Designing Stable Income

Why composition matters in lifetime pension pools

BY JEAN-FRANÇOIS BEGIN AND BARBARA SANDERS
Photo Credit: Adobe Stock

Retirement systems in Canada and the United States have been moving away from traditional defined-benefit pensions and toward arrangements in which individuals bear responsibility for managing their retirement savings.1

Many retirees now reach retirement with accumulated balances in defined contribution plans or similar vehicles, but without a clear path for turning those assets into reliable lifetime income.

The two options normally considered by retirees (i.e., individual drawdown strategies and, to a lesser extent, life annuities) both raise concerns. Annuities provide longevity protection but can reduce initial income or liquidity relative to an unguaranteed pool, whereas individual drawdown strategies preserve flexibility but leave retirees exposed to longevity risk. Lifetime pension pools have therefore attracted some attention recently as a middle-ground solution: They provide income for life through risk pooling, but without a guarantee.

Lifetime pension pools are built on a simple idea: By sharing longevity risk across many retirees, members can receive income for life without requiring a traditional insurance guarantee. The usual intuition is that bigger pools should produce more stable outcomes. That intuition is useful, but incomplete. In practice, members are not identical—they join at different ages, bring different amounts of retirement savings, and may have different mortality expectations. A recent SOA report, “Understanding Heterogeneity in Lifetime Pension Pools: Implications for Practice,” shows that these differences can materially affect income stability. In short, pooling retirees is not just a numbers game.

QUICK REFRESHER ON LIFETIME PENSION POOLS

Lifetime pension pools—also known as variable payment life annuities, dynamic pension pools, and retirement tontines—allow retirees to convert a lump sum of savings into income for life. A retiree contributes assets to the pool and, in return, receives regular payments for as long as they live. This helps address one of the central risks of retirement income planning: outliving one’s savings, especially when assets are managed through an individual drawdown strategy.

The notion of pooling is central to how these arrangements work. Members share longevity risk with one another rather than transferring it to a third party. Because no one guarantees a fixed income level, benefits must adjust over time to reflect the pool’s actual mortality and investment experience. If experience is favourable, benefits may rise; if more members survive than expected or investments underperform, benefits could be cut.

There are many ways to design a lifetime pension pool and to update benefits over time. This article focuses on a design similar to the variable payment life annuities offered by the Faculty Pension Plan at the University of British Columbia and the Australian Retirement Trust’s Lifetime Pension. For simplicity, it is assumed that benefits are adjusted once a year, with the revised benefit paid immediately after the pool’s annual experience is measured. This example is illustrative, since actual designs may differ.

Suppose a new member joins a lifetime pension pool at the beginning of a year, making a lump sum contribution from their retirement savings to the pool. Their lump sum is converted into an initial lifetime income; this conversion is based on pricing, which typically includes a mortality assumption derived from a life table and an assumed rate of return, often called the hurdle rate. Specifically, the member’s initial benefit is calculated by dividing their contribution to the pool by an annuity factor based on the hurdle rate and life table.

Then, at the end of each year, the benefit is updated to reflect the pool’s actual experience. This update applies two experience-based adjustments to the previous benefit: one for mortality experience and one for investment experience.

Figure 1. Typical benefit update rule

Source: Authors

The mortality experience adjustment (green box in Figure 1) reflects how the pool’s actual survival experience compares with expectations. Loosely speaking, if more members die than expected, the assets left behind by those members are shared among the survivors, pushing benefits upward. If more members survive than expected, benefits are adjusted downward so that the pool can continue paying lifetime income to the remaining members.

The investment experience adjustment (purple box in Figure 1) reflects how the pool’s assets performed relative to the hurdle rate used to set benefits. If investment returns exceed the hurdle rate, benefits tend to increase. If returns fall short, benefits tend to decrease.2

WHY STABILITY MATTERS

The benefit-update rule highlights a central feature of lifetime pension pools: Income can change from year to year as experience emerges. This flexibility allows the pool to provide lifetime income without a traditional guarantee, but it also means that members bear some variability in their payments.

