Hierarchical clustering is an unsupervised learning algorithm that uses the distances between data points and assigns them into clusters with similar traits. This study applies hierarchical clustering to funds in order to group them if they demonstrate similar statistical behaviour in their returns. By testing the hypothesis that there are differences in average returns between assets belonging to the same group, it will be evaluated whether it is possible to predict such differences in future returns as well.
Moreover, the research evaluates whether the fund fee can be used as a predictor of relative performance between funds. This will answer a question of substantial importance - will funds with higher fees on average outperform similar funds with lower fees