Good article.<p>I'll nitpick by disagreeing with his assertion that asset allocation decisions weren't analysed for skill vs luck. We were certainly analysing our performance into decision types and their results for skill or luck. The results were pretty conclusive. We started to (mildly) overweight our skill decision types and underweight our luck decision types and the result was some very successful portfolios for our clients.<p>I can't say whether all investment houses were doing it, probably not, but I'm pretty sure the houses known for their quant work would have been.