For retirees, this variability matters. Many retirement expenses are recurring and difficult to adjust quickly, so a significant benefit reduction may be disruptive even if the expected level of income remains attractive. A pool that provides a high initial benefit but exposes members to frequent or significant reductions may therefore be less appealing than one that starts at a lower level but provides a more stable income path.3

This shifts attention from the level of benefits alone to the stability of those benefits over time. In practice, the relevant issue is not only how much income the pool can pay today, but how confident members can be that this income will remain reasonably close to its expected level.4

One way to answer this question is to use a time-based measure of stability. In the report, the stable income period is considered, which measures how long benefits are expected to remain within a chosen range around their initial level, with a chosen level of confidence.5 For example, a pool operator might ask how many years benefits are likely to remain within 10% of their initial level with 95% confidence. This type of measure translates benefit variability into a language that is easier to communicate: not just how much income might fall, but how long income is likely to remain stable.

Once stability is viewed this way, pool composition becomes important. If benefit changes depend on the pool’s mortality and investment experience, then members’ ages and initial benefit levels can all affect the stability of income over time.

THREE PRACTICAL LESSONS

The “Understanding Heterogeneity” report points to three practical lessons for pool operators:

  1. Scale matters, but the stability gained from adding members depends on the characteristics of those members.
  2. Differences in initial benefit levels can weaken the diversification gains from pooling.
  3. The importance of heterogeneity depends on the investment strategy, because investment risk can dominate other sources of benefit variability.

POOL SIZE HELPS, BUT COMPOSITION MATTERS

Larger lifetime pension pools generally provide more stable income because mortality experience is diversified across more members. As members are added, the pool becomes less exposed to the mortality experience of any one individual, and benefit adjustments tend to become more predictable.

However, pool size is not the whole story. The stability gained from adding members depends on the characteristics of the members being added. In particular, adding members of different ages can have very different effects on the stability of future benefits.

Figure 2 illustrates this point. The figure shows the stable income period as additional members are added to an existing pool of 100 65-year-old members. Each line represents a different age for the additional members. The upward slope of the lines confirms the basic diversification effect: adding more members generally improves stability. But the lines do not increase at the same rate. Adding younger members yields larger improvements in stability, while adding older members produces smaller gains and may remain below the stability achieved by a pool composed of more similar members. The result informs pool analysis and does not imply that operators should exclude older members.

Figure 2. Stability as a function of the number of additional members and their ages

Source: “Understanding Heterogeneity in Lifetime Pension Pools: Implications for Practice,” Section 4. For pool operators, growth should therefore not be assessed only by counting the number of members. An operator considering an expansion, merger or new entrant group should also consider how the additional members change the pool’s age profile.

DIFFERENCES IN BENEFIT LEVELS TEND TO REDUCE STABILITY

Age is not the only source of heterogeneity in a lifetime pension pool. Members may also differ in the lump sum they contribute to the pool and, as a result, in the size of the benefits they receive. These differences matter because benefit adjustments are applied across the pool, but the relative importance of each member in the pool is not the same.

The results show that differences in initial benefit levels tend to reduce stability. When members have similar benefit levels, the pool’s mortality experience is spread more evenly across the membership. But when one group has much larger or much smaller benefits than the rest of the pool, the experience of that group can have a disproportionate effect on benefit adjustments. This can shorten the period over which the benefits are stable.

Figure 3 illustrates this effect by showing how the stable income period changes as new members are added to an existing pool. In this figure, all members are the same age, so the only difference between the curves is the initial benefit level of the added group. The purple curve represents the case where the additional members have the same initial benefit as the existing members. The blue and green curves represent cases where the additional members have lower or higher initial benefits, respectively.

The figure shows that adding members with similar benefit levels produces the strongest stability gains. When the added group has much lower or much higher benefits, the stable income period is lower. In the case where the additional members have much higher benefits, stability can even fall at first before improving as the added group becomes larger. This reflects the fact that a small group with large benefits can have a disproportionate effect on the pool’s experience.

