總體資訊的發佈與匯率的關聯性：以新台幣兌美元為例

Macroeconomic News Announcement and Exchange Rate: Evidence in NTD/USD

林佳京(Chia-Ching Lin)

指導教授 ： 吳博欽

中原大學/商學院/國際貿易研究所/碩士(2009年)

https://doi.org/10.6840/cycu200900079

摘要

本文根據Ehrmann and Fratzsher(2005)所建立的實證模型並加以修改，探討美國工業生產指數、美國貿易餘額、台灣躉售物價指數、台灣貿易餘額的未預期部份，及台灣加權股價指數與央行干預如何影響新台幣/美元匯率，實證期間為1998年1月至2007年12月。實證上採用對稱波動模型(GARCH Model)及非對稱波動模型(Threshold GARCH、Exponential GARCH及Power GARCH Model)進行分析。
實證結果，可獲得下列的結論：
一、 就總體資訊發佈的未預期部份對匯率的影響而言，美國工業生產指數的上升會導致美元貶值。此符合凱因斯學派的觀點，由於工業生產指數某種程度代表所得，若所得增加，進口會增加，本國貨幣會貶值。其次，美國貿易餘額的改善有利於美元升值，而台灣躉售物價指數的上升會使新台幣貶值。最後，台灣貿易餘額對匯率的影響符號為正向顯著，與預期的不一致，可能的原因為台灣近年貿易順差萎縮及央行干預所造成。
二、 台灣加權股價指數上升會使新台幣升值，此符合Solnik(1987)國內股價上漲對匯率產生長期負面效果的觀點。央行干預對匯率的影響為負向，顯示央行為減低因匯率變動所造成不利於經濟發展的影響，會在外匯市場進行干預使匯率變動方向縮小。
三、 不對稱效果係數雖然符合預期方向但皆不顯著，可能的原因為這段期間政府積極干預外匯市場。當好、壞消息的發佈使新台幣/美元波動更大時，進出口商的利潤波動則增加，此時其投資不確定也增加，最後勢必影響一國的經濟成長，政府為避免這種情況會適度地進場干預，以降低匯率波動幅度過大而造成不利於經濟發展的影響。故匯率波動對好、壞消息反應的差異程度就不那麼明顯。

關鍵字

未預期部份 ； 總體資訊發佈 ； 匯率 ； 非對稱波動模型 ； 對稱波動模型

並列摘要

In this paper, we follow Ehrmann and Fratzsher(2005) framework and modify it to analyze the impacts of unexpected component of U.S. industrial production index, U.S. trade balance, Taiwan wholesale price index, Taiwan trade balance, plus the Taiwan stock index and central bank intervention on NTD/US exchange rate. The empirical period is from January 1998 to December 2007; in empirical estimation, we utilize the symmetric volatility model (GARCH Model) and asymmetric volatility models (TGARCH, EGARCH and PGARCH Model) to evaluate.
The empirical results are as follows:
1. As far as the impacts of unexpected component of macroeconomic news announcement on the exchange rate are concerned, a rise in U.S. industrial production index will lead to devaluate the US dollar. This corresponds with Keynesian school’s viewpoint. Because the industrial production index represents incomes to some extent, if incomes increase, imports can raise. Consequently, the domestic currency will depreciate. In addition, an improvement in U.S. trade balance will bring about an appreciation of the US dollar; however, an increase in Taiwan wholesale price index will cause a depreciation of NTD. Finally, the sign of Taiwan trade balance on the exchange rate is positive significantly, which is inconsistent with expectation. The reasons may be the decrease of Taiwan trade surplus in recent years and central bank intervention.
2. A rise in the Taiwan stock index will result in an appreciation of NTD, which is consistent with Solnik’s (1987) viewpoint: an increase in the domestic stock index will generate a negative effect on the exchange rate in the long term. The effect of central bank intervention on the exchange rate is negative, which shows that the central bank will adopt intervention tool in the foreign exchange market to reduce the fluctuation of the exchange rate that generates unfavorable effects on economic development.
3. Although the asymmetry coefficient matches the expected direction but not significant, it may be that the government intervened the foreign exchange market aggressively during this period. When good or bad news announcement makes NTD/US exchange rate more volatile, profit volatility of impostors and exporters increases; meanwhile their investment uncertainty will also increase, and finally it will influence the economic growth. The government will intervene the market appropriately based on the above reasons to reduce the unfavorable impacts by the volatile foreign exchange market. Therefore, the different extent of exchange rate reaction to good or bad news will be not so obvious.

並列關鍵字

symmetric volatility model ； asymmetric volatility model ； unexpected component ； exchange rate ； macroeconomic news announcement

參考文獻

楊凱文(2004)，「貨幣學派均衡匯率偏離之非線性動態調整」，中原大學國際貿易研究所碩士論文。

連伯瑋(2007)，「歐元匯率與總體經濟指標關係之實證研究」，台灣大學財務金融研究所碩士論文。

Andersen, T., T. Bollerslev , F. Diebold, and C.Vega (2003), “Micro effects of macro announcements: real-time price discovery in foreign exchange,” American Economic Review , 391, 38–62.

Black, F. (1976), “Studies of stock, price volatility change,” Proceedings of the American Statistical Association: Business and Economic Statistics Section, 177-181.

Bollerslev, T. (1986), “Generalized autoregressive conditional heteroscedasticity,” Journal of Econometrics, 31, 307-327.