本文探討全球風險與金融市場狀態是否影響外匯因子報酬之可預測性，並比較不同模型架構在匯率報酬樣本外預測上的表現。研究樣本涵蓋2010年1月至2024年12月，以加幣、瑞士法郎、歐元、英鎊及日圓相對美元之匯率報酬為研究對象。本文建構利差、動能與價值三類外匯因子，並納入恐慌指數、芝加哥聯儲經調整全國金融狀況指數、信用利差與期限利差等全球狀態變數，以檢驗外匯因子的預測效果是否具有狀態依賴性。
實證方法以歷史均值模型作為主要對照基準，並依序比較普通最小平方法、脊迴歸、最小絕對收縮和選擇算子迴歸、交互項模型、隨機森林、梯度提升、前饋式神經網路，以及月頻與週頻長短期記憶模型。本文採用擴展視窗進行樣本外預測，並以平均平方預測誤差、均方根誤差、平均絕對誤差、方向準確率及損失差異檢定評估模型表現；此外，亦透過依預測報酬排序所建立之長短部位投資組合，檢驗模型預測是否能轉化為經濟績效。
實證結果顯示，普通最小平方法與正則化模型整體上未能穩定優於歷史均值模型。部分外匯因子與全球狀態變數的交互項在樣本內具有統計顯著性，顯示外匯因子的預測效果可能隨金融條件與市場風險狀態而改變；然而，加入交互項後並未帶來穩定的樣本外預測改善。非序列機器學習模型與月頻長短期記憶模型亦未在整體樣本中展現顯著優勢。週頻長短期記憶模型的敏感度分析則顯示，52週回看期間在不同設定中具有相對較佳的預測表現，但其整體預測誤差仍高於歷史均值模型。投資組合回測結果亦顯示，多數模型無法產生優於對照模型的風險調整後報酬。
綜合而言，外匯因子報酬具有經濟理論基礎與狀態依賴特徵，但相關訊號仍不足以穩定轉化為樣本外預測優勢與可持續的投資績效。本文結果呼應匯率預測文獻所強調之高雜訊、結構不穩定與低訊號特性，並顯示模型複雜度的提高不必然改善外匯報酬預測。全球風險與金融條件變數雖未能提供穩定的精確預測，仍可作為理解外匯因子脆弱性與調整風險曝險的重要參考。

This study examines whether global risk and financial conditions affect the predictability of foreign exchange factor returns. The sample covers January 2010 to December 2024 and includes the Canadian dollar, Swiss franc, euro, British pound, and Japanese yen against the U.S. dollar. Carry, momentum, and value factors are combined with the Volatility Index, the Adjusted National Financial Conditions Index, the credit spread, and the term spread.
Using the historical mean model as the benchmark, this study compares ordinary least squares, ridge and Lasso regressions, interaction models, random forest, gradient boosting, feedforward neural networks, and monthly and weekly long short-term memory models. Forecasts are generated through an expanding-window procedure and evaluated using forecast errors, directional accuracy, loss-differential tests, and long-short portfolio performance.
The results show that the linear and regularization models do not consistently outperform the historical mean benchmark. Although several interaction terms are statistically significant in-sample, they do not provide stable out-of-sample improvements. The machine learning models and the monthly long short-term memory model also fail to demonstrate a significant overall advantage. Among the weekly specifications, the 52-week lookback window performs relatively better, but its forecasting error remains higher than that of the benchmark. Portfolio results likewise show no persistent improvement in risk-adjusted returns.
Overall, foreign exchange factor returns exhibit state-dependent characteristics, but the predictive signals are too weak and unstable to generate consistent forecasting or investment gains.

