Yes, decision attribution would show the bottom-up source of portfolio returns and improve performance attribution. Now can decision attribution apply to return risk as well ? If one defines risk as the difference between beginning of period return expectations and end of period return results then this is a measure of investment risk that investment managers can be accountable for. Note that this definition of return risk ignores benchmark returns because investors shoulder the absolute return variability of the portfolios they invest in, so this should be most relevant to them. This investment risk would be better explained with decision attribution. Kudos to the author for writing about it here.