯率制度，釘住一籃子貨幣並設定幅度狹窄之浮動區間「微調」。匯價調整時機，除衡酌自身經濟面的條件，尚須納入國際政治面的考量，由中共官方談話研判，人民幣升值，將是無預警的，出其不意的。 / In this paper monthly data from 1990 through October 2004 are used to establish real effective exchange rate indices of RMB for export, import and bilateral trading respectively. These real effective exchange rate indices are established taking the particular political and economic conditions in China into consideration, based on the increasingly liberation of importation and exportation after joining WTO, and using China’s import and export volume as well as its CPI. The theory used is doubtlessly correct and there is no much dispute on the data referred herein. To adhere with what the China government proclaimed: “a flexible exchange rate meeting the market economics”, currencies of its top 12 trade partners are selected for a basket of currencies instead of pegging to US dollar, in forming the real effective exchange rate indices. These indices can effectively reflect RMB’s true value, and measure China’s foreign trade competition. According to the result of verification, in comparing with the base period (2000), the indices for 1990 – 1993 were mostly less than 100, representing that RMB was overvalued. In 1994 the exchange rates were unified, resulting in a structural change on foreign exchange rate, RMB was undervalued. Then, these indices fell gradually, and the exchange rate tended to become balanced in 1997. From 1997, while the Asian financial crisis happened, till 2002, there was no much deviation from the balanced exchange rate, i.e., even there was any out of balance, it was adjusted properly within a relatively short time. From the second half of 2002 till the end of 2004, the tendency of undervaluation was ascertained, by the deviation was only about 5%, the range was not so much. Proclaiming that it is going to “peacefully rise”, China should assume more international liabilities, and the appreciation of RMB is a non-reversible trend. To avoid interference to its economic development, China has no choice but to adopt an aggressively control on its floating exchange rate regime, pegging to a basket of currencies and setting up a relatively narrow range of tunnel for “snaking”. In addition of its own political economy, international political situation must be taken into consideration for timing of its exchange rate adjustment. From some China government officials’ statements, it can be seen that appreciation of RMB would be done without any warning in advance and unexpectedly.

本文欲以平滑轉換自我迴歸模型(Smooth Transition Autoregressive Model，簡稱STAR模型)及時變平滑轉換自我迴歸模型(Time-varying Smooth Transition Autoregressive Model，簡稱TV-STAR模型)兩種非直線性模型為工具，剖析人民幣實質有效匯率之動態結構。 實證結果得知，在長期干預的情況下，人民幣實質有效匯率拒絕線性檢定且其為LSTAR模型，故可知人民幣實質有效匯率在轉換過程具有不對稱之特性；其次，利用“Specific-to-General-to-Specific”篩選過程得知，若是預測人民幣實質有效匯率，並不一定需要利用到比較複雜的TV-STAR模型；因為樣本外預測，短期間，一階自我迴歸模型的表現可能並不遜於複雜模型；長期而言，則似以STAR模型表現較佳。

90年代末東亞金融危機造成多國貨幣大幅貶值，銀行紛紛倒閉。基本上金融危機可分為通貨危機(Currency Crisis)與銀行危機(Banking Crisis)；通貨危機是指當年中任一季名目匯率貶值超過25%，且貶值幅度比前一季超過10個百分點。諸多實證文獻顯示，高估一國匯率為其通貨崩潰之先驅；又由於近年來新興國家快速開放資本市場，以致於成為危機之導火線。為分析此一現象，本文首先編製金融帳權數之新台幣實質有效匯率指數，並將東亞之台灣、印尼、韓國、菲律賓、泰國等五國之匯率、相對物價（各國與美國物價）、金融帳餘額等變數做共整合關係檢定，觀察三個變數的長期均衡關係，再將誤差項加入模型中，建構向量誤差模型。實證結果發現，金融帳與相對物價對匯率有顯著之影響力。 / The 1997 East Asian Crises had made exchange rate depreciations and bank bankruptcies. Broadly speaking, it can be divided into currency crisis and banking crisis. Nominal exchange rate of any season in a year, which is depreciated over 25% and 10% than last season, is called a currency crisis. Lots of papers demonstrate that overvaluation is a precursor of a currency crash. Furthermore, developing countries have opened capital markets so rapidly that it became the tinderbox of crises. To analyze the phenomenon, this thesis first compile Taiwan’s financial weighted real effective exchange rate index, then examine exchange rates, relative prices (compare to American consumer price index), and net financial account of Taiwan, Indonesia, Korea, Philippine, and Thailand with cointegrated test to identify the long run equilibrium relationships between variables; then adding error terms into models to estimates vector error correction model (VECM). The empirical results show that financial account and relative price influence exchange rate significantly.