Time Series versus Cross Sectional Momentum

We empirically examine the differences between time-series and cross-sectional momentum based on the stocks included in the German Prime All Share Index. Our method is mainly based on the findings of Goyal and Jegadeesh (2017) and Kim et al. (2016). In particular, we examine the relevance of volatility scaling and net-long positions for comparing the two momentum strategies. With equally-sized long and short positions, we observe cross-sectional momentum outperforming time-series momentum for a sample period from 1996 to 2018. Net-long positions and volatility scaling contribute both to even higher returns. However, we find that not the weighting in volatility scaling, but the time-varying investment into risky assets causes significant changes in returns. All in all, no unequivocal evidence for the superiority of the timeseries momentum could be determined suggesting that abnormal returns in the study of Moskowitz et al. (2012) were not due to the anomaly itself.