摘要
匯率走勢的預測與匯率波動的分析，始終是國際金融議題之一。本文區分兩部分，分別探討匯率走勢的預測與匯率波動的分析。 有關匯率走勢預測部分，兩種動態因子模型各別從大量的總體經濟變數中，萃取共同因子來幫助匯率預測並比較預測能力的表現。藉由動態因子模型的假設，比較古典因子模型(classical factor model)與 Pena-Box 模型的穩健性(robustness)。在模型錯置的條件下，發現 Pena-Box 模型比古典因子模型更穩健。這兩模型估計共同因子的方法不同，分別為主成分分析 (Principal Componemt Analysis; PCA) 與 2-SCM (two-directional Scalar Component Model)。PCA 因子是擷取同一時期變數間的關係，而 2-SCM 因子則是擷取跨期變數之間的關係。由於 2-SCM 因子受到時間因素的影響，因此，在時間序列上 Pena-Box 模型表現會比古典因子模型穩健。在實證上，本文藉由兩國基本面之時間序列資料分別粹取 PCA 因子與 2-SCM 因子，並進一步在匯率預測方程式上加入雙邊匯率與共同因子之遞延項。樣本內的預測，無論使用 2-SCM 因子或 PCA 因子的預測能力都比隨機漫步優越，而且使用 2-SCM 因子作為匯率預測的外生變數會比用 PCA 因子表現更佳。然而，在樣本外的預測，利用 2-SCM 因子幫助匯率預測，其預測表現無明顯優於隨機漫步。 有關匯率波動分析，藉由電子外匯交易系統 EBS 之日內高頻資料，記錄每 15 分鐘內 JPY/USD 匯率之買賣價差，資料長度由西元 2003 年 01 月 01 日至西元 2005 年 12 月 31 日止，檢視外匯市場交易機制的活動性與匯率波動的關係。利用買賣價差資料配適 Periodic-GARCH (P-GARCH) 模型，當匯率交易次數與價格改變次數，兩者與買賣價差有負向關係時，匯率的波動大小與買賣價差便有顯著且正向關係，此項結果與過去的實證研究相似。另外，也發現在東京外匯市場的交易期間，交易活動曲線會呈現 U 型現象；而倫敦外匯市場交易期間的交易活動曲線則為倒 U 字型。同時，買賣價差受未預期新聞發佈的影響，買賣價差的波動曲線呈現反 U 字型。由電子外匯交易系統 EBS 所取得的買賣價差，與其他電子外匯交易系統或交易商間之外匯交易市場的資料，買賣價差的活動行為無明顯的差異。
並列摘要
This dissertation investigates two issues in the international financial market. One is the forecasting of exchange rate movement, the other is the analysis of volatility exchange rate. To forecast the exchange rate movement, adopting dimension reduction to extract useful predictors from a large set of monthly macroeconomic time series is a way to improve forecasting accuracy. Part of studies compares two methods for extracting predictors, including the well-known classical factor model and the Pena-Box Model, which is a dynamic factor model. Compared with the classical factor model, the Pena-Box Model is more robust with respect to misidentifying models since it captures the time-effect relationship of original variables. Both simulations and empirical studies on forecasting the JPY/USD exchange rate confirm the advantages of the Pena-Box Model. To analyze the relation trading activities with exchange rate volatility on the intraday behavior of the 15-minute exchange rate bid-ask spreads in the EBS electronic brokerage market. Part of studies investigates the market institutional effects of the determining role of trading activities and exchange rate volatility on the intraday behavior of the 15-minute JPY/USD and EUR/USD exchange rate bid-ask spreads in the EBS electronic brokerage market for the period from January 1, 2003 to December 31, 2005. We find that the exchange rate volatility both significantly and positively affects the bid-ask spreads, while the numbers of deals and quotation changes negatively affect the bid-ask spreads. These results are similar to those of past studies. We also find that a U-shaped pattern exists in the Tokyo trading hours and an inverted U-shaped pattern in the London trading hours, in addition to a round clock inverted U-shaped pattern of spread volatility. This inverted U-shaped pattern may be caused by unexpected news arrivals. The spread behavior of the EBS global electronic broking market is not different from that of other electronic interdealer quotation markets and electronic broking markets.
參考文獻
延伸閱讀
- 林昆良（2007）。外匯市場技術分析之研究〔碩士論文，國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342/NTU.2007.03249
- （2010）。預測論壇：經濟預測的觀點、方法與表現。臺灣經濟預測與政策，41(1)，155-188。https://www.airitilibrary.com/Article/Detail?DocID=17298849-201010-201011170001-201011170001-155-188
- 姜淑美、鄭婉秀、邱建良（2003）。外資交易行為、股市及匯市動態關係之研究。風險管理學報，5(1)，45-64。https://doi.org/10.30003/JRM.200303.0003
- 吳依蒨（2012）。人民幣匯率之動態預測〔碩士論文，國立暨南國際大學〕。華藝線上圖書館。https://doi.org/10.6837/NCNU.2012.00256
- Chow, M. K. (1992). Technical analysis of the foreign exchange market [master's thesis, The University of Hong Kong]. Airiti Library. https://www.airitilibrary.com/Article/Detail?DocID=U0029-1812201200009708