Which investment choice is most important? Is it the allocation among broad asset classes, country allocation, the choice of sectors, or the selection of individual securities? We argue that many investors have a false impression of the relative importance of these choices for two reasons. First, they typically fail to distinguish between the activities investors choose to emphasize and those with the greatest potential to influence investment results. Second, to the extent they measure the potential influence of an investment activity, they often base it on contrived portfolios that are unobtainable. We simulate thousands of realistic portfolios to determine the natural dispersion of performance arising from different investment activities, which enables us to evaluate the hierarchy of investment choice from a normative perspective.