[01] Carol Anne Hargreaves, Department of Statistics and Applied Probability, National University of Singapore, Singapore. [02] Chen Leran, Department of Statistics and Applied Probability, National University of Singapore, Singapore.

Background: Stock market prediction is one of the most challenging tasks since the financial time series is highly volatile, noisy, non-linear and dynamic in nature. Long-Short-Term-Memory (LSTM) is a deep learning technique that focuses on sequential learning. Researchers have shown that the LSTM makes good prediction in the S&P 500 stocks, the Hong Kong stocks and the US intra-day stock data. However, there is limited research in the Australian stock market. Objective: This study aims to apply the LSTM technique to predict the stock price movement in the Australian Stock Market and to identify which stocks to buy for a profitable portfolio. Methodology: We analyzed 400 stocks and selected the top 5 stocks to buy and trade, based on the predictions of the LSTM, Regression Tree (CART) and the Auto Regressive Integrated Moving Average (ARIMA) techniques. Results: The results showed that the LSTM, a deep learning neural network algorithm, outperformed the CART and ARIMA-Time Series algorithms by achieving a return rate of thirty-five percent, a Sharpe ratio of 2.13 and a maximum drawdown of 0.34. The LSTM portfolio had a higher overall return rate of 35% versus the Australian market index of 21%. Conclusion: The LSTM networks made more accurate predictions on stock prices than the ARIMA time series and regression trees (CART). This may be because the LSTM networks is good at processing sequential data, extracting useful information and dropping unnecessary information. Further, the LSTM had a relatively more stable return compared to the ARIMA model. This is because a deep learning model is more capable of extracting the non-linear relationship in the data. In addition, the LSTM model stock portfolio outperformed the stock market index and generated profits over three consecutive time periods.

Deep Learning, Deep Neural Network, Long-Short-Term-Memory (LSTM), ARIMA-Time Series, Classification and Regression Tree (CART), Stock Portfolio, Stock Prediction

