The S&P SPIVA studies are indeed wrong, as are the Morningstar studies. These studies use peer groups. All peer groups have biases, but most do not know about the most insidious bias. Classification bias is in all peer groups. It results from the fact that peer groups all have funds that do not belong. This fact is easy to see if you look, although almost no one does. For example, Zephyr has produced style plots of all the funds in a given Morningstar peer group. If you're looking at large value funds, you'd expect the members of this peer group to cluster in the upper left quadrant, but they don't. There are several reasons for classification bias, but suffice it to say that this bias cannot be corrected. It's inherent in all peer groups regardless of what the provider does. Classification bias is particularly pronounced and problematic in hedge fund peer groups. Hedge funds are unique. The word "unique" means without peers.

Classification bias causes funds to win or lose because they don't belong, rather than because they're good or bad. In http://www.ppca-inc.com/pdf/Peer-Groups-Will-Mess-You-Up.pdf I show how I have predicted winners and losers long before the SPIVA studies are published. I can make these predictions because these studies have nothing to do with skill, & everything to do with the impact of classification bias.