本文試圖對美元兌南非幣匯率進行預測，並找出對於南非匯率有影響力之變數。研究期間為1996年至2021年，共26年期間之月資料。使用變數範圍涵蓋總體經濟變數、技術指標、商品價格與國際指數等領域共32種變數。本文以兩種模型預測次月的匯率，其一是逐步迴歸模型，另一個是機器學習中的XGBoost模型。由實證結果顯示，迴歸模型於研究期間之預測準確率為56.25%，年化報酬率達6%；而XGBoost模型預測準確率為67.71%，年化報酬率達20.07%。兩種模型選取的變數中，金銅比和標準普爾500指數同樣都是選取頻率最高的變數，顯示其對於南非匯率具有相當程度的影響力。雖然迴歸模型的預測準確率較低，不過由實證結果發現該模型於匯率波動較大的時段能保持不錯的預測能力，因此若能改善模型於匯率波動較小的時段之表現，迴歸模型仍可作為一種方便且容易使用的匯率預測模型。

This paper attempt to forecast the exchange rate of the US dollar against the South African Rand and try to find the determinant of the Rand. The study period is from 1996 to 2021 with monthly frequency data. The variables used in this paper including macroeconomic variables, technical analysis indexes, several commodity prices and global stock indexes. This paper uses two models to predict the exchange rate in the next month, stepwise regression model and XGBoost respectively. The empirical results show that the prediction accuracy of XGBoost is 67.71% and the annualized rate of return is 20.07% which are much better than stepwise regression model (with 56.25% prediction accuracy and 6% annualized rate of return). Gold to Copper Ratio and S&P 500 Index are the most frequently picked by both of models. It can be seen that the variables have considerable influence on the South African Rand.

論文審定書 .................................................................................................................... i
摘要 ................................................................................................................................ ii
Abstract ........................................................................................................................... iii
第一章 緒論 ................................................................................................................... 1
第一節 研究背景與動機 ...................................................................................... 1
第二節 研究目的 .................................................................................................. 3
第三節 研究對象與範圍 ...................................................................................... 4
第四節 研究流程 .................................................................................................. 5
第二章 文獻探討 .......................................................................................................... 7
第一節 匯率理論 .................................................................................................. 7
第二節 影響匯率變數 ........................................................................................ 10
第三節 機器學期相關文獻 ................................................................................ 15
第三章 研究方法 ........................................................................................................ 17
第一節 研究架構 ................................................................................................ 17
第二節 變數定義 ................................................................................................ 18
第三節 迴歸模型 ................................................................................................ 32
第四節 XGBoost演算法 ................................................................................... 36
第五節 超參數調整與模型表現評估 .............................................................. 40
第六節 交易策略 ............................................................................................... 43
第四章 實證結果 ........................................................................................................ 44
第一節 迴歸分析實證結果 ............................................................................... 44
第二節 XGboost模型實證結果 ....................................................................... 51
第三節 模型表現比較 ………........................................................................... 58
第五章 結論 ................................................................................................................ 60
第一節 結論 ........................................................................................................ 60
第二節 未來研究方向 ........................................................................................ 61
第六章 參考文獻 ........................................................................................................ 62

圖次
圖1.4.1研究流程圖....................................................................................................... 6
圖3.1.1研究架構......................................................................................................... 17
圖3.3.1迴歸模型資料滾動方式................................................................................ 35
圖3.4.1 XGBoost資料滾動方式................................................................................ 39
圖3.4.2交叉驗證方式................................................................................................. 40
圖3.6.1交易策略......................................................................................................... 43
圖4.1.1迴歸模型報酬表現......................................................................................... 49
圖4.1.2逐步迴歸選取變數......................................................................................... 50
圖4.2.1 XGBoost模型報酬表現................................................................................ 56
圖4.2.2 XGBoost模型特徵重要度............................................................................ 57

 
表次
表1.3.1 模型滾動方式................................................................................................. 4
表3.2.1 使用變數種類................................................................................................ 18
表3.3.1 變數處理方式................................................................................................ 32
表3.5.1 混淆矩陣........................................................................................................ 40
表4.1.1 迴歸模型預測結果........................................................................................ 44
表4.1.2 迴歸模型績效表現........................................................................................ 48
表4.2.1 XGBoost模型預測結果................................................................................ 51
表4.2.2 XGBoost模型績效表現................................................................................ 54
表4.3.1 模型表現比較................................................................................................ 58
表4.3.2 模型於不同漲跌幅度之表現....................................................................... 58
表4.3.3 模型選取變數比較........................................................................................ 59

