Forecasting stock prices using sentiment information in annual reports - A neural network and support vector regression approach
Part of : WSEAS transactions on business and economics ; Vol.10, No.4, 2013, pages 293-305
Issue:
Pages:
293-305
Author:
Abstract:
Stock price forecasting has been mostly realized using quantitative information. However, recent studies have demonstrated that sentiment information hidden in corporate annual reports can be successfully used to predict short-run stock price returns. Soft computing methods, like neural networks and support vector regression, have shown promising results in the forecasting of stock price due to their ability to model complex non-linear systems. In this paper, we apply several neural networks and ε-support vector regression models to predict the yearly change in the stock price of U.S. firms. We demonstrate that neural networks and ε-support vector regression perform better than linear regression models especially when using the sentiment information. The change in the sentiment of annual reports seems to be an important determinant of long-run stock price change. Concretely, the negative and uncertainty categories of terms were the key factors of the stock price return. Profitability and technical analysis ratios have significant effect on the long-run return, too.
Subject (LC):
Keywords:
stock price, forecasting, prediction, sentiment analysis, annual report, neural networks, e-support vector regression
Notes:
Περιέχει πίνακες, σχήματα και βιβλιογραφία
References (1):
- [1] S.J. Grossman, R.J. Shiller, The Determinantsof the Variability of Stock Market Prices, TheAmerican Economic Review, Vol.71, No.2,1981, pp. 222–227.[2] K. Michalak, P. Lipinski, Prediction of HighIncreases in Stock Prices using NeuralNetworks, Neural Network World, Vol.15,2005, pp. 359–366.[3] K.J. Kim, Financial Time Series Forecastingusing Support Vector Machines,Neurocomputing, Vol.55, 2003, pp. 307–319.[4] G. Zhang, B.E. Patuwo, M.Y. Hu, Forecastingwith Artificial Neural Networks: The State ofthe Art, International Journal of Forecasting,Vol.14, 1998, pp. 35–62.[5] Y. Yoon, J. Swales, Prediction Stock PricePerformance: A Neural Network Approach, In:Proc. of the 24th Annual Hawaii Int. Conf. onSystem Science, 1991, pp.156–162.[6] J. Yao, H.L. Poh, Forecasting the KLSE Indexusing Neural Networks, In: Proc. of the IEEEInt. Conf. on Neural Networks, Vol.2, 1995, pp.1012–1017.[7] N. Baba, M. Kozaki, An Intelligent ForecastingSystem of Stock Price using Neural Networks,In: Proc. of the Int. Joint Conf. on NeuralNetworks, Vol.1, 1992, pp. 371–377.[8] J.H. Choi, M.K. Lee, M.W. Rhee, Trading S&P500 Stock Index Futures using a NeuralNetwork, In: Proc. of the Annual Int. Conf. onArtificial Intelligence Applications on WallStreet, New York, 1995, pp. 63–72.[9] D. Enke, S. Thawornwong, The Use of DataMining and Neural Networks for ForecastingStock Market Returns, Expert Systems withApplications, Vol.29, No.4, 2005, pp. 927–940.[10] X.B. Yan, Z. Wang, S. Yu, Y. Li, Time SeriesForecasting with RBF Neural Network, In:Proc. of the 4th Int. Conf. on Machine Learningand Cybernetics, Guangzhou, Vol.8, 2005,pp.4680–4683.[11] K. Chen, H. Lin, T. Huang, The Prediction ofTaiwan 10-year Government Bond Yield,WSEAS Transactions on Systems, Vol.8, No.9,2009, pp. 1051–1060.[12] L.J. Cao , F.E.H. Tay, Support Vector Machinewith Adaptive Parameters in Financial TimeSeries Forecasting, IEEE Transactions onNeural Networks, Vol.14, No.6, 2003, pp.1506–1518.