新台幣與亞洲主要貨幣兌換美元匯率波動關聯性之分析

An Analysis on the Fluctuation Relationship of New Taiwan Dollar and Major Asian Currencies Against US Dollar

江金全(Jin-Chyuan Jiang)

指導教授 ： 陳達新

國立臺北大學/商學院/國際財務金融碩士在職專班/碩士(2009年)

摘要

迎接地球村來臨，外匯市場走向國際化及全球化，因此國際金融市場之間彼此有著高度的互動性與整合性，所以單一匯率已經無法完全隔絕與其他各國匯率變動，故本研究將針對新台幣與亞洲主要強勢貨幣間之匯率變動，利用E-View計量工具，分析彼此互動關係，做為全球化佈局的企業及投資人規避匯率風險之參考。本研究取樣期間自2004年1月1日至2008年12月31日之新台幣、日圓、人民幣及韓圜兌換美元匯率之日收盤價資料，樣本數共1,190筆。運用此四種貨幣兌換美元匯率之日變動率資料探討相互間匯率變動是否存在因果關係？彼此間又有何種因果關係？及彼此間之衝擊影響，何者自身解釋度較強？或何者被其他貨幣匯率變動解釋度較高？經由實證研究結果發現如下：一.新台幣與亞洲主要貨幣匯率變動率之Granger因果關係檢定新台幣匯率變動僅具有單向領先與日圓匯率變動關係；新台幣匯率變動僅具有單向領先與人民幣匯率變動關係；新台幣匯率變動與韓圜匯率變動間、日圓匯率變動與韓圜匯率變動間，具有雙向因果關係；日圓匯率變動與人民幣匯率變動間、人民幣匯率變動與韓圜匯率變動間，不具有任何因果關係。二.新台幣與亞洲主要貨幣之外匯市場衝擊反應分析 日圓、新台幣和人民幣匯率變動率三變數在衝擊反應上皆具有效性，都是在接受資訊時當日反應最大，接下來再收斂為零，且對自身的衝擊效果仍較其他變數明顯；而韓圜匯率變動率方面，在接受資訊時當日反應亦最大，惟未逐漸收斂為零，顯示其非完全有效性。在長期累積效果上，新台幣的衝擊反應會對其本身、日圓及人民幣產生正向的累積效果，日圓的衝擊反應會對其本身、新台幣及人民幣產生正向的累積效果，人民幣的衝擊反應會對其本身及新台幣產生正向的累積效果，韓幣的衝擊反應則會對其本身及新台幣產生正向的累積效果外，其他貨幣衝擊反應在累積效果方面均呈負向的累積效果。三.新台幣與亞洲主要貨幣之預測誤差變異數分解新台幣、日圓及人民幣等三種貨幣其自身解釋的程度都相當高(皆高於96%)，表示均不易由外在因素來加以解釋及影響，顯示此三種貨幣之外生性都很強。在變數間的相互影響及關聯性，新台幣匯率變動對韓圜匯率變動率的解釋比例達23.61%，因此顯示新台幣對韓圜有單向的影響。

關鍵字

單根檢定、因果關係檢定、向量自我迴歸模型、衝擊反應、誤差變異數分解

並列摘要

The advent of the global village has lead the foreign exchange market towards internationalization and globalization, and therefore, there exists a high degree of interaction and integration among international financial markets. Single currency can no longer be completely isolated from other countries or exchange-rate fluctuation. This study, by using E-Veiw tool, analyzes the correlation between New Taiwanese Dollar and exchange-rate fluctuation of Asian currencies, to provide currency risk-aversion reference for multi-national corporations and investors.

This study takes daily closing price for the exchange rates of New Taiwanese Dollar (NTD), Japanese Yen, RMB and Korean Won against U.S. Dollar as samples. There are a total of 1,190 pieces of samples in a sampling period from 2004/1/1 to 2008/12/31. This study uses daily data on the exchange-rate of these four currencies against U.S. Dollar, to discuss questions such as followings: whether there exist causal relationships between exchange-rate fluctuations? If so, what kind of causal relationship exists? What relationship demonstrates stronger self-explanation under mutual impacts? Or which currency demonstrates higher degree of explanation by fluctuation in other currency exchange-rate?
Results found through empirical research are as follows:
1. Granger causal relationship test of NTD and major Asian currencies rates of exchange rate movements only has one-way relationship and relationship with Japanese Yen currency exchange-rate; there are two-way causal relationship between NTD-Won currency exchange-rate and Yen-Won currency exchange-rate; there does not exist any relationship between Yen-RMB currency exchange-rate or RMB-Won currency exchange-rate.
2. Foreign exchange market’s impact response analysis on NTD and major Asian currencies
Three factors of Japanese Yen, RMB and NTD currency exchange-rates all demonstrates effectiveness and all show highest level of response on the day of data reception, gradually turn into zero and still demonstrate more profound self-impact than other factors. In the aspect of the fluctuation rate in Korean Won currency exchange rate also demonstrates highest response, but does not gradually become zero, which does not represent full-effectiveness. In long-term accumulated effects, impact response of NTD demonstrates positive accumulated response towards itself, Japanese Yen and RMB. Impact response of Japanese Yen demonstrates positive accumulated response towards itself, NTD and RMB. Impact response of RMB demonstrates positive accumulated result towards itself and NTD. Impact response of Korean Won, besides demonstrating positive accumulated results, also demonstrates negative accumulated results towards other currency impacts.
3. The decomposition of predicted variables among New Taiwan Dollar and major Asian
currencies.For New Taiwan Dollar, Japanese Yen and RMB, the self-explained level for these three currencies are exceptionally high (all above 96%). This shows that the erogeneity of these three currencies is tremendously strong. As for the correlation and connection among variables, the explained percentage of the currency fluctuation in New Taiwan Dollar to South Korean Won reaches 23.61%. Therefore, it illustrates that New Taiwan Dollar has single direction effect to South Korean Won.

並列關鍵字

Panel Unit Root Test、Granger Causality Test、Vector Autoregression Model、

參考文獻

1.林俊宏（1997），「外匯匯率預測-不同模型之比較」，國立成功大學企業管理研究所碩士論文。

2.Bilson, J.F.O.(1981), The Speculative Efficiency Hypothesis, Journal of Business, vol. 3, pp.435-451.

3.Bollerslev, T.(1986), Generalized Autoregressive Conditional Heteroskedasticity, Journal of Econometrics, vol. 31, pp.307-327.

4. Ballie, R. T., Bollerslev, T. (1989), The message in daily foreign exchange rates: A conditional-variance tale, Journal of Business and Economic Statistics, vol. 7, pp.297-305.

5.Campa, J.M., Chang, K.(1995), Testing the expectations hypothesis on the term structure of volatilities in foreign exchange options, Journal of Finance, vol. 50, pp. 529-547.