The conventional wisdom is back-to-front. There is good reason to argue that investors should start to diversify their exposure to passive managers, while diversifying their exposure to active manager less.
Investors above a certain size typically use several professional managers to run various portfolios in order to reduce the risk that one manager will fail in some way. Traditionally, this diversification process centers on the actively managed parts of the portfolio, leaving any passively managed assets in the hands of a single firm.
But while running indexed funds may look simple, there is still plenty of risk involved, both organizational and operational. For example, during the recent credit crisis some index providers had significant losses in their securities lending programs, which hurt fund returns and in several cases limited the ability of clients to withdraw their money.
It's relatively inexpensive to diversify this enterprise risk from passive funds. Dividing institutional-sized passive mandates among two or three managers is likely to increase asset management fees by only a basis point or two. By contrast, dividing a large portfolio among several active managers can drive up fees substantially, since large portfolios typically get fee discounts.
There may also be unintended investment consequences of using two or more active managers in any given asset class. Typically, institutions select managers that deploy different investment approaches, expecting that this will diversify their exposures. But what if they don't?
Let's consider a case where the investor has appointed two local equity managers for mandates of equal size. One manager approaches investing from the bottom-up, while the other has more of a top-down process. The two portfolios have completely different holdings, except for Stock X. Both have positions in Stock X that are one percentage points larger than the stock's benchmark weight, so the combined portfolios also has an position in Stock X that is one percentage point larger than the stock's benchmark weight.
The 1% overweight in Stock X creates far more idiosyncratic risk in the combined portfolio than the same 1% overweight would in either of the underlying portfolios on their own, as the display below shows. Combining the two portfolios reduces the active weight of most stocks in the combined portfolio, because a 2% overweight in any of the other stocks becomes a 1% overweight in the whole. In addition, the combination is likely to diversify sector exposures and other systematic risks, such as market cap or style.
As a result, Stock X ends up dominating the risk profile of the combined portfolio. In the example shown , which was drawn from a particular stock in a particular index, it would account for 9.4% of the active risk in the combined portfolio, even if it were only 4.1% and 7.0% of the active risk in the two separate portfolios.
Rethinking manager diversification
With enough managers in an asset class, the result could be a quasi-index fund with a handful of outsized and unintentional idiosyncratic risks. While each active manager will seek to construct an optimal portfolio trading off risk and return versus a benchmark, the odds of the combined portfolio being optimal are likely to be both random and small.
This investment problem can be mitigated in one of two ways. First, the investor could try to select managers that are very different. Nonetheless, the risk of overlapping security positions amplifying unintentional idiosyncratic risk would still increase with the number of managers. This suggests that the investor should limit the selection of active managers within each asset class.
Second, the investor could introduce an overlay to manage the unintended consequences of overlapping positions. This would create another layer of fees, but might well be worth the money.
Either way, investors need to think through the issues involved in manager diversification carefully. Sometimes, counter-intuitively, fewer managers can actually lead to a better alignment of risk and return.
The views expressed herein do not constitute research, investment advice or trade recommendations and do not necessarily represent the views of all AllianceBernstein portfolio- management teams.
Patrick Rudden is head of blend strategies at AllianceBernstein.
