Many pension plan Statements of Investment Policies & Goals (SIP&G) include an investment performance goal for their investment manager, based on beating the increase in the Consumer Price Index (CPI) plus some specified target, typically between 3% and 5%. In six of the past 10 years, more than 99% of the investment managers in Canada managing balanced funds achieved this goal and in two of the past 10 years over 99% of the investment managers failed to achieve this goal.1 It might appear that such a performance objective has no value, as pension funds are rarely even close to it.

However, there is a need for a pension plan to have a CPI-related target, but it must be distinguished from the measure that might be used to evaluate an investment manager's performance.2

There are two sets of performance objectives for pension assets, which have quite distinct purposes, that are often inadvertently mixed. One objective, such as earning more than the CPI plus 3%, is referred to as the plan objective and is important for the funding of the plan. For instance, in a plan where the benefits are adjusted for inflation, and the actuary assumes that the fund will earn 3% more than inflation, a target such as CPI plus 3% is important. If the fund earns more than CPI plus 3%, then the next time the actuary performs an actuarial valuation and reports on the financial status of the plan, the actuary will report good news: the fund did better than what was required to fund the benefits. Conversely, if the fund earns less than CPI plus 3%, the actuary will likely report bad news.

On the other hand, the investment manager earning more than CPI plus 3% doesn't say much at all about the investment manager's performance. The markets have all performed well in most years in the 1980s and 1990s, and a CPI + 3% target has been easy to beat.

Most of a fund's result is determined by the extent of its exposure to the financial markets. The exposure to the Canadian stock market might be, for example, 30%, 50% or 80% of the fund. These long-term exposures are not necessarily the same in all plans and should be determined by the plan sponsor according to their own tolerance for risk. They should not be determined by the investment manager. The investment manager may be given direction to manage the equity percentage over the short term (for instance within a range of 35% to 65% around the long-term target of 50%), to add value through these short-term decisions. By doing so, the investment manager might be expected to add some amount of value, over the amount the fund would earn by keeping the equity percentage fixed at 50%. This target of adding value is the investment manager's performance objective.3

So what is the relationship between the plan objective and the investment manager's performance objective? A plan objective such as CPI plus 3% is not created in a vacuum. Although all plan sponsors would want their pension fund to earn CPI plus 5% rather than CPI plus 3%, for a pension plan to earn a higher return it would have to be invested more aggressively. A pension plan invested conservatively to avoid risk would not reasonably expect to earn much more than CPI plus 3%. But a pension plan invested aggressively for high long term returns could easily earn CPI plus 5% over the long term, depending on the economic environment.

Different plans have different plan objectives because of differing tolerances to risk. For instance, the sponsor of a pension plan with a large deficit could be very concerned about steep declines in the financial markets, as such declines could increase the deficit further, which could lead to higher than acceptable contribution rates. Conversely, the sponsor of a pension plan with a large surplus would not likely be as concerned about market declines which, although such declines would reduce the surplus, might not require additional contributions.

The benchmark portfolio is the link between the plan objective and the investment manager's performance objective. It is a portfolio of the asset classes (such as stocks and bonds) expected to be held in the fund over the long-term at percentages referred to as the "long term asset mix" or "policy asset mix".

The benchmark portfolio follows from the plan objective. Pension plan sponsors will adopt different benchmark portfolios (and long-term asset mixes) for their plans because of their different tolerance to risk. For instance, the plan with benchmark portfolio A may have little tolerance for risk, and has therefore adopted a portfolio with less exposure to Canadian equities and more money market securities than the average pension fund. Such a fund would have lower investment returns on average than other pension funds, but also a lower chance of losing money. They would have a lower rate of return target, such as CPI plus 3%, because they would be invested primarily in lower risk assets.

A pension plan that could accept poor investment results over the short term would adopt a portfolio that looked like benchmark portfolio B. This portfolio has more Canadian equity exposure than the typical Canadian pension fund, so would likely have a higher rate of return over the long term than the typical fund, but also a greater chance of losing money. They would have a higher rate of return target, such as CPI plus 5%, because they would have a higher concentration of higher-risk higher-reward assets.

