The role of conditioning information upon the unconditional performance of active (tactical) portfolio strategies is analysed in two methodological settings: the UE approach developed by Hansen an Richard (1987) and Ferson and Siegel (2001), and the linear (or multiplicative) approach which was widely used in empirical tests (scaling returns) and theoretically justified Brandt and Santa Clara (2006), and Brandt, Santa Clara and Valcanov (2009) and others. In an earlier project, the theoretical foundations were developed to apply the tests to the case without riskfree asset. In this project, we address the testability of the two approaches and the empirical propoerties of the unconditionally efficient portfolio weights' estimates.