一、中文文獻
1. 王界舜 (2010)。外匯技術指標獲利績效分析與預測。天主教輔仁大學經濟學研究所碩士論文。
2. 呂怡慧 (2021)。美國Nasdaq指數變動方向之預測。國立中山大學金融創新產業碩士專班碩士論文。
3. 徐維志 (2015)。以隨機森林為模式之美金/歐元匯率交易預測研究。天主教輔仁大學統計資訊學系應用統計碩士在職專班碩士論文。
4. 胡玉茹 (2015)。南非幣匯率變動因素之研究。國立中正大學經濟學系國際經濟學碩士論文。
5. 梁煜傑 (2020)。利用機器學習預測臺幣匯率。國立中央大學經濟學系碩士論文。
6. 黃建豪 (2014)。高息貨幣匯率與黃金期貨間之波動外溢現象及避險效果探討。國立中正大學財務金融研究所碩士論文。
7. 趙翊伶 (2010)。CRB商品指數與高息貨幣匯率之關係。國立中正大學財務金融研究所碩士論文。
8. 賓秀娟 (2015)。南非幣、澳幣、人民幣及新台幣之關連性研究。國立中正大學經濟學系國際經濟學碩士論文。
9. 葉松炫 (2000)。運用類神經網路預測匯率。國立中山大學財務管理學系研究所碩士論文。
10. 鄭嬿真 (2011)。南非之匯率實證研究。國立高雄應用科技大學金融資訊研究所碩士論文。
11. 簡任申 (2017)。類神經網路在股票預測之獲利可能性研究-以台灣50成分股為例。國立雲林科技大學財務金融系碩士論文。

二、英文文獻
1. Amano, R. and van Norden, S. (1998), Exchange rates and oil prices. Review of International Economics, 6(4), 683–694.
2. Aron, J., Elbadawi, I. and Kahn, B. (1997). Determinants of the Real Exchange Rate in South Africa. Centre for the Study of African Economies Working Paper Series, 1997-16.
3. Bloomberg, S.B. and Harris, E.S. (1995). The commodity–consumer price connection: fact or Fable? Economic Policy Review, 1(3), 21–38.
4. Coudert, V., Mignon, V. and Penot, A. (2007). Oil Price and the Dollar. Energy Studies Review, 15(2), 48-65.
5. Chen, T., and Guestrin, C. (2016). XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, 785-794.
6. Dibooglu, S. (1996). Real disturbances, relative prices, and purchasing power parity. Journal of Macroeconomics, 18(1), 69-87.
7. Dauvin M. (2014). Energy prices and the real exchange rate of commodity-exporting countries. International Economics, 137, 52–72.
8. Fama, E.F. (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal of Finance, 25(2), 383-417.
9. Frankel, J. (2007). On The Rand: Determinants Of The South African Exchange Rate. South African Journal of Economics, 75(3), 425-441.
10. Galeshchuk, S. (2016). Neural networks performance in exchange rate prediction. Neurocomputing, (172), 446–452.
11. LeBaron, B. (1999). Technical trading rule profitability and foreign exchange intervention. Journal of International Economics, 49(1), 125-143

12. Lui, Y.H. and Mole, D. (1998). The use of fundamental and technical analyses by foreign exchange dealers: Hong Kong evidence. Journal of International Money and Finance, 17(3), 535-545.
13. Liu, C., Hou, W. and Liu, D. (2017). Foreign Exchange Rates Forecasting with Convolutional Neural Network. Neural Process Lett, 46, 1095–1119.
14. Neely, C.J., & Weller, P.A. (2011). Technical Analysis in the Foreign Exchange Market. Federal Reserve Bank of St. Louis Working Paper Series, Working Paper 2011-001B
15. Ozkan F. (2012). A Comparison of the Monetary Model and Artificial Neural Networks in Exchange Rate Forecasting. Business and Economics Research Journal, 3(1), 1-27.
16. Sadorsky, P. (2000). The empirical relationship between energy futures prices and exchange rates. Energy Economics, 22(2), 253-266.
17. Saayman, A. (2007). The Real Equilibrium South African Rand/US Dollar Exchange Rate: A Comparison of Alternative Measures. International Advances in Economic Research, 13(6), 183-199.
18. Szakmary, A.C. and Mathur I. (1997). Central Bank Intervention and Trading Rule Profits in Foreign Exchange Markets. Journal of International Money and Finance, 16(4), 513-535.
19. Saacke, P. (2002). Technical Analysis and the Effectiveness of Central Bank Intervention. Journal of International Money and Finance, 21(4), 459-479.
20. Sermpinis, G., Stasinakis, C., Theofilatos, K. and Karathanasopoulos, A. (2015). Modeling, forecasting and trading the EUR exchange rates with hybrid rolling genetic algorithms support vector regression forecast combinations. European Journal of Operational Research, 247(3), 831–846.

21. Throop, A.W. (1993). A generalized uncovered interest parity model of exchange rates. Economic Review - Federal Reserve Bank of San Francisco 2.
22. Takaendesa , P. (2006). The Behaviour and Fundamental Determinants of The Real Exchange Rate in South Africa. Unpublished Master’s Thesis, Rhodes University.
23. Zhou, S. (1995). The Response of Real Exchange Rates to Various Economic Shocks. Southern Economic Journal, 61(4), 936–954.