本篇論文試圖放寬Krugman (1991)、Bertola and Caballero(1992)、和胡春田及林鴻裕(1997)的模型假設並將其一般化。我們透過改寫Mundell-Fleming的開放總體架構(IS-LM-BP model),並再考量資本的移動程度後,藉此模型討論資本移動程度對於匯率波動以及蜜月效果的存續有何影響。 本文可以得出三個主要結論: (1)當民眾完全信任央行會監守匯率目標區政策時,蜜月效果將會存在,且當資本的移動程度越大,則蜜月效果將會越小;反之亦然。 (2)當民眾認為央行很有可能僅重整匯率目標區的中心點時(即 q > 1/2 ),蜜月效果將不會存在,且當資本的移動程度越大,匯率的波動也會更形劇烈。反之,倘若民眾認為央行重整中心點的可能性不大時(即 q < 1/2 ),則蜜月效果將會存在,但是無法確定資本移動程度與蜜月效果的關係。 (3)當央行可能同時採行放寬目標區以及在寬幅不變下、進行重整的匯率政策時,若央行干預市場基要的程度不大時(p < 1),則體系存在蜜月效果;反之,若央行干預市場基要的程度很大時(p > 1),則蜜月效果將不存在。
In this thesis, we tried to generalize the model provided by Krugman (1991), Bertola and Caballero(1992), and Hu and Lin(1997). In order to discuss the effect of the capital mobility on the exchange rate volatility, we employed an open macroeconomic model a la Mundell-Fleming (IS-LM-BP) model. Besides, the sensitivity of capital mobility on honeymoon effect was also discussed in the thesis. Three major theoretical findings of this thesis are as follows: (1) The result shows that there is still a honeymoon effect in a target zone system with imperfect capital mobility. Furthermore, we can also find that the honeymoon effect and the degree of the capital mobility change in reverse direction. (2) If the central bank realigns the central rate of the exchange rate target zones, there will exist the honeymoon effect when q<1/2 . On the contrary, there will not exist the honeymoon effect when q>1/2 , furthermore, the honeymoon effect and the degree of the capital mobility change in the same direction at the same time. (3) If the central bank widens the band width of the exchange rate target zones, there will exist the honeymoon effect when p<1 . On the contrary, there will not exist the honeymoon effect when p>1 .