本文透過文字探勘的方法，自Reuters新聞網站爬取與歐元、英鎊及美元相關的外匯新聞，利用情緒分析計算每篇報導的新聞情緒指標，試圖分析新聞情緒與匯率之間的關聯性。首先，本文以迴歸模型探討新聞情緒與匯率報酬率的同期(contemporaneous)關係與預測(predictive)能力；第二，我們建構LSTM神經網路模型預測EUR/USD、GBP/USD的隔日匯率收盤價格(one-step-ahead closing price)，並與傳統的計量模型進行比較。此外，本文也檢視將新聞情緒因子納入LSTM神經網路模型後，是否能提升預測表現，藉此顯示新聞情緒為一項有效的匯率預測因子。最後，對於機構法人或散戶投資人而言，匯率預測的目的不外乎是從交易中獲利，因此我們以LSTM神經網路建構交易策略，並以情緒指標作為進出場決策之濾網。本研究的實證結果發現，在迴歸分析中，新聞情緒與匯率報酬率具有同期的解釋能力，而在預測迴歸式中則不具統計顯著性；匯率預測方面，當LSTM神經網路模型在納入新聞情緒因子後，確實能夠提升預測表現。最後，在LSTM交易策略中，以情緒指標作為濾網能夠大幅提高策略的績效表現，顯示新聞情緒不論是在LSTM模型的預測或交易策略的建構上均為一項具有關鍵性的重要因子。

This paper utilizes text-mining methods to crawl forex news related to Euro, British pound, and U.S. dollar from the Reuters news website. We use sentiment analysis for the calculation of news sentiment indicators for each report, attempting to analyze the correlation between news sentiment and foreign exchange rates. First, we create 3 regression models to explore the contemporaneous relationship and the predictive power of news sentiment. Second, we build an LSTM neural network model to predict the one-step-ahead closing price of EUR/USD, GBP/USD exchange rates, and compare it with the traditional econometric model. Furthermore, we also examine whether the inclusion of the news sentiment factor into the LSTM neural network model can enhance the forecasting performance, thereby showing that news sentiment is an effective exchange rate predictor. Last but not least, in terms of institutional or individual investors, the main purpose of exchange rate prediction is nothing more than gaining profits from trading. Therefore, we build trading strategies based on LSTM neural networks and exert sentiment indicators as the trading signal’s filters. The empirical results show that in the regression analysis, there’s a contemporaneous relation between news sentiment and exchange rate return. Nevertheless, for the predictive regression, they are not statistically significant. As for exchange rate forecasting, when the LSTM neural network model incorporates news sentiment factors, it can indeed improve the forecasting performance. Finally, for the LSTM trading strategy, the use of sentiment indicators as the trading signal’s filters can tremendously improve its performance, indicating that news sentiment is a crucial factor both for forecasting the exchange rates movement and for trading strategy.

Antweiler, W., & Frank, M. Z. (2004). Is all that talk just noise? The information content of internet stock message boards. The Journal of Finance, 59(3), 1259-1294.
Baker, M., & Wurgler, J. (2006). Investor sentiment and the cross‐section of stock returns. The Journal of Finance, 61(4), 1645-1680.
Baker, M., & Wurgler, J. (2007). Investor sentiment in the stock market. Journal of economic perspectives, 21(2), 129-152.
Bao, W., Yue, J., & Rao, Y. (2017). A deep learning framework for financial time series using stacked autoencoders and long-short term memory. PloS one, 12(7), e0180944.
Brown, G. W., & Cliff, M. T. (2005). Investor sentiment and asset valuation. The Journal of Business, 78(2), 405-440.
Chen, S. S., & Hsu, C. C. (2019). Do stock markets have predictive content for exchange rate movements? Journal of Forecasting, 38(7), 699-713.
Chung, S.-L., Hung, C.-H., & Yeh, C.-Y. (2012). When does investor sentiment predict stock returns? Journal of empirical finance, 19(2), 217-240.
Cocianu, C.-L., & Avramescu, M.-Ş. (2020). The Use of LSTM Neural Networks to Implement the NARX Model. A Case Study of EUR-USD Exchange Rates. Informatica Economica, 24(1), 5-14.
Dhamija, A. K., & Bhalla, V. (2011). Exchange rate forecasting: comparison of various architectures of neural networks. Neural Computing and Applications, 20(3), 355-363.
Dzielinski, M. (2011). News sensitivity and the cross-section of stock returns. NCCR FINRISK(719).
Evans, M. D., & Lyons, R. K. (2008). How is macro news transmitted to exchange rates? Journal of Financial Economics, 88(1), 26-50.
