Combining bag-of-words and sentiment features of annual reports to predict abnormal stock returns
- 246 Downloads
Automated textual analysis of firm-related documents has become an important decision support tool for stock market investors. Previous studies tended to adopt either dictionary-based or machine learning approach. Nevertheless, little is known about their concurrent use. Here we use the combination of financial indicators, readability, sentiment categories, and bag-of-words (BoW) to increase prediction accuracy. This paper aims to extract both sentiment and BoW information from the annual reports of US firms. The sentiment analysis is based on two commonly used dictionaries, namely a general dictionary Diction 7.0 and a finance-specific dictionary proposed by Loughran and McDonald (J Finance 66:35–65, 2011. doi: 10.1111/j.1540-6261.2010.01625.x). The BoW are selected according to their tf–idf. We combine these features with financial indicators to predict abnormal stock returns using a multilayer perceptron neural network with dropout regularization and rectified linear units. We show that this method performs similarly as naïve Bayes and outperforms other machine learning algorithms (support vector machine, C4.5 decision tree, and k-nearest neighbour classifier) in predicting positive/negative abnormal stock returns in terms of ROC. We also show that the quality of the prediction significantly increased when using the correlation-based feature selection of BoW. This prediction performance is robust to industry categorization and event window.
KeywordsStock return Prediction Text mining Sentiment Neural network
This study was funded by the scientific research project of the Czech Sciences Foundation (Grant No. 16-19590S).
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
- 8.Myskova R, Hajek P (2016) The effect of managerial sentiment on market-to-book ratio. Transform Bus Econ 15:80–96Google Scholar
- 10.Hajek P, Olej V (2013) Evaluating sentiment in annual reports for financial distress prediction using neural networks and support vector machines. In: Iliadis L, Papadopoulos H, Jayne C (eds) Communications in computer and information science. Springer, Berlin, pp 1–10Google Scholar
- 23.Butler M, Kešelj V (2009) Financial forecasting using character n-gram analysis and readability scores of annual reports. In: Gao Y, Japkowicz N (eds) Lecture notes in computer science. Springer, Berlin, pp 39–51Google Scholar
- 24.Hart RP (2001) Redeveloping DICTION: theoretical considerations (new). In: West MD (ed) Theory, method, and practice in computer content analysis. CT Ablex, Westport, pp 43–60Google Scholar
- 27.Hinton GE, Srivastava N, Krizhevsky A, et al (2012) Improving neural networks by preventing co-adaptation of feature detectors, pp 1–18. ArXiv e-prints: arXiv:1207.0580
- 29.Hajek P, Bohacova J (2016) Predicting abnormal bank stock returns using textual analysis of annual reports: a neural network approach. In: Jayne C, Iliadis L (eds) Communications in computer and information science. Springer, Aberdeen, pp 67–78Google Scholar
- 30.Demers E, Vega C (2014) Understanding the role of managerial optimism and uncertainty in the price formation process: evidence from the textual content of earnings announcements. doi:http://dx.doi.org/10.2139/ssrn.1152326
- 50.Feuerriegel S, Ratku A (2016) Analysis of how underlying topics in financial news affect stock prices using latent dirichlet allocation. In: Bui TX, Sprague RH (eds) 49th Hawaii international conference on system sciences. IEEE, Kauai, pp 1072–1081Google Scholar
- 59.Yang Y, Pedersen JO (1997) A comparative study on feature selection in text categorization. In: Machine learning working then conference, pp 412–420Google Scholar
- 64.Nam J, Kim J, Loza Mencía E et al (2014) Large-scale multi-label text classification: revisiting neural networks. In: Calders T, Esposito F, Hullermeier E, Meo R (eds) Lecture notes in computer science. Springer, Berlin, pp 437–452Google Scholar
- 68.Maas AL, Hannun AY, Ng AY (2013) Rectifier nonlinearities improve neural network acoustic models. In: Dasgupta S, McAllester D et al (eds) Proceedings of the 30th international conference on machine learning. JMLR, Atlanta, pp 1–6Google Scholar
- 69.Jaitly N, Hinton G (2011) Learning a better representation of speech soundwaves using restricted boltzmann machines. In: ICASSP on IEEE international conference on acoustics, speech and signal processing. IEEE, Prague, pp 5884–5887Google Scholar
- 72.Taddy M (2015) Document classification by inversion of distributed language representations. In: Proceedings of the 53rd annual meeting of the association for computational linguistics, pp 45–49Google Scholar
- 73.Wong FMF, Liu Z, Chiang M (2014) Stock market prediction from WSJ: Text mining via sparse matrix factorization. In: 2014 IEEE international conference on data mining. IEEE, pp 430–439Google Scholar
- 78.Tang D, Wei F, Yang N, et al (2014) Learning sentiment-specific word embedding for twitter sentiment classification. In: Proceedings of the 52nd Annual meeting of the association for computational linguistics. Association for Computational Linguistics, Baltimore, pp 1555–1565Google Scholar