Figure 3. Stability as a function of the number of additional members and their initial benefit levels

Source: “Understanding Heterogeneity in Lifetime Pension Pools: Implications for Practice,” Section 4

This means that pool operators should pay attention not only to how many members are added or how old they are, but also to the size of the benefits they bring into the pool. Large differences in benefit levels can weaken diversification and make benefit adjustments more volatile.

INVESTMENT RISK CAN DOMINATE HETEROGENEITY

The results also show that the importance of heterogeneity depends on the investment strategy of the pool. When the pool is invested conservatively, mortality experience plays a relatively large role in benefit adjustments. In that setting, differences in age and initial benefit levels can have a visible effect on income stability.

The picture changes when the pool is exposed to more investment risk. If assets are invested in a more volatile portfolio, such as a typical balanced pension fund, investment experience can become the dominant source of benefit variability. In other words, the effect of market returns may be much larger than the effect of differences among members. As a result, the stable income period may be driven more by asset allocation than by the age or benefit-level composition of the pool.

This does not mean that heterogeneity becomes irrelevant. Rather, it means that its practical importance depends on the broader design of the pool. In a low-risk investment environment, pool composition may be a first-order design issue. In a higher-risk investment environment, however, investment policy may be the main driver of benefit stability, with heterogeneity playing a secondary role.

DESIGN IMPLICATIONS

We believe the practical implication is that heterogeneity should be treated as a design issue rather than as a secondary detail. Pool operators should not assess a lifetime pension pool only by looking at the number of members or the expected level of benefits. They should also consider how stable those benefits are likely to be under different membership compositions. Measures such as the stable income period can help make the impact of heterogeneity easier to observe and interpret.

This type of analysis can be useful before launching a pool, expanding an existing pool, or merging groups of members. For example, if a new group has a very different age or much higher or lower benefits than the existing membership, the operator can assess whether the new composition improves or weakens stability. The goal is not necessarily to avoid heterogeneity, but to understand when it may create undesirable volatility.

Planning to Read More on Pensions?

Read The Actuary Canada article, “From Pension Risk to Alignment.”

Read the report “Understanding Heterogeneity in Lifetime Pension Pools: Implications for Practice.” at SOA.org.

Several design tools can help manage these effects. One option is modified entry pricing, where the initial benefit or contribution terms are adjusted to reflect the characteristics of different groups.6 This can help rebalance the pool when some groups would otherwise have a disproportionate effect on benefit adjustments. Another option is segmentation, where members with substantially different characteristics are placed in separate pools.

Pool operators may also consider contribution limits to prevent a small number of members with very large benefits from having too much influence on the pool’s experience.7 These limits do not eliminate heterogeneity, but they can reduce the risk that one financially significant group drives benefit adjustments for everyone else.

Finally, communication is essential. Member communications need not reproduce every technical detail, but should clearly explain the adjustment formula, why benefits are variable, why they may change, and which factors affect their stability. Clear communication about the role of mortality experience, investment experience and pool composition can help set realistic expectations. In practice, good design is not only about producing stable outcomes, it is also about ensuring that members understand the nature of the arrangement they are joining.

CONCLUDING REMARKS

The practical lesson is not that heterogeneity should be avoided. Real-world pools will inevitably include members with different characteristics. Rather, heterogeneity should be measured, managed, and communicated. Before creating, expanding, or merging pools, practitioners may want to consider not only how many members will be included, but also how their ages, benefit levels, and risk profiles interact. Larger pools can improve stability, but only when the added heterogeneity is well balanced. For lifetime pension pools, good design begins with knowing who is in the pool.

Jean-François Bégin, FSA, FCIA, PhD, is an associate professor at Simon Fraser University. He is based in Vancouver, Canada.

Barbara Sanders, FSA, FCIA, MSc, is an associate professor at Simon Fraser University. She is based in Vancouver, Canada.

Statements of fact and opinions expressed herein are those of the individual authors and are not necessarily those of the Society of Actuaries or the respective authors’ employers.

Copyright © 2026 by the Society of Actuaries, Chicago, Illinois.