由於長短期記憶(Long-short term memory, LSTM)能夠非常良好地處理以及預測具有時間序列特性的資料，近幾年開始被應用在預測未來的股價。許多研究皆顯示，LSTM的預測能力比傳統的計量經濟以及機器學習模型還精準。然而，我們認為在文獻中，仍有許多待補強的地方。本研究透過納入更多有意義的變數作為LSTM的輸入變數以及去除輸入變數的雜訊，來增進模型的預測能力。此外，本研究首先嘗試揭露LSTM在數據中所學到的知識。我們依據模型對股價的預測值將股票分成”看漲組”和”看跌組”，並分別對這兩個組別進行統計驗證。
根據我們的研究結果，首先，我們成功透過增加輸入變數以及去除雜訊的方法，改善LSTM模型的預測能力。此外，我們注意到不同個股之間的預測力差異很大，這個現象是由於測試集中存在著LSTM尚未學習到的特徵所產生的。最後，我們發現LSTM學習到一些市場的異常現象，像是價格反轉、過度反應以及反應不足。在過去幾個交易日內，具有高報酬、高波動性以及高成交量的個股，在不久的將來容易發生價格反轉的現象。並且，市場對於新聞的好消息容易過度反應而對壞消息則是反應不足，這樣的現象造成股價容易被高估並在不久的將來反轉。然而，上述的異常現象在大型股之間並不顯著。

Many studies exhibit that LSTM can outperform other traditional econometric and machine learning models. However, there are still somewhere insufficient we would like to make up to the literature. For the first time, we include more meaningful input variables and separate the noise from the input variables of the LSTM model, attempting to improve the predictability of the models. In addition, in order to reveal the knowledge that the LSTM model learn, we assign the stocks into bullish group and bearish group based on the predictive values of the stock prices, and then exhibit the descriptive statistics for each two group. Based on our empirical results, first of all, we acquire more precise predictive results after extending and denoising the input variables of the LSTM model. Furthermore, we notice the existence of the unlearned characteristics in the testing sets weakens the predictability of the LSTM model. Finally, we find our LSTM model captures the patterns of commonly known market anomalies about price contrarian, overreaction and underreaction. However, the market anomalies referred above are not quite significant among large cap stocks.

論文審定書 i
摘要 ii
ABSTRACT iii
LIST OF FIGURES vi
LIST OF TABLES vii
Chapter 1 INTRODUCTION 1
1.1 Background Information 1
1.2 Research Purpose 2
1.3 Research Contribution 3
Chapter 2 LITERATURE REVIEW 4
2.1 Stock price prediction 4
2.2 Text Mining Information in News 6
2.3 Long Short-Term Memory 9
Chapter 3 METHODOLOGY 12
3.1 Description of Data and Software 12
3.2 Forecasting Model 14
3.3 Selection and Preprocessing of Features 22
3.4 Training the Model 28
3.5 Empirical Method 31
Chapter 4 EMPIRICAL RESULTS 32
4.1 Predictive Performance 32
4.2 Details about What the LSTM Learned 38
CONCLUSION 47
REFERENCES 49

Rolf W.Banz (1981). The relationship between return and market value of common stocks. Journal of Financial Economics, Volume 9, Issue 1, Pages 3-18.

Donald B.Keim and Robert F.Stambaugh (1986). Predicting returns in the stock and bond markets. Journal of Financial Economics, Volume 17, Issue 2, Pages 357-390.

Richard H. Thaler (1987). Anomalies: The January Effect. Journal of Economic Perspectives, Volume 1.

John Y. Campbell, Robert J. Shiller (1988). Stock Prices, Earnings and Expected Dividends. The Review of Financial Studies, Vol. 1, No. 3, pp. 195-228.

Eugene F.Fama and Kenneth R.French (1989). Journal of Financial Economics,
Volume 25, Issue 1, Pages 23-49.

Malcolm Baker and Jeffrey Wurgler (2002). The Equity Share in New Issues and Aggregate Stock Returns. The Journal of Finance, Volume 55, Issue 5.

Daniel Kahneman and Amos Tversky (1979). Prospect Theory: An Analysis of Decision under Risk. Econometrica, 1979, vol. 47, issue 2.

Robert Shiller (1984). Stock Prices and Social Dynamics. Cowles Foundation Discussion Papers from Cowles Foundation for Research in Economics, Yale University.

Sendhil Mullainathan and Richard H. Thaler (2000). Behavioral Economics. NBER Working Paper No. w7948.

Andrei Shleifer and Robert W. Vishny (2012). A Survey of Corporate Governance. The Journal of Finance, Volume 52, Issue 2.

Dan Sullivan (2001). Document Warehousing and Text Mining: Techniques for Improving Business Operations, Marketing, and Sales.

Andrei Shleifer, Lawrence H. Summers and James Bradford DeLong (1990). Positive Feedback Investment Strategies and Destabilizing Rational Speculation. Journal of Finance, 1990, vol. 45, issue 2, 379-95.

