Unraveling the effect of engagement and consistency in the results of the M6 forecasting competition

Abstract

The M6 competition evaluated investment performance over a period of one year, contributing to the efficient market hypothesis debate. This paper provides further insights into the outcomes of the competition by unraveling the effect that team engagement and performance consistency had on the final results. First, we identify three different types of engagement and investigate their relationship with portfolio efficiency, also making useful observations about the learning effect implied by a re-submission process. Then, we analyze the monthly performance of the teams and determine whether it aligned with their global performance or was affected significantly by extreme instances. Our results suggest that consistency is more important than engagement for making profitable investments. Nevertheless, we identify many cases where both regular portfolio updates and luck provided an advantage.