London Market Monitor – 30 September 2021
September delivers some ups and downs: Equity and bond markets mostly fell, risk-free rates rose, UK inflation increased, and volatility notched upward.
All pension plans face a constant tension between:
There is generally a tradeoff between these three elements. For example, larger benefits require larger contributions or larger investment returns. Increasing investment returns generally involves increasing risk in the asset portfolio. As a second example, lowering investment risk generally lowers investment returns, which leads to either smaller benefits or larger contributions.
Asset losses during the Great Recession highlighted this tension. In one extreme case, over the two years of 2007 and 2008, a Plan went from 125% to 76% funded based on unsmoothed market assets. The Plan Trustees, who felt very secure at the beginning of 2007 with a 25% reserve, had to make some very difficult decisions by the beginning of 2009 when the Plan was 76% funded. Those Trustees were not alone. After 2008, many plans had to increase contributions and reduce benefits at the same time.
Fast forward to 2021, over the past decade many plans experienced:
This has greatly improved the funded percentage of many plans. Not surprisingly, many multiemployer plan trustees are now asking: Is it finally time to improve benefits? To answer this question, trustees should first address the big picture and revisit their plan’s balance between benefits, contributions, and risks.
In this article, we discuss what “100% funded” means for a pension plan, considerations when setting an appropriate funding reserve, thoughts on improving benefits, de-risking considerations, and a suggestion for a possible path forward.
A funded percentage is calculated by taking the value of the assets divided by the measured value of the obligation, or the liability. A funding percentage of 100% simply means that the current value of assets is equal to the measured liability for benefits earned to date in a pension plan. Said another way, 100% funded means that there are enough assets to pay for the last benefit payment to the last participant the day before they die if all future experience follows the assumptions. To truly understand what 100% funded means though, we need to understand what the measured liability is for a plan.
The measured liability for a pension plan is a single number that represents the target value of assets today needed to pay promised benefits. It is calculated using a set of actuarial assumptions, which are regularly reviewed and updated. These assumptions include investment returns, how long participants will work, when participants will start receiving retirement checks, and how long participants will live in retirement. In general, actuarial assumptions are set to be the “best estimate.” Ultimately, liability calculations will only be accurate to the extent that future experience follows the best estimate assumptions.
So, is it appropriate to target a 100% funded percentage? On the surface, yes, however the reality is that being 100% funded may or may not be enough to pay for all future benefits. A 100% funded percentage today leaves no room for experience worse than the best estimate assumptions in the future.
For illustration, let’s consider two recent bear markets. During the Dot Com Bust, many plans averaged less than 0.0% over 2000, 2001, and 2002. This was much less than their best estimate assumption for earnings over those three years. During the Great Recession, many plans earned negative 25% or less in 2008. Either one of these events could have reduced a 120% funded percentage to less than 100%. Suddenly, even being 100% funded can feel precarious.
The past year, 2020, has been a good reminder of how experience may not meet our expectations. The COVID-19 pandemic brought two surprises to most pension plans: (1) more deaths than expected, and (2) drastically low investment returns for the first three months, followed by outstandingly high returns in the last nine months. Looking forward, we are left wondering how COVID-19 will impact our future experience. After COVID-19, it is not clear if we will see more deaths due to damaged health, or fewer deaths because those already in poor health were hit hardest by COVID-19.
There is concern that the strong market returns from 2019-2020 have lowered future expectations. Specifically, even though the S&P 500 earned 31.5% in 2019 and 18.4% in 2020, these S&P increases were almost entirely driven by increases in stock prices, as opposed to actual increases in company earnings including paid dividends. This means that most of the combined return for 2019 and 2020 was based on investors paying a higher price for earnings.
Investment returns over the next few years will therefore be below average unless (a) there is a large increase in earnings, or (b) investors continue to be willing to pay a high price for earnings. A common message currently delivered by investment consultants at trust meetings is that the Price to Earnings (P/E) ratios are high compared to historic norms. Meanwhile, the good returns in the bond market throughout 2019 and 2020 were driven by decreasing yields. When bond yields go down, bond prices go up, but expected future returns on bonds go down. All of this has led investment experts to lower their expectations for future investment returns. There may be low investment returns in the next few years that hurt plans’ funding.