[13] Ch.K. Volos, I.M. Kyprianidis, S.G.Stavrinides, I.N. Stouboulos, L. Magafas, A.N.Anagnostopoulos, Nonlinear FinancialDynamics from an Engineer’s Point of View,Journal of Engineering Science andTechnology Review, Vol.4, 2011, pp. 281–285.[14] Ch.K. Volos, I.M. Kyprianidis, S.G.Stavrinides, I.N. Stouboulos, SynchronizationPhenomena in Coupled Nonlinear SystemsApplied in Economic Cycles, WSEASTransactions on Systems, Vol.11, No.12, 2012,pp. 681–690.[15] F. Neri, Learning Predictive Models forFinancial Time Series by Using Agent BasedSimulations, Lecture Notes in ComputerScience, Vol.7190, 2012, pp. 202–221.[16] F. Neri, Agent based Modeling under Partialand Full Knowledge Learning Settings toSimulate Financial Markets, AICommunications, Issue 4, Vol.25, 2012, pp.295–304.[17] P.Ch. Chang, Ch.H. Liu, A TSK Type FuzzyRule based System for Stock Price Prediction,Expert Systems with Applications, Vol.34,No.1, 2008, pp. 135–144.[18] R. Tsaih, Y. Hsu, C.C. Lai, Forecasting S&P500 Stock Index Futures with a Hybrid AISystem, Decision Support Systems, Vol.23,No.2, 1998, pp. 161–174.[19] G. Armano, M. Marchesi, A. Murru, A HybridGenetic-Neural Architecture for Stock IndexesForecasting, Information Science, Vol.170,No.1, 2005, pp. 3–33.[20] V. Olej, Prediction of the Index Fund byTakagi-Sugeno Fuzzy Inference Systems andFeed-Forward Neural Network, In: Proc. of the5th WSEAS Int. Conf. on Artificial Intelligence,Knowledge Engineering and Data Bases,Madrid, Spain, 2006, pp. 7–12.[21] R.K. Lai, Ch.Y. Fan, W.H. Huang, P.Ch.Chang, Evolving and Clustering FuzzyDecision Tree for Financial Time Series DataForecasting, Expert Systems with Applications,Vol.36, 2009, pp. 3761–3773.[22] P. Hajek, Forecasting Stock Market Trendusing Prototype Generation Classifiers, WSEASTransactions on Systems, Vol.11, No.11, 2012,pp. 671–680.[23] J.Y. Campbell, J. Ammer, What Moves theStock and Bond Markets? A VarianceDecomposition for Long-Term Asset Returns,The Journal of Finance, Vol.48, No.1, 1993,pp. 3–37.[24] J.Y. Campbell, Y. Hamao, Predictable StockReturns in the United States and Japan: AStudy of Long-term Capital Market Integration,The Journal of Finance, Vol.47, No.1, 1992,pp. 43–69.[25] J.Y. Campbell, R.J. Shiller, Valuation Ratiosand the Long-run Stock Market Outlook: AnUpdate. No.w8221, National Bureau ofEconomic Research, 2001.[26] Z. Ding, C.W.J. Granger, R.F. Engle, A LongMemory Property of Stock Market Returns anda New Model, Journal of Empirical Finance,Vol.1, No.1, 1993, pp. 83–106.[27] P. Bak, M. Paczuski, M. Shubik, PriceVariations in a Stock Market with ManyAgents, Physica A: Statistical Mechanics andits Applications, Vol.246, No.3, 1997, pp. 430–453.[28] N. Barberis, A. Shleifer, R. Vishny, A Modelof Investor Sentiment, Journal of FinancialEconomics, Vol.49, No.3, 1998, pp. 307–343.[29] F. Neri, Quantitative Estimation of MarketSentiment: A Discussion of Two Alternatives,WSEAS Transactions on Systems, Vol.11,No.12, 2012, pp. 691–702.[30] J. Bollen, H. Mao, X. Zeng, Twitter MoodPredicts the Stock Market, Journal ofComputational Science, Vol.2, No.1, 2011, pp.1–8.[31] P.C. Tetlock, Giving Content to InvestorSentiment: The Role of Media in the StockMarket, Journal of Finance, Vol.62, 2007, pp.1139–1168.[32] E.A. Demers, C. Vega, Soft Information inEarnings Announcements: News or Noise?Working Paper, INSEAD, 2010.[33] S. Das, M. Chen, Yahoo!for Amazon: OpinionExtraction from Small Talk on the Web.Working Paper, Santa Clara University, 2001.