|
Home > Library > Stable Times > Volume 5, Issue 2
The quarterly publication of the Stable Value Investment Association
Second Quarter 2001 • Volume 5 Issue 2
Assessing Multi-Manager Stable Value Funds
By Victoria M. Paradis, CFA JPMorgan Fleming Asset Management
A multi-manager stable
value fund has more than one firm involved in investment decision-making.
Specialty stable value managers make investment decisions when they choose
GICs or bonds. Active fixed income managers make investment decisions
during the continual portfolio management process. Stable value funds
may have either or both types of managers involved in fund management.
This article explores three aspects of multi-manager funds:
- The case for a single manager fund
- Why consider a multi-manager structure?
- Reaching an optimal multi-manager solution.
The case for a single manager fund
A single manger fund, whether a specialty stable value manager or an active
fixed income manager, offers numerous key benefits to a stable value fund:
- Maximum control
of investment decisions and risks. One decision-maker can more effectively
keep tabs on all positions in a fund.
- Consistency
with other DC plan options. Most DC plans offer a variety of choices
in an investment fund line-up. Each fund typically has a single investment
manager. Participants diversify their manager risk when they choose
multiple funds. Why should stable value be inconsistent?
- Cost
considerations. Separate account fee schedules typically get lower as
portfolio size increases. Splitting a portfolio means that marginally
more assets are invested at higher fees, less at the lower tail end
fees.
- Portability
of today's products enables simple replacement. Today's products are
far more flexible and portable than historically. If results with a
manager are not satisfactory, it is simple to seamlessly replace the
firm.
- Policy is
the predominant source of investment results. The critical fund design
decision to get right is one of investment policy, or investment guidelines.
Participants will always be better off with a single manager and an
optimal investment policy than five managers combining to deliver a
sub-optimal investment strategy.
Why
consider a multi-manager structure?
There are two key reasons for choosing a multi-manager structure:
- access to different skill sets
- excess return diversification
1) Plans with
specialty stable value managers typically invest in Traditional GICs, Treasury,
Agency, Asset-Backed, and simple Mortgage-Backed Securities. There can be
numerous diversification and return benefits from accessing other sectors
- ones that fixed income managers have the expertise, team, systems, and
resources to handle. These could include the following sectors in addition
to those available from specialty stable value managers: investment grade
corporate bonds, prepayment sensitive mortgages, private placements, non-US,
and below investment grade debt. Pairing both styles within a fund can be
effective, but there are important considerations.
First, it's critical to consider the total portfolio results. The overall
fund manager should roll up the portfolio holdings to see the total sector
allocations. We often see distortions when managers are combined. With a
predominant position with specialty stable value managers and a token allocation
to active fixed income portfolios, we tend to see portfolios that are concentrated
in Asset-Backed securities with very low allocations to investment grade
corporate bonds. Is this intentional? If the active portfolios permit extended
sectors such as non-US and below investment grade, often those net positions
are less than 1-2% of the total fund. Is that meaningful? These allocations
should be thought out in advance of manager assignments.
2) The other reason for multi-manager funds is to diversify excess return
and risk. All managers are not alike. It is possible for some managers to
consistently produce results that have higher returns and lower risk than
others. Selecting multiple managers reduces the "luck factor" at getting
that excess return.
Reaching an optimal multi-manager decision
There exist useful quantitative approaches to deciding on multiple managers,
to help answer questions like:
- What styles complement
each other?
- How many managers
are optimal?
- How do you allocate
between managers?
The first thing to
assess is complementary styles. Different investment styles perform better
in different market conditions. Since markets are continually changing,
the idea is to get styles that will work together well. Statistically,
you get this by looking at correlation of excess returns between managers.
You want to find low correlation, where a correlation of "1" is the highest,
"0" implies there is no relationship between manager returns, and "-1"
means two managers portfolios move in opposite directions (the ideal).
The chart summarizes an actual project that looked at a total fund, including
market value returns of both fixed income and stable value managers. This
analysis found that it is possible to find negative correlation between
active fixed income and stable value managers, which is a worthwhile goal.
If a fund were to combine low-correlation managers in equal proportions,
it is quite possible to produce results that have enhanced risk-adjusted
returns. To the Fund, this means less style-specific risk, reduced short-term
volatility of returns, and increased consistency of returns.
Excess Return Correlations
|
|
Fixed Income manager A
|
Fixed Income manager B
|
Fixed Income manager C
|
Fixed Income manager D
|
Stable value manager
|
|
Fixed Income manager A
|
1.0
|
|
|
|
|
|
Fixed Income manager B
|
0.95
|
1.0
|
|
|
|
|
Fixed Income manager C
|
0.71
|
0.54
|
1.0
|
|
|
|
Fixed Income manager D
|
(0.28)
|
(0.35)
|
(0.15)
|
1.0
|
|
|
Stable value manager
|
(0.12)
|
(0.18)
|
(0.21)
|
0.58
|
1.0
|
One common question
is how many managers do you need to have effective diversification? Research
demonstrates that a Fund's total excess return risk decreases as the number
of managers increases, but eventually flattens. Typically, the most significant
enhancement occurs when funds move from 1 to 2 to 3 managers. There is
not much benefit after five managers are combined. The benefit is most
pronounced when the managers have a low correlation.
The last step is to determine what percentage allocation makes sense.
Can a fund do better than the old "1/n" formula? There exist optimization
programs that combine returns, volatility, and correlation data to solve
for any stated criteria, including maximum return, minimum risk, highest
information ratio, and lowest tracking error. The output is an optimal
allocation for a stated historical period. Running such an optimization
over a series of rolling periods will identify trends and suggest an optimal,
efficient portfolio strategy. The net results can be powerful. Analysis
we have seen has demonstrated the potential for enhancing returns by almost
1% with lower risk.
Finally, there are other factors to consider when implementing a multi-manager
analysis.
- Objectivity and
independence. Who's doing the analysis? It's important to have independence
between the individuals performing manager research and investment management.
This will ensure full access to the important data and the necessary
objectivity.
- Ongoing risk management.
This is not just a set-up process. Effective multi-manager funds require
continual risk monitoring and re-assessment.
- Overall duration
control. With multiple moving parts within a fund, it's important to
have a consolidated approach to ensuring that the total duration is
within the appropriate ranges. Futures overlay programs can ensure that
decisions by various managers do not allow the fund to get "offside".
- Consolidated reporting.
Last, the plan sponsor needs meaningful, transparent summary reporting,
as compared with a hodge podge of reports that do not show the total
picture. The net result should be presented with a meaningful perspective.
Read Next: Dealing with Diminished Expectations
|
|
