Mark and Tom present a variety of charts showing the performance of several funds managed by individuals listed as Morningstar Managers of the Year. They discuss the subsequent performance of these funds after receiving star status, showing that the funds’ alphas (performance over a benchmark) over time were not statistically significant. Mark talks about Step 3: Stock Picking (from his 12-Step Recovery Program for Active Investors), explaining why stock picking and manager picking may not be helpful behaviors and how choosing a manager based on short-term past performance may not lead to optimal fund returns for the investor.

See Step 3: Stock Picking for further explanation of manager alpha over time, leading to luck being the best explanation of previous success. 

See chart references here.

The formula to calculate the number of years needed to determine if an alpha is statistically significant can be found here.  

Also read Brad Steiman's two articles on statistical significance: What's the Significance? and Paradox of Skill.