@inproceedings{9346fcbc927d4a3dad23d09d3db826ef,
title = "Predicting future earnings change using numeric and textual information in financial reports",
abstract = "The main propose of this study is to build a more powerful earning prediction model by incorporating risk information disclosed in the textual portion of financial reports. We adopt the single-index model developed by Weiss, Naik and Tsai as a foundation. However, other than the traditionally used numeric financial information, our model adds textual information about risk sentiment contained in financial reports. We believe such a model can reduce specification errors resulting from pre-assuming linear relationship, thus can predict future earnings more accurately. The empirical results show that the modified model does significantly improve the accuracy of earning prediction.",
keywords = "Earnings prediction, Risk sentiment, Single-index model, Textual information",
author = "Chen, {Kuo Tay} and Chen, {Tsai Jyh} and Yen, {Ju Chun}",
year = "2009",
doi = "10.1007/978-3-642-01393-5_7",
language = "???core.languages.en_GB???",
isbn = "9783642013928",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "54--63",
booktitle = "Intelligence and Security Informatics - Pacific Asia Workshop, PAISI 2009, Proceedings",
note = "Pacific Asia Workshop on Intelligence and Security Informatics, PAISI 2009 ; Conference date: 27-04-2009 Through 27-04-2009",
}