Abstract
This article aims to empirically support the hypothesis that behavioural patterns are key determinants of market movements. We developed a model for predicting market psychology which is based on the application of a self-organizing network algorithm. The estimated model is applied to a mechanical trading system, which independently adopts investment decisions based on the current daily data. The model was tested on the data for daily trading on the Slovenian stock market as an example of an emerging capital market. The performance of the model supports the suggested hypothesis.