期刊名称：International Journal of Economics and Finance Studies
电子版ISSN：1309-8055
出版年度：2014
卷号：6
期号：1
出版社：Social Sciences Research Society
摘要：The prediction of business failure is a widely studied subject in financial literature. Many earlier studies on this topic employed statistical methods such as multiple discriminant analysis, logit and probit to predict corporate failure using past financial data (especially the ratio data). However, there has been a recent surge in academic interest in the use of artificial intelligence (AI) methods to predict financial distress. Numerous studies documented that AI methods outperform traditional methods. Majority of these studies used data from established markets, the number of studies on emerging market data is rather limited and only a handful of studies employed Turkish data for analysis. This study aims to contribute to the literature by applying the artificial neural networks to predict deletions from Istanbul Stock Exchange (ISE) National 100 Index. The sample is constructed using the quarterly fundamental data of the companies listed in this index the period between January 2008 and December 2012. We employed Neural Networks (NN), logit and probit to predict deletions from index one quarter before they have occurred. Results show that although the logit provides slightly better in-sample predictions, all of the methods fail to identify deletions in the out-of-sample periods
关键词：Neural Networks; Genetic Programming; Business Failure ; Prediction; Emerging Markets