[34] G.F. Kohut, A.H. Segars, The President's Letterto Stockholders: An Examination of CorporateCommunication Strategy, Journal of BusinessCommunication, Vol.29, No.1, 1992, pp. 7–21.[35] R. Feldman, S. Govindaraj et al.,Management's Tone Change, Post EarningsAnnouncement Drift and Accruals, Review ofAccounting Studies, Vol.15, 2010, pp.915–953.[36] F. Li, The Information Content of ForwardLookingStatements in Corporate Filings - ANaïve Bayesian Machine Learning Approach,Journal of Accounting Research, Vol.48, No.5,2010, pp. 1049–1102.[37] A. Huang, A. Zang et al., Informativeness ofText in Analyst Reports: A Naïve BayesianMachine Learning Approach. Working Paper,The Hong Kong University of Science andTechnology, 2010.[38] T. Loughran, B. McDonald, When Is a LiabilityNot a Liability? Textual Analysis, Dictionaries,and 10-Ks, The Journal of Finance, Vol.66,No.1, 2011, pp. 35–65.[39] P. Hajek, V. Olej, Evaluating Sentiment inAnnual Reports for Financial DistressPrediction using Neural Networks and SupportVector Machines. In: Proc. of the 14thEngineering Applications of Neural Networks(EANN’13), 2013, (in press).[40] R. Feldman, M. Fresko et al., Text Mining atthe Term Level. In: Proc. of the 2nd Europ.Symposium on Principles of Data Mining andKnowledge Discovery, 1998, pp. 65–73.[41] G.H. John, R. Kohavi, K. Pfleger, IrrelevantFeatures and the Subset Selection Problem. In:Proc. of the 11th Int. Conf. on MachineLearning, 1994, pp. 121–129.[42] M.A. Hall, Correlation-Based Feature SubsetSelection for Machine Learning. Hamilton,University of Waikato, 1998.[43] S. N. Sivananda, S. N. Deepa, Introduction toGenetic Algorithms. Berlin, Springer Verlag,2007.[44] R.P. Lippman, An Introduction to Computingwith Neural Nets, IEEE ASSP Mag., Vol.4,1987, pp. 4–22.[45] F. Rosenblatt, The Perceptron, a ProbabilisticModel for Information Storage andOrganization in the Brain, Psychol. Rev.,Vol.65, No.6, 1958, pp. 386–408.[46] S. Haykin, Neural Networks: A ComprehensiveFoundation. New Jersey, Prentice-Hall Inc.,1999.[47] D.S. Broomhead, D. Lowe, MultivariateFunctional Interpolation and AdaptiveNetworks, Complex Systems, Vol.2, 1988, pp.321–355.[48] N. Cristianini, J. Shawe-Taylor, AnIntroduction to Support Vector Machines andother Kernel-based Learning Methods.Cambridge University Press, 2000.[49] A. Smola, J. Scholkopf, A Tutorial on SupportVector Regression, Statistics and Computing,Vol.14, 2004, pp. 199–222.[50] V.N. Vapnik, The Nature of StatisticalLearning Theory. Springer, New York, 1995.[51] P. Duspiva, J. Novotný, Utilization ofQuantitative Methods in the Decision MakingProcess of a Manager. Scientific Papers of theUniversity of Pardubice, Vol.17, No.2, 2010,pp. 63–69.[52] O. Rejnuš, Finanční Trhy. Key Publishing,Ostrava, 2011, (in Czech).[53] P. Hajek, V. Olej, Credit Rating Modelling byKernel-based Approaches with Supervised andSemi-Supervised Learning, Neural Computing& Applications, Vol.20, No.6, 2011, pp. 761–773.[54] H. Lin, K. Chen, Soft Computing Algorithmsin Price of Taiwan Real Estates, WSEASTransactions on Systems, Vol.10, No.10, 2011,pp. 342–351.[55] P. Hajek, V. Olej, Municipal RevenuePrediction by Ensembles of Neural Networksand Support Vector Machines, WSEASTransactions on Computers, Vol.9, No.11,2010, pp. 1255–1264.[56] K. Chang, S.S. Wang, P. Ke, Y.R. Huang, Y.Zhen, The Valuation of Futures Options forEmissions Allowances under the TermStructure of Stochastic Multi-Factors, WSEASTransactions on Systems, Vol.11, No.12, 2012,pp. 661–670.
