Striking the Proper Balance in Multi-Manager Portfolios

Diversification is widely viewed as essential to achieving superior risk-adjusted returns, although diversification is not a guarantee against losses. There are multiple dimensions of diversification: individual security, industry, sector, capitalization, investment style, asset class and even country. Diversification also can apply to the use of multiple investment managers with differing investment approaches. But at what point does an investor cross over from having enough managers to having too many?

There is no single or simple answer to the question of how many managers are ideal, as the portfolio construction process must weigh many elements. Among the relevant factors are the portfolio’s investment objectives, the size of a particular asset class and the size of the investment portfolio.

Specific investment objectives are a driving force in determining the number of investment managers used. Among the questions for an investor to consider are:

• What is your objective for return?
• What is your investment time horizon?
• How much tolerance do you have for volatility?
• How comfortable are you with off-benchmark exposure?

The answers to these questions help to shape the size and scope of a multi-manager program, explains Shannon Eidson, FSA, CFA, investment program strategist at Northern Trust, Chicago.

“Institutional investors have different goals and tolerances,” Eidson notes. “A foundation may focus on spending policy while a defined benefit plan considers funded status as well as the potential implementation of an investment program focused on derisking as funding levels improve,” he says.

“If you have a very narrow mandate — a small, finite universe of securities — there’s not a lot you can do to represent different types of strategies and types of assets compared with a much broader universe,” adds Jessica Hart, director of portfolio construction at Northern Trust, Stamford, Conn. “The size of an investor’s investment program also might influence the number of managers used.”

Maximizing Portfolio Value

One of the goals of a multi-manager portfolio is to maximize returns while minimizing risk. To that end, each manager should be able to add value relative to a benchmark return, net of fees, rather than merely replicate the benchmark’s performance.

Tracking error is a conventional way to evaluate a manager’s potential to add value. Although a high tracking error could produce a higher rate of return, the manager also could be introducing a higher level of risk to the portfolio. A multi-manager program, however, can blend managers with diverse styles and low correlations between their respective portfolios. As a result, the overall portfolio could more closely reflect the target benchmark in terms of characteristics while potentially producing a higher risk-adjusted return.

Another way of evaluating a manager’s potential to add value is by measuring active share, or the percentage of a portfolio aligned to a particular benchmark. What a manager does not own is just as important as what is owned when thinking about returns relative to the benchmark. For example, a portfolio with zero overlap to the benchmark has 100% active share. Portfolios with an active share greater than 60% are generally considered truly active. Active share is also one way to assess the optimal number of managers for a portfolio.

“If a manager is highly skilled, and has achieved better returns than others, that is something that an investor may want to tap into,” says Jennifer Kamp Tretheway, CFA, managing director of manager of manager services at Northern Trust, Chicago. “However, you don’t want to have too much single manager risk.”

The chart on page 17 compares the percentage of tracking error that comes from stock selection for an overall mix (Y axis) relative to the highest allocation to a single manager for the mix (X axis). It is important to be cognizant of manager specific risk, both at the organizational and portfolio level. Therefore, an investor may be able to reduce the amount of risk from stock selection — but better diversify the manager concentration risk — by moving from Mix 1 to the portfolios represented by Mix 2, 3, or 4. The same analysis can be used to determine the impact of adding and subtracting managers.

How Many Is Too Many?

Ultimately, there is no universally correct number of investment managers in a multi-manager portfolio. In fact, at some point simply adding managers to increase diversification will bring little if any benefit.

The process of building a multi-manager portfolio includes extensive analysis and rigorous monitoring of manager correlations, stock selection, tracking error and active share. Once established, the program should be reviewed on a regular basis to assess the need for changes as a result of changing markets and as a result of any changes in manager execution or client requirements. As part of this process, it can be useful to employ customized benchmarks to assess the performance of individual managers as well as the overall blended portfolio.

↑ back to top ↑