Benchmark portfolios A and B are just two examples of such portfolio.5 Over the long term benchmark portfolio B should do better than benchmark portfolio A. The rates of return will vary, depending on the economic environment. Historical results for the 43-year period 1956-1998 are shown in Table 2.

This table shows:

1. In the most recent period 1983-1998, just about any benchmark portfolio beat CPI + 3%. The markets performed much better than inflation;

2. In earlier periods, benchmark portfolio A could not even achieve CPI + 3%. Pension funds needed exposure to stocks to achieve that level of return;

3. In all periods, benchmark portfolio B performed better than benchmark portfolio A, as may be expected.

The Benchmark Portfolio and Performance Objective Link

Assuming the investment manager is given the mandate to invest in stocks and bonds, to actively pick securities within each of those assets classes, and to opportunistically switch between the two assets classes to add value, what is the investment manager's target return?

Some pension plan sponsors evaluate their investment manager's performance by comparing their results to those of other pension funds. This is referred to as "relative performance", and is appropriate when the other funds in the comparison are comparable. A plan with a benchmark portfolio such as benchmark portfolio A using relative performance would want to compare their pension fund performance to other plans with similar benchmark portfolios.

Alternatively, the investment manager's performance might be evaluated against the benchmark portfolio itself. For instance, benchmark portfolio A is investable: each asset class in the portfolio can be purchased with index funds, and held in the proportions set out for each of the asset classes. This "passive" fund is a suitable benchmark for the pension fund. A reasonable target for the investment manager with a traditional balanced fund mandate, consistent with that of the top-performing managers in the last 10-15 years, might be to beat the return of the benchmark portfolio by some amount such as 0.50% to 0.75% each year, before investment management fees.

When the investment manager reports on their results, the fact that the manager, like most other managers, beat the plan objective (such as CPI plus 3%) doesn't necessarily mean the investment manager did well. It just means that when the actuary reports on the next valuation, the report will have good news: the fund did better than the rate assumed by the actuary in the valuation.

More important to the evaluation of the investment manager's performance is the comparison of the fund return to that of the benchmark portfolio. Did the manager beat the benchmark portfolio of the plan, and if so, by how much? In what areas did the manager do well, and in which areas did the manager do poorly?

The plan objective and the investment manager's performance objective are distinct, but both are important. They are linked through the benchmark portfolio. The plan objective is the rate of return that the fund should achieve to maintain the plan at a reasonable cost. The benchmark portfolio should be chosen to achieve this objective. The investment manager's performance objective is to beat the benchmark portfolio by a specified amount.

1. Rates of return of typical (median) funds can be compared to typical CPI-related targets. In only 1992 and 1998 were the median returns close to the CPI-related target.

2. The analysis in this article describes the performance objectives that might be set for an investment manager managing a balanced fund of stocks and bonds. The investment manager's performance objective depends on the mandate given to the investment manager.

3. An investment manager's performance objective depends on the investment manager's mandate. For a balanced fund of stocks and bonds, the manager might be expected to add 1.00% of value in excess of the benchmark. A manager managing purely a bond portfolio might be expected to add 0.50%; whereas a manager managing purely an equity portfolio might be expected to add 1.50%.

4. Most pension plans will have more asset classes than these, including US stocks, international stocks, and possibly mortgages and real estate. Benchmark portfolios A and B were limited to 3 asset classes to simplify the discussion.

5. In developing a benchmark portfolio, a plans sponsor will need to choose asset classes and the mix of them. The choice of asset classes depends on return expectations and risk, the availability of such assets in the market, the need for liquidity, the ease of explaining such assets to the Trustees of the pension plan. The mix of asset classes in the Benchmark Portfolio depends on the plan sponsors risk tolerance.

Harry Satanove is an investment consultant in Vancouver.