A defined benefit pension plan makes a promise that is both simple and powerful: lifetime income for participants.

That promise is what makes a pension plan different from a defined contribution plan. A defined contribution plan—for example, a 401(k)—is also a pool of assets, but the participant’s benefit is directly tied to the value of their own individual account. A defined contribution plan sponsor may promise to make contributions on a participant’s behalf, but the promise generally ends there. If the account grows, the participant has more. If the account declines, the participant has less.

A traditional defined benefit pension plan works differently. Lifetime income is the promise under the plan. The plan does not promise participants “whatever the trust can afford.” It promises a defined benefit, usually payable for life, and often with optional forms of payment that may provide continued payments to a spouse or beneficiary upon the participant’s death.

That distinction is the starting point for understanding pension gain/loss.

The cost of the pension promise is uncertain by nature

Even though the plan document clearly defines the benefit formula, the total cost of the pension promise cannot be precisely known in advance.

The plan must pay benefits:

• In unknown amounts
• Beginning at unknown retirement dates
• In unknown payment forms
• For unknown lengths of time
• To an unknown number of participants and beneficiaries

Some participants will terminate employment before retirement. Some will work to normal retirement age. Some will retire early. Some will elect a single life annuity. Others will elect a joint and survivor form. Some participants will live longer than expected; others will not. Active participants in pay-related plans may receive salary increases that are higher or lower than assumed.

Because the future is uncertain, the actuary cannot simply determine the future benefits with perfect knowledge. Instead, the actuary measures the pension obligation using assumptions and actuarial judgment (guided by our training and the principles and best practices that have been developed over time by the actuarial profession).

Why do actuarial assumptions matter?

A pension liability is the present value of expected future benefit payments.

That phrase has two important parts.

First, the actuary estimates the expected future benefit payments. That requires assumptions about the covered population. Common assumptions include:

• Mortality
• Retirement age
• Termination or withdrawal
• Disability
• Form of payment
• Spouse or beneficiary coverage
• Salary increases if benefits are pay-related
• Future hours or service, where relevant

Second, the actuary discounts those expected payments to the valuation date to determine the present value. This reflects the time-value of money. For many multiemployer pension-funding valuations, the discount rate is tied to the expected long-term investment return on plan assets. In plain English, the liability is the amount of money the plan would need today, assuming it earns the expected investment return in the future, to make the expected future benefit payments based on the benefits earned to-date.

Actuaries develop assumptions to quantify the uncertainty around these demographic and economic future outcomes. The assumptions are periodically reviewed and refined by the actuary to ensure that they remain reasonable for the intended purpose.

Assumptions are not just technical details buried in an actuarial report. They are central to the measurement of the promise.

How is pension liability expected to change?

A pension liability is not static from year to year; it changes over time in some expected and predictable ways. Even if every assumption were exactly met, the liability would change.

In general, the liability is expected to increase during the year because participants earn additional benefits during the year. This is the normal cost, or the value of benefits newly accrued.

The liability is also expected to increase because time passes. A benefit payment that was 20 years away at the start of the year is now 19 years away at the end of the year. This is often described as interest cost, or the time-value of money.

At the same time, the liability is expected to decrease when benefits are paid. Once a monthly pension payment is made, the plan no longer owes that payment in the future. The payment releases liability.

A simplified liability roll-forward from year to year is described in Figure 1.

Gain/loss analysis is the process of explaining why the actual result differed from the expected result. Beyond simply helping the actuary in the review of the overall results, this analysis also is an opportunity to validate the continued reasonableness of the actuarial assumptions and can help the actuary detect potential defects in the underlying census data.

What are experience gains and losses?

Actuarial assumptions attempt to predict future changes in demographics. When reality is observed over time, the difference between the actual experience and the actuarial assumption creates an experience gain or loss.

If the difference decreases the liability or improves the funded position, it is referred to as an actuarial gain. If it increases the liability or worsens the funded position, it is called an actuarial loss.

Figure 2 lists several common gains and losses that pensions experience.

For multiemployer trustees, it is especially important to separate investment gain/loss from liability gain/loss. A plan may have strong asset returns but still experience liability losses due to retirements, mortality, form-of-payment elections, or data corrections. Conversely, a plan may experience favorable liability experience but still see funded status decline because investment returns were poor.