[01] Adebiyi, A. A., Adewumi, A. O., Ayo, C. K (2014). Stock Price Prediction using the ARIMA Model. UKSim-AMSS 16th International Conference on Computer Modelling and Simulation, DOI 10.1109/UKSim.2014.67, pages, 105 - 111 [02] Adebiyi, A. A., Adewumi, A. O., Ayo, C. K (2014). Comparison of ARIMA and artificial neural networks models for stock price prediction. Journal of Applied Mathematics, Volume 2014, Article ID 614342, 7 pages, https://www.hindawi.com/journals/jam/2014/614342/ [03] Ballings, M., Van den Poel, D., Hespeels, N., Gryp, R (2015). Evaluating multiple classifiers for stock price direction prediction. Expert Systems with Applications, Vol. 42, page, 7046–7056. [04] Bao, W., Yue, J. and Rao, Y (2017). A deep learning framework for financial time series using stacked autoencoders and long short term memory. PLoS ONE, Vol. 12 (7): e0180944. https://doi.org/10.1371/journal.pone.0180944 [05] Barak, S., Modarres, M (2015). Developing an approach to evaluate stocks by forecasting effective features with data mining methods. Expert Systems with Applications, Vol. 42, page, 1325–1339. [06] Boroovkova, S., Tsiamas, I (2019). An ensemble of LSTM neural networks for high-frequency stock market classification. Journal of Forecasting, v. 38, page, 600–619. DOI: 10.1002/for.2585 [07] Brida, J. G., & Risso, W. A. (2010). Hierarchical structure of the German stock market. Expert Systems with Applications, Vol. 37(5), pages, 3846–3852. [08] Chang, T. S (2011). A comparative study of artificial neural networks, and decision trees for digital game content stocks price prediction. Expert Systems with Applications, Vol. 38, page, 14846–14851. [09] Chen, K., Zhou, Y., Dai, F (2015). A LSTM-based method for stock returns prediction: A case study of china stock market. 2015 IEEE International Conference on Big Data (Big Data), Santa Clara, CA, page. 2823–2824. [10] Chen, S., Ge, L (2019). Exploring the attention mechanism in LSTM-based Hong Kong stock price movement prediction. Quantitative Finance, Vol. 19, pages, 1507–1515. DOI: 10.1080/14697688.2019.1622287. [11] Fischer, T., Krauss, C (2018). Deep learning with long-short-term-memory networks for financial market predictions. European Journal of Operational Research, Vol. 270, pages, 654–669. [12] Gao, P., Zhang, R, Yang, X (2020). The Application of Stock Index Price Prediction with Neural Network. Mathematical and Computational Applications — Open Access Journal, 25, 53; doi:10.3390/mca25030053 [13] Gashler, M., Giraud-Carrier, C. G. & Martinez, T. R. (2008). Decision tree ensemble: small heterogeneous is better than large homogeneous. In Seventh International Conference on Machine Learning and Applications, pages, 900–905. Washington, DC: IEEE Computer Society. [14] Graves, A., Liwicki, M., Fernandez, S., Bertolami, R., Bunke, H., Schmidhuber, J (2009). A novel connectionist system for unconstrained handwriting recognition. IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 31, No.5, pages, 855-868. [15] Guo Z, Wang H, Liu Q, Yang J (2014). A Feature Fusion Based Forecasting Model for Financial Time Series. Plos One, Vol 9(6), pages, 172-200. [16] Hargreaves, C. A (2017). Machine Learning Application in the financial Markets Industry. Indian Journal of Scientific Research, Volume 17, Issue 1, pages, 253-256. [17] Hargreaves, C. A., Yi, H (2012). Does the use of Technical & Fundamental Analysis improve Stock Choice?: A Data Mining Approach applied to the Australian Stock. Market. IEEEXPLORE Conference Proceeding. https://ieeexplore-ieee-org/document/6396537 [18] Hochreiter, S., Schmidhuber, J (2006). Long short-term memory. The MIT Press Journals, Vol. 9, pages, 1735–1780. https://doi-org.libproxy1.nus.edu.sg/10.1162/neco.1997.9.8.1735 [19] Huang, W., Nakamori, Y., Wang, S. Y (2005). Forecasting stock market movement direction with support vector machine. Computers & Operations Research. Vol. 32(10), Pages, 2513-2522. [20] Kim, K-J, Han, I (2000). Genetic algorithms approach to feature discretization in artificial neural networks for the prediction of stock price index. Expert Systems with Applications, Vol. 19, pages, 125–132. https://doi.org/10.1016/S0957-4174(00)00027-0 [21] Lahmiri, S (2012). Linear and nonlinear dynamic systems in financial time series prediction. Management Science Letters 2, pages, 2551–2556 [22] Levin, N, Zahavi, J (2001). Predictive modeling using segmentation. Journal of Interactive Marketing, Vol. 15, pages. 2–22. [23] Loh, W.-Y (2014) Fifty years of Classification and Regression Trees. International Statistical Review, Vol. 82, Issue 3, pages, 329–348. doi:10.1111/insr.12016 [24] Lu, C. J., Lee, T. S., Chiu, C. C (2009). Financial time series forecasting using independent component analysis and support vector regression. Decision Support Systems. Vol. 47(2), pages, 115-125. [25] Lui, Y (2019). Novel volatility forecasting using deep learning long-short-term-memory recurrent neural networks. Expert Systems with Applications, Vol. 132, pages. 99–109. [26] Mallikarjuna, M, Prabhakara Rao, R (2019). Evaluation of forecasting methods from selected stock market returns. Financial Innovation. https://jfin-swufe.springeropen.com/articles/10.1186/s40854-019-0157-x [27] Ou, P, Wang, H (2009). Prediction of stock market index movement by ten data mining techniques. Modern Applied Science, Vol. 132, pages. 28–42. [28] Pai, P.-F., Lin, C.-S (2005). A hybrid ARIMA and support vector machines model in stock price forecasting. Omega, Vol. 33, pages. 497–505. [29] Prasaddas, S., Padhy, S (2012). Support Vector Machines for Prediction of Futures Prices in Indian Stock Market. International Journal of Computer Applications. Vol. 41(3), pages, 22-26. [30] Pyo, S, Lee, J, Cha, M. Jang, H (2017). Predictability of machine learning techniques to forecast the trends of market index prices: Hypothesis testing for the Korean stock markets. PLoS ONE 12(11): e0188107. https://doi.org/10.1371/journal.pone.0188107 [31] Stoean, C, Paja, W, Stoean, R, Sandita, A (2019). Deep architectures for long-term stock price prediction with a heuristic-based strategy for trading simulations. https://doi.org/10.1371/journal.pone.0223593. [32] Singh, R.; Srivastava, S (2017). Stock prediction using deep learning. Multimedia Tools and Applications, Vol. 76, pages. 18569–18584. [33] Tabachnick, B. G., Fidell, L. S (2013). Using Multivariate Statistics. Upper Saddle River, NJ, USA, 6th edition. [34] Takeuchi, L., Lee, Y.-Y (2013). Applying deep learning to enhance momentum trading strategies in stocks. [35] Ture, M., Tokatli, F., Kurt, I (2009). Using Kaplan–Meier analysis together with decision tree methods in determining recurrence-free survival of breast cancer patients. Expert Systems with Applications, Vol. 36, pages, 2017–2026. [36] Wang, B., Huang, H., Wang, X (2012). A novel text mining approach to financial time series forecasting. Neuro-computing, Vol. 83(6), pages, 136-145. [37] Yin, Q, Zhang, R, Liu, Y, Shao, X (2017). Forecasting of stock price trend based on CART and similar stock. International Conference on Systems and Informatics (ICSAI). DOI: 10.1109/ICSAI.2017.8248523