本研究使用總經因子和個別外匯因子之交乘項作為LSTM模型的因子，希望藉由深度學習模型來捕捉總經因子和個別外匯因子的互動，並比較其對於外匯超額報酬之解釋力和傳統四因子(利差、動能、價值、市場因子)在線性模型(OLS)上對外匯超額報酬之解釋力的差異。而在因子的部分，本文做了特徵篩選的處理，希望能提升模型的預測力，最後在比較樣本外R^2時，發現LSTM模型的表現優於OLS模型。

接著，將預測力較好的LSTM模型進行策略交易，把LSTM模型預測出的國家超額報酬進行排列，買入預測前25%的國家貨幣，賣出預測後25%的國家貨幣，進而和傳統價值、動能及利差交易策略建構的投資組合做比較，並以夏普比率(Sharpe Ratio)及卡馬比率(Calmar Ratio)當作績效的衡量，最後在結果上發現LSTM模型建立的投資組合績效優於傳統價值、動能及利差因子進行的交易策略。另外，本文最終也探討因子之重要度，發現和利率相關的總經因子對於外匯超額報酬有不錯的預測能力。

This paper used the covariates which are the product of macroeconomic factors and specific foreign exchange factors to train LSTM model, and author hopes to capture the interaction between macroeconomic factors and specific foreign exchange factors through LSTM model. Additionally, author applied feature selection method, trying to enhance the prediction of models. The purpose of using LSTM model with covariates and OLS model with four traditional factors is to compare the prediction of foreign exchange return. Finally, LSTM model performed better than OLS model in the values of coefficient of determination.

Furthermore, the paper used the outcomes predicted by LSTM model to trade in currency markets and tried to compare the performance made by value trade, momentum trade and carry trade. All strategies were made to buy the currencies in the top quarter of predictions and to sell currencies in the bottom quarter of prediction. Author used Sharpe ratio and Calmar ratio to measure the performance of all strategies, finding that the strategy made by LSTM model outperformed than other strategies. This paper also explored the importance of factors, and it turned out that the factors related to interests predicted well in foreign exchange return.

[1] Asness, C.S., Moskowitz, T.J., & Pedersen, L. (2013). Value and Momentum Everywhere. Journal of Finance, 68(3), 929-985.
[2] Bryan, K., Dacheng, X. & Shihao, G. (2019). Empirical Asset Pricing via Machine Learning. The Review of Financial Studies, 33(5), 2223-2273.
[3] Burnside, C., Eichenbaum, M., Kleshchelski, I., Rebelo, S. (2006). The returns to currency spec- ulation. NBER Working Paper 12489.
[4] Brunnermeier, M.K., Nagel, S., & Pedersen, L.H. (2009). Carry Trades and Currency Crashes. NBER Macroeconomics Annual, 23, 313-347.
[5] Burnside, C., Eichenbaum, M., & Rebelo, S. (2011). Carry Trade and Momentum in Currency Markets. Annual Review of Financial Economics, 3(1), 511-535.
[6] Chaboud, A.P., & Wright, J.H. (2005). Uncovered interest parity: it works, but not for long. Journal of International Economics, 66(2), 349-362.
[7] Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.).
[8] Cover, T.M. and Hart, P.E. (1967) Nearest Neighbor Pattern Classification. IEEE. Transactions on Information Theory, 13, 21-27.
[9] Fama, E.F., & French, K.R. (1993). Common risk factors in the return on stocks and bonds. Journal of Financial Economics, 33(1), 3-56.
[10] Fang G. Liu W. Wang L. (2020). A machine learning approach to select features important to stroke prognosis. Computational Biology and Chemistry, 88, 107316.
[11] Filippou, I., & Taylor, M. P. (2017). Common Macro Factors and Currency Premia Journal of Financial and Quantitative Analysis, 52(4), 1731-1763.
[12] Batista, G. & Monard, M. C. (2003). An Analysis of Four Missing Data Treatment Methods for Supervised Learning. Applied Artificial Intelligence 17: 519–533.
[13] Huang, K., Qiao, M., Liu, X., Liu, S., Dai, M. (2019). Computer Vision and Metrics Learning for Hypothesis Testing: An Application of Q-Q Plot for Normality Test.
[14] Jegadeesh, N., & Titman, S. (1993). Returns to Buying Winners and Selling Losers: Implications for Stock Market Efficiency. Journal of Finance, 48(1), 65-91.
[15] Kroencke, T.A., Schindler, F., & Schrimpf,A. (2014). International diversification benefits with foreign exchange investment styles. Review of Finance, 18(5),1847- 1883.
[16] Lintner, J. (1965). The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets. Review of Economics and Statistics, 47(1), 13-37.
[17] Lakonishok, J., Shleifer, A., & Vishny, R. W. (1994). Contrarian Investment, Extrapolation, and Risk. Journal of Finance, 49(5), 1541-1578.
[18] Lustig, H., Roussanov, N., & Verdelhan, A. (2011). Common Risk Factor in Currency Markets. Review of Financial Studies, 24(11), 3731-3777.
[19] Menkhoff, L., Sarno, L., Shmeling, M., & Schrimpf, A. (2012b). Currency Momentum Strategies. Journal of Financial Economics, 106(3), 660-684.
[20] Menkhoff, L., Sarno, L., Shmeling, M., & Schrimpf, A. (2016). Currency value. Review of Financial Studies, 30(2), 416-441.
[21] Moosa, I. A. (2010). The Profitability of Carry Trade - La redditività del carry trade. Economia Internazionale / International Economics, 63(3), 361-380.
[22] Nelson, M. Q., Pereira, A. C. M., & de Oliveira, R. A. (2017). Stock markets price movement prediction with LSTM neural networks. International Joint Conference on Neural Networks (IJCNN). Piscataway, NJ: IEEE, pp. 1419-1426.
[23] Okunev ,J., & White, D. (2003). Do Momentum-Based Strategies Still Work in Foreign Currency Markets?. Journal of Financial and Quantitative Analysis, 38(2), 425-447.
[24] Qi, L., Khushi, M., Poon, J. (2020). Event-driven LSTM for forex price prediction. In Proceedings of the 2020 IEEE Asia-Pacific Conference on Computer Science and Data Engineering (CSDE), Gold Coast, Australia, 16-18.
[25] Raza, A. (2015). Are Value Strategies Profitable in the Foreign Exchange Market.
[26] Rana, M., Uddin, M., & Hoque, M. (2019). Effects of Activation Functions and Optimizers on Stock Price Prediction using LSTM Recurrent Networks. Proceedings of the 2019 3rd International Conference on Computer Science and Artificial, 354-358.
[27] Sharpe, W.F. (1964). Capital Asset Prices: A Theory of Market Equilibrium Under Conditions of Risk. Journal of Finance, 19(3), 425-442.
[28] Shapiro, S.S., Wilk, M.B. (1965). An analysis of variance test for normality (Complete samples). Biometrika 52, 591–611.