“…It is expected that the defi ned trend can contribute to improved model stability, in order to eliminate the problem of invariance. Empirical results have shown that the inclusion of stability factors (variables which can offset temporary disturbances and volatilities of the economy) can improve the classifi cation accuracy of forecasting models (Dambolena et al, 1980;Betts et al, 1987;Kahya et al 1999). Several of the variables were seen to be changes not just in ratios, but also trends (Edmister, 1972;Blum, 1974;Lau, 1987;Shumway, 2001;Shin et al, 2005;Neves et al, 2006;Muller et al, 2009;Molina et al, 2009).…”