The key issue relating to employing multiple managers is the fact that diversification of risk may have been so extreme that there is no potential for added value. In the example above, an uncontrolled diversification between multiple generalist managers resulted in 98% of the fund’s performance being a function of whether the market went up or down, and not from any added-value from the managers.
Rules of Thumb for Combining More Than One Manager:
Rule #1 Know what the aggregate portfolio looks like
Diversifying between funds with generalist mandates can lead to duplication of holdings and little control over whether the sum of the managers adds value. (You may as well invest in an index fund)
Rule #2 Diversify between manager styles not just managers
Top performing managers for given periods often have similar styles. Selecting a team of top quartile managers may simply result in a compounded bet on one investment style and not diversify risk at all.
Rule #3 In selecting more than one manager, the top consideration is how they blend with other managers.
Top quartile return is irrelevant for selecting managers that blend effectively. Trustees need to determine what range of manager styles they require to achieve their objectives first, then identify which manager is best within that given style.
Rule #4 Only a controlled structure can protect against random outcomes.
Many trustees and consultants can assemble a top team of managers to blend together and then lose all potential value by failing to manage that blend over time. Because managers are constantly changing what they buy and sell, the fund must have some structure in place for dynamically managing the blend of managers.
“Who’s in charge of making the stew?”42
What’s wrong with split funding?
If each of your managers is allowed to make the asset allocation decision themselves, then who determines whether all this activity adds up to something better? Surely the only possible outcome is a purely random outcome.43