台灣地區為一小型開放經濟,資本移動程度頗高,在影響匯率的市場基要中,僅單純考量貨幣面與貿易面,已無法正確反映出匯率的波動,倘能再加上不確定性之金融面訊息如公眾預期心理所帶來的衝擊,應更有助於匯率預測效率之提昇。另外,台灣屬海島型經濟,為緩和匯率過度波動影響進出口,央行通常會對匯市採行干預措施,致使匯率實際值並非為市場機能所決定的均衡值。
在各種匯率預測理論的研究中,當貨幣供給額、利率、國際收支及躉售物價指數差距等一般性變數,對匯率波動的效果已被探討的同時,至於公眾預期心裡對匯率走勢的影響,則少被國內文獻所提及。有關這方面認知有限的原因之一,是缺乏將公眾預期心理予以量化的評估技術。
本研究選取日圓匯率、韓圜匯率及歐元匯率作為公眾預期心理之量化指標,並應用多變量ARIMA轉移函數模式對三者在定態性檢定、共整合檢定、交叉相關函數、參數估計、殘差值及AIC值檢定等方面做比較分析,檢測出韓圜匯率為預測新台幣匯率之公眾預期心理的最佳量化指標,並藉以建立新台幣匯率之預測模式。此一發現亟盼能補強先前貨幣學派結構化模型(Monetary Structural Model)與隨機漫步模式(Random Walk Model)未能納入公眾預期心理致引發對匯率預測能力不足之爭議。
Taiwan is a typical small, open economy with a highly degree of free capital flow. Among the key market fundamentals of market for foreign-currency exchange, it is not enough to reflect the accuracy of exchange rate by just reviewing the factors on the monetary and trade approach. It could be much helpful to improve the accuracy of exchange rate forecasting if uncertain impact from the financial market such as individual’s expectation behavior could be included. In addition, Taiwan is a so-called “Island-Economy”. In order to slow down the impact of rapidly fluctuations of exchange rate on the import and export. Central Bank of Taiwan used to have an active policy to stabilize the market for foreign-currency exchange, so that there might be difference between actual exchange rate and equilibrium exchange rate.
While the effects of several general variables such as money supply, interest rate, balance of payments and wholesale price index etc.. On the fluctuations of the exchange rate have been well documented, less is known about the effects of individual expectation behavior onto the determination of exchange rate. One reason for this limited knowledge has been the lack of a measurement technique for quantifying this variable.
In this report, we try to quantify this variable and adopt MARIMA transfer function model for an empirical research. It would be helpful to lengthen the shortness of exchange rat forecasting efficiency by using monetary structural model and random walk model. I am convinced that this particular portion of the paper could be thought of as seeking to expand the analytical inadequacy in prior related researches and make a contribution to new insights.