Why does the expected active return grows with the square root of breadth

E(active return) = IC * square root of breadth * active risk

If IC is one, it means that there is a perfect correlation between forecasted and realized active returns and we would have that active risk is realized every time. Therefore, expected returns would simply be the active risk times the number of independant decisions. What is the justification for the square root?

It says in the curriculum to look for Buckle 2004 for a more detailed proof, but the full paper is not available online.

Thanks,