Gregory W. Brown and Michael T. Cliff (2004). Investor sentiment and the near-term stock market. Journal of Empirical Finance, 2004, vol. 11, issue 1, 1-27.

Sepp Hochreiter and Jurgen Schmidhuber (1997). LONG SHORT-TERM MEMORY. Neural Computation 9(8):1735 – 1780.

Felix A. Gers, Jurgen Schmidhuber and Fred Cummins (1999). Learning to Forget: Continual Prediction with LSTM. Technical Report IDSIA-01-99.

Felix A. Gers, Nicol N. Schraudolph and Jurgen Schmidhuber (2002). Journal of Machine Learning Research 3, 115-143.

Fischer Thomas and Krauss Christopher (2017). Deep learning with long short-term memory networks for financial market predictions. FAU Discussion Papers in Economics, No. 11.

Murtaza Roondiwala, Harshal Patel and Shraddha Varma (2015). Predicting Stock Prices Using LSTM. International Journal of Science and Research (IJSR).

Eugene F. Fama (1970). Efficient Capital Markets: A Review of Theory and Empirical Work. The Journal of Finance, Vol. 25, No. 2, pp. 383-417.

Heiko Jacobs (2015). What Explains the Dynamics of 100 Anomalies? Journal of Banking and Finance, Volume 57, Pages 65–85.

Palangi H, Ward R, Deng L. (2016), Distributed Compressive Sensing: A Deep Learning Approach. IEEE Transactions on Signal Processing. 2016; 64(17):4504–18.

Gal, Y., Ghahramani, Z. (2016), A theoretically grounded application of dropout in re-cur- rent neural networks. In: Advances in Neural Information Processing Systems. pp. 1019–1027.

V. N. Vapnik. The Nature of Statistical Learning Theory. Springer, New York, 1995.

K. R. Muller, A. Smola, G. Ratsch, B. Scholkopf, J. Kohlmorgen and V. N. Vapnik. Predicting time series with support vector machines. ICANN, 999-1004, 1997.

T. B. Trafalis and H. Ince, 2000, Support vector machine for regression and applications to financial forecasting. IJCNN2000, 348-353.

Chang, Y., M. Metghalchi, & C. Chan. (2006). Technical Trading Strategies and Cross-national Information Linkage: The case of Taiwan Stock Market. Applied Financial Economics, 16, pp.731–743.

M. Metghalchi, Y. Glasure, X. Garza-Gomez, Ch. ChenFor (2012). The Austrian Stock Market, International Journal of Economics and Finance Vol. 4, No. 1.

Guo, J.T., & Wu, R.C. (1998). Financial liberalization and the exchange-rate exposure of the Taiwanese firms: A nonparametric analysis. Multinational Finance Journal, 2(1), 37–61.

Chiao, C., Hung, K., & Nwanna, G. I. (2001). The impact of market liberalization on firms’ exchange rate exposure. Competitiveness Review, 11(1), 40–52.

T Singh, S Mehta and M.S.Varsha (2010). Macroeconomic factors and stock returns: Evidence from Taiwan. Journal of Economics and International Finance Vol. 2(4), pp.217-227.

WL Hsieh (2015). Predicting Company Revenue Trend Using Financial News.

B.L. Bowerman, R.T. O’Connell, A.B. Koehler (2004). Forecasting, time series and regression: An applied approach. Thomson Brooks/Cole, Belmont, CA.

S Kim, H Kim (2016). A new metric of absolute percentage error for intermittent demand forecasts. International Journal of Forecasting, Volume 32, Issue 3, July–September, Pages 669-679.

Andrew W. Lo and A. Craig MacKinlay (1990). When are Contrarian Profits Due to Stock Market Overreaction? The Review of Financial Studies Vol. 3, No. 2, pp. 175-205.

Malcolm Baker and Jefferey Wurgler (2006). Investor Sentiment and the Cross‐Section of Stock Returns. The Journal of Finance, Volume 61, Issue 4.

Charles M.C. Lee and Bhaskaran Swaminathan (2000). Price Momentum and Trading Volume. The Journal of Finance, Volume 55, Issue 5.

Wesley S.Chan (2003). Stock price reaction to news and no-news: drift and reversal after headlines. Journal of Financial Economics, Volume 70, Issue 2, Pages 223-260.

Massoud Metghalchi, Yung-Ho Chang and Xavier Garza-Gomez (2012). Technical Analysis of the Taiwanese Stock Market. International Journal of Economics and Finance, Vol. 4, No. 1.

Brad M. Barber and Terrance Odean (2008). All That Glitters: The Effect of Attention and News on the Buying Behavior of Individual and Institutional Investors. The Review of Financial Studies, Volume 21, Issue 2, 1 April 2008, Pages 785–818.

X. Frank Zhang (2006). Information Uncertainty and Stock Returns. Journal of Finance, 2006, vol. 61, issue 1, 105-137.