Garcia, D. (2013). Sentiment during recessions. The Journal of Finance, 68(3), 1267-1300.
Hasselgren, A., Peltomäki, J., & Graham, M. (2020). Speculator activity and the cross-asset predictability of FX returns. International Review of Financial Analysis, 72, 101561.
He, G., Zhu, S., & Gu, H. (2020). The Nonlinear Relationship between Investor Sentiment, Stock Return, and Volatility. Discrete Dynamics in Nature and Society, 2020.
Heston, S. L., & Sinha, N. R. (2017). News vs. sentiment: Predicting stock returns from news stories. Financial Analysts Journal, 73(3), 67-83.
Ho, T.-W. (2020). Machine Learning is Not as Good as Expected in Time Series Prediction: Evidence from Multistep Dynamic Forecasting. Available at SSRN 3496138.
Huang, D., Jiang, F., Tu, J., & Zhou, G. (2015). Investor sentiment aligned: A powerful predictor of stock returns. The Review of Financial Studies, 28(3), 791-837.
Kim, Y. M., & Lee, S. (2020). Exchange rate predictability: A variable selection perspective. International Review of Economics & Finance, 70, 117-134.
Loughran, T., & McDonald, B. (2011). When is a liability not a liability? Textual analysis, dictionaries, and 10‐Ks. The Journal of Finance, 66(1), 35-65.
Ma, C., Xiao, S., & Ma, Z. (2018). Investor sentiment and the prediction of stock returns: a quantile regression approach. Applied Economics, 50(50), 5401-5415.
Narayan, P. K., Sharma, S. S., Phan, D. H. B., & Liu, G. (2020). Predicting exchange rate returns. Emerging Markets Review, 42, 100668.
Nartea, G. V., Bai, H., & Wu, J. (2020). Investor sentiment and the economic policy uncertainty premium. Pacific-Basin Finance Journal, 64, 101438.
Nassirtoussi, A. K., Aghabozorgi, S., Wah, T. Y., & Ngo, D. C. L. (2015). Text mining of news-headlines for FOREX market prediction: A Multi-layer Dimension Reduction Algorithm with semantics and sentiment. Expert Systems with Applications, 42(1), 306-324.
Nofsinger, J. R. (2001). The impact of public information on investors. Journal of Banking & Finance, 25(7), 1339-1366.
Parot, A., Michell, K., & Kristjanpoller, W. D. (2019). Using Artificial Neural Networks to forecast Exchange Rate, including VAR‐VECM residual analysis and prediction linear combination. Intelligent Systems in Accounting, Finance and Management, 26(1), 3-15.
Ranaldo, A., & Somogyi, F. (2021). Asymmetric information risk in FX markets. Journal of Financial Economics, 140(2), 391-411.
Rogoff, K. S., & Stavrakeva, V. (2008). The continuing puzzle of short horizon exchange rate forecasting.
Schmeling, M. (2009). Investor sentiment and stock returns: Some international evidence. Journal of empirical finance, 16(3), 394-408.
Semiromi, H. N., Lessmann, S., & Peters, W. (2020). News will tell: Forecasting foreign exchange rates based on news story events in the economy calendar. The North American Journal of Economics and Finance, 52, 101181.
Shapiro, A., Sudhof, M., & Wilson, D. (2017). Measuring News Sentiment, Federal Reserve Bank of San Francisco Working Paper 2017-01. Accessed, 17, 51.
Smales, L. A. (2015). Time-variation in the impact of news sentiment. International Review of Financial Analysis, 37, 40-50.
Stambaugh, R. F., Yu, J., & Yuan, Y. (2012). The short of it: Investor sentiment and anomalies. Journal of Financial Economics, 104(2), 288-302.
Tetlock, P. C. (2007). Giving content to investor sentiment: The role of media in the stock market. The Journal of Finance, 62(3), 1139-1168.
Yasir, M., Durrani, M. Y., Afzal, S., Maqsood, M., Aadil, F., Mehmood, I., & Rho, S. (2019). An intelligent event-sentiment-based daily foreign exchange rate forecasting system. Applied Sciences, 9(15), 2980.
何宗武. (2012). R 財經計量入門手冊. 前程文化.
陳旭昇. (2020). 國際金融：理論與實證. 雙葉書廊.
黃裕烈, & 管中閔. (2019). 美國聯準會會議紀要的文字探勘與台灣經濟變數預測. 經濟論文叢刊, 47(3), 363-391.
楊奕農. (2017). 時間序列分析：經濟與財務上之應用(三版). 雙葉書廊.
廖四郎, & 賴嘉蔚. (2019). 卷積神經網路預測時間序列能力分析. 期貨與選擇權學刊, 12(3), 139-190.