I believe the issue is one of managing expectations. If a manager is hired to "be" the market or play a specific role (such as the "core" or a "satellite") within a portfolio, then it is fair to impose tracking error, information ratio and other relative comparisons, in my opinion. On the other hand, if the starting point is a shared expectation of what will make both parties "happy," there is still room for quantitative performance attribution but the relationship is far sturdier to the vagaries of benchmarking.

Active management has become a "loser's game," in the words of the illustrious former head of the CFA Institute, Charlie Ellis. This is because there is far more uniformity in the participants' skills and resources, that "winners" are established more by avoiding mistakes (losing less) rather than going for the winning points.

All managers can't all be above-average - this is by definition. Yet, all managers have the potential to serve a client based on what they bring to the table consistently. Surviving cycles of performance reviews will take faith from both parties and on both parties, so that the "flavor of the year" game can be avoided. This comes back to properly setting and managing expectations.