Abankwa, S., & Blenman, L. P. (2015). FX liquidity risk and carry trade returns [Working paper]. SSRN. https://doi.org/10.2139/ssrn.2662955
Ayitey Junior, M., Appiahene, P., & Appiah, O. (2022). Forex market forecasting with two-layer stacked long short-term memory neural network (LSTM) and correlation analysis. Journal of Electrical Systems and Information Technology, 9, Article 14. https://doi.org/10.1186/s43067-022-00054-1
Banti, C., Phylaktis, K., & Sarno, L. (2012). Global liquidity risk in the foreign exchange market. Journal of International Money and Finance, 31(2), 267–291. https://doi.org/10.1016/j.jimonfin.2011.11.010
Botha, I., & Pretorius, M. (2009). Forecasting the exchange rate in South Africa: A comparative analysis challenging the random walk model. African Journal of Business Management, 3(9), 486–494. https://doi.org/10.5897/AJBM09.260
Brunnermeier, M. K., Nagel, S., & Pedersen, L. H. (2008). Carry trades and currency crashes. NBER Macroeconomics Annual, 23, 313–347. https://doi.org/10.1086/593088
Bunčák, T. (2016). Exchange rates forecasting: Can jump models combined with macroeconomic fundamentals help? Prague Economic Papers, 25(5), 527–546. https://doi.org/10.18267/j.pep.581
Cheung, Y.-W., Chinn, M. D., & Garcia Pascual, A. (2005). Empirical exchange rate models of the nineties: Are any fit to survive? Journal of International Money and Finance, 24(7), 1150–1175. https://doi.org/10.1016/j.jimonfin.2005.08.002
Chipunza, K. J., Muguto, H. T., Muguto, L., & Muzindutsi, P.-F. (2020). Stock returns and valuation ratios at sector level in South Africa: The regime-switch modelling approach. Global Business Review. Advance online publication. https://doi.org/10.1177/0972150920976641
Clark, T. E., & West, K. D. (2007). Approximately normal tests for equal predictive accuracy in nested models. Journal of Econometrics, 138(1), 291–311. https://doi.org/10.1016/j.jeconom.2006.05.023
Engel, C., & West, K. D. (2005). Exchange rates and fundamentals. Journal of Political Economy, 113(3), 485–517. https://doi.org/10.1086/429137
Evans, M. D. D., & Lyons, R. K. (2002). Order flow and exchange rate dynamics. Journal of Political Economy, 110(1), 170–180. https://doi.org/10.1086/324391
Hansun, S., Putri, F. P., Khaliq, A. Q. M., & Hugeng, H. (2022). On searching the best mode for forex forecasting: Bidirectional long short-term memory default mode is not enough. IAES International Journal of Artificial Intelligence, 11(4), 1596–1606. https://doi.org/10.11591/ijai.v11.i4.pp1596-1606
Hassanzadeh, M., & Mousavi, S. (2023). Real effective exchange rate misalignment and currency crisis in Iran. Future Business Journal, 9, Article 3. https://doi.org/10.1186/s43093-023-00182-8
Juvenal, L., & Santos Monteiro, P. (2025). Currency risk premia and exporter dynamics (IDB Discussion Paper No. IDB-DP-1089). Inter-American Development Bank.
Kim, C.-J., & Park, C. (2016). Real exchange rate dynamics and the Taylor rule: Importance of Taylor-rule fundamentals, monetary policy shocks, and risk-premium shocks [Working paper].
Lustig, H., Roussanov, N., & Verdelhan, A. (2011). Common risk factors in currency markets. The Review of Financial Studies, 24(11), 3731–3777. https://doi.org/10.1093/rfs/hhr068
Mancini, L., Ranaldo, A., & Wrampelmeyer, J. (2013). Liquidity in the foreign exchange market: Measurement, commonality, and risk premiums. The Journal of Finance, 68(5), 1805–1841. https://doi.org/10.1111/jofi.12053
Mark, N. C. (1995). Exchange rates and fundamentals: Evidence on long-horizon predictability. American Economic Review, 85(1), 201–218.
Meese, R. A., & Rogoff, K. (1983). Empirical exchange rate models of the seventies: Do they fit out of sample? Journal of International Economics, 14(1–2), 3–24. https://doi.org/10.1016/0022-1996(83)90017-X
Meese, R. A., & Rose, A. K. (1991). An empirical assessment of non-linearities in models of exchange rate determination. The Review of Economic Studies, 58(3), 603–619. https://doi.org/10.2307/2298014
Menkhoff, L., Sarno, L., Schmeling, M., & Schrimpf, A. (2012a). Carry trades and global foreign exchange volatility. The Journal of Finance, 67(2), 681–718. https://doi.org/10.1111/j.1540-6261.2012.01728.x
Menkhoff, L., Sarno, L., Schmeling, M., & Schrimpf, A. (2012b). Currency momentum strategies. Journal of Financial Economics, 106(3), 660–684. https://doi.org/10.1016/j.jfineco.2012.04.015
Menkhoff, L., Sarno, L., Schmeling, M., & Schrimpf, A. (2017). Currency value. The Review of Financial Studies, 30(2), 416–441. https://doi.org/10.1093/rfs/hhw067
Ni, L., Li, Y., Wang, X., Zhang, J., Yu, J., & Qi, C. (2019). Forecasting of forex time series data based on deep learning. Procedia Computer Science, 147, 647–652. https://doi.org/10.1016/j.procs.2019.01.189
Pan, G.-G., Shiu, Y.-M., & Wu, T.-C. (2025). Time-varying predictability of TAIEX volatility. Review of Derivatives Research, 28, Article 6. https://doi.org/10.1007/s11147-025-09212-9
Penzin, D. J., & Salisu, A. A. (2026). Financial stress and exchange rate volatility in Nigeria: A predictability approach. Quality & Quantity, 60, 3223–3236. https://doi.org/10.1007/s11135-025-02389-z
Rossi, B. (2013). Exchange rate predictability. Journal of Economic Literature, 51(4), 1063–1119. https://doi.org/10.1257/jel.51.4.1063
許彧嘉（2018）。《高維度多層 GRU 遞迴類神經網路於新台幣兌美元即時匯率預測研究》〔碩士論文，輔仁大學〕。