Neuro-Fuzzy Approach To Fundamental Analysis
The purpose of this study was to perform fundamental analysis and cross-sectional prediction of stock return with neuro-fuzzy. Also this study tries to understand the investors' process of financial statement analysis by interpreting the neuro-fuzzy model rules. The data set consisted of firms traded on the Istanbul Stock Exchange (ISE) in Turkey during the period of 1992-1999. Validation of the neuro-fuzzy model is conducted at the portfolio level. Even though there wasn't any statistically significant difference, the neuro-fuzzy model provides slightly higher return than benchmark portfolios. Also this approach exposes how investors select those firms which have low P/E ratios but high gross profit and/or operating profit.