Abstract
The classical portfolio theory suggests that higher returns of an asset are justified by the higher risk it carries, supported by multi-factor cross-sectional regressions. By investigating the time series and frequency transformation of the Russell 3000 constituents, this study shows that there are weak links between risk and return, as well as trade volume and return. Only 19,02% (13,45%) constituents have significantly positive (negative) risk and return relationships. In addition, only 7,66% (12,77%) of the returns are positively (negatively) related to trade volume. We use the cross-wavelet power spectrums to provide additional evidence on the weak links. The conclusions from cross-sectional analysis might lead to the asset misallocation in a time series setting.