The key question is not “Did we have a gain or a loss?”

The better question is “What caused it, and is it likely to happen again?”

A sustained pattern of consistent gains or losses typically points to assumptions that are not the best fit for the population and could be indicative of a liability measurement that is materially overstated or understated.

Assumption changes are different from pension experience

Experience gain/loss looks backward. It compares what happened during the year with what the prior assumptions predicted.

Assumption changes look forward. They occur when the actuary changes the basis used to measure future obligations.

For example, an experience study might show that participants are retiring earlier than previously assumed. The year in which that happened may generate a retirement loss. But if the actuary updates the retirement assumption to reflect earlier retirements in the future, that assumption change may create an additional increase in liability.

That does not mean something “new” happened twice. It means the plan first experienced a deviation from the old assumption and then changed the measurement basis to better reflect expected future behavior. The goal of the assumption change is to reduce future gains or losses and have a better measurement of the true liability associated with the pension promise.

Assumption changes can create large one-time gains or losses. Common examples are listed in Figure 3.

ASOP No. 27 emphasizes that actuaries should evaluate relevant data, consider plan-specific factors, select reasonable assumptions, and review the assumption set for consistency. It also notes that materiality matters; not every assumption requires the same level of refinement if it would not materially affect the measurement.

Why do pension experience studies matter?

A single year of gain or loss may be noise. A pattern is different.

If a plan has mortality losses year after year, retirees may be living longer than the mortality assumption predicts. If the plan repeatedly has retirement losses, participants may be retiring earlier than assumed. If pay-related benefits consistently produce salary losses, the salary scale may be too low. If form-of-payment experience repeatedly increases liability, the assumed election pattern may not reflect actual participant behavior.

This is where experience studies are valuable. A gain/loss analysis tells trustees what happened. An experience study helps determine whether the assumptions should be changed.

Trustees do not need to become actuaries. But they should understand enough to ask good questions.

• Are the gains and losses isolated, or are they recurring?
• Which assumptions are driving the largest deviations?
• Are the deviations credible enough to justify an assumption change?
• Are plan-specific assumptions being used where appropriate?
• Are there administrative or data issues creating apparent gains or losses?
• Are investment losses clearly separated from demographic and liability losses?
• Are assumption changes being shown separately from annual experience?

These questions matter because assumptions affect contribution requirements, funding projections, zone status, withdrawal liability, benefit improvement decisions, and long-term intergenerational equity.

Why this matters for trustees

For a pension plan, the cost of the plan is ultimately the cost of the benefits and expenses paid. The actuarial valuation does not create that cost; it measures it.

But the quality of that measurement matters.

If liabilities are understated for too long, the plan may appear healthier than it really is. Contributions may be set too low, benefit improvements may be adopted too confidently, and future trustees may inherit a funding problem that could have been addressed earlier.

If liabilities are overstated, the plan may collect more than needed from the current generation of contributing employers and active workers. That may strengthen the trust, but it can also raise fairness concerns if today’s contributors fund benefits more conservatively than necessary while benefit levels remain lower than they otherwise could have been.

Neither underfunding nor overfunding is ideal. Both affect real people: retirees, active participants, beneficiaries, contributing employers, and future generations.

The trustee’s call to action

When your actuary presents the annual actuarial valuation results, do not fear asking questions to more fully understand how your plan’s specific circumstances drove the changes to your liabilities and your funded status.

Ask what changed. Ask why it changed. Ask whether the result was expected. Ask whether the same pattern has appeared before. Ask whether the assumptions still fit the plan. Ask whether a demographic experience study is warranted. Ask whether the gain/loss is being driven by investments, participant behavior, mortality, data, benefit payments, or assumption changes.

A pension plan is a long-term promise funded through a long-term trust. Gain/loss analysis is one of the best tools trustees have for understanding whether the plan’s actual experience is tracking the assumptions used to measure that promise.

The goal is not to eliminate gains and losses. That is impossible. Assumptions will never be met perfectly.

The goal is to understand them; learn from them; and use that information to keep the fund solvent, healthy, and fair across generations.

Related insight for pension plan sponsors: Dear Actuary: Should I consider an experience study to reduce cost volatility in my corporate pension plan?