Abstract: Generalizations of traditional preference criteria such as the Sharpe ratio, the information ratio and the Jensen alpha are obtained by maximizing a certain equivalent excess return (CER) under relevant investment conditions. They are increasing functions of CERs and therefore equivalent criteria. They are consistent with utility theory and are applicable to any investment choice. That is not the case for many other popular preference criteria (e.g., Omega index, Sortino ratio, expected shortfall and so-called ?coherent' preference criteria). Most are incompatible with expected utility maximization and therefore best avoided.