As COVID-19, the Dot Com Bust, and the Great Recession made clear, actuarial assumptions are not always met. A 100% funded percentage one year does not always mean 100% the next year.
This section started with the question, what does 100% funded mean?
This raises the question of how to create a reserve for adverse deviations from best estimate assumptions.
Two methods to create a reserve are through implicit and explicit reserving.
One way to create an implicit reserve is to include a margin for adverse deviation in the actuarial assumptions. For example, if the median expected investment return for a plan is 5.5%, measuring the liabilities at 5.0% or 4.5% would provide a cushion against adverse investment returns.
An easier reserve to understand is an explicit reserve “target.” For example, a plan may target assets equal to 120% of the measured liabilities using best estimate assumptions. This reserve would create stability in the event of experience below expectations such as a market downturn. Often, explicit reserves are built by contributing more than the annual cost of benefits. In the rest of this article, we will focus on the explicit reserve.
Answering this question requires understanding the plan’s risk tolerance. It may be desirable to target a very large reserve, or it may be acceptable to target 100% funded with no reserve at all. A plan’s ability to withstand adverse events, such as bear markets, will depend on many factors including: How the plan is invested, the size and stability of future contributions, and the plan’s benefit design. While some of these factors are outside of trustee control, trustee decisions directly impact others, such as asset allocation.
A more conservative asset allocation will tend to need a smaller reserve than a more aggressive asset allocation. However, a conservative asset allocation is also likely to have a lower expected investment return, which will ultimately make benefits more expensive.
Deterministic and stochastic projections can provide great insight into how these various factors interact and help trustees determine a target reserve. Deterministic projections look at one specific set of potential future experience including investment returns and contributory hours. Varying the investment returns in a deterministic projection to specific levels shows how the plan would perform in specific situations. For instance, a Plan could look at projections using (a) the 25th percentile for the plan’s specific asset allocation, (b) the plan’s actual returns from the Dot Com Bust, or (c) the Great Recession. As shown earlier, these may imply a reserve of 20% or more is needed.
Stochastic projections produce a probability distribution of future results based on the plan’s asset allocation. One stochastic projection can provide multiple probabilities for the modeled result such as the 5th, 25th, 50th, 75th and 95th percentiles. Stochastic projections can be used to test the probability, given different levels of reserves, that contribution rates may need to increase, or benefits may need to decrease. As an example, trustees may wish to examine the probability that contributions will need to double given a specific starting level of reserves. This in turn will help trustees decide on the appropriate size of a target reserve.
Once the plan has decided on a target reserve, the next steps to consider are balancing: (a) progress toward building up the target reserve, (b) improving benefits, (c) contribution level, and (d) de-risking.
After trustees understand the plan’s risk tolerance and set a target reserve, then the focus can turn to if, when, and how to improve benefits. A set of guidelines for when benefit improvements should be considered is often called a “funding and benefits policy” and will help trustees make consistent and prudent funding decisions. For example, a funding and benefits policy could say that:
Once trustees decide a plan is ready to improve benefits, there are multiple ways in which improvements can be made.
One type of benefit improvement is to increase all benefits going forward for each future year of service. This is most beneficial for working members and has no impact on members who have retired. This improvement has no immediate impact on the measured liability. Instead, as benefits are earned from year-to-year, the measured liability will gradually go up as the additional benefits are earned.
A second type of benefit improvement is to increase benefits earned in the past. These are often called backfills. They do not have any effect on future accruals. A backfill will immediately impact the measured liability, as the service has already been earned. Therefore, the funded percentage will decrease unless additional contributions are made to pay for the improvement.
A third type of improvement is to adopt an alternative plan design. These may not directly increase the size of benefits but can decrease risk so that the plan can provide larger benefits without exceeding its risk tolerance. For instance, benefits can be adjusted to match the value of the assets accumulated to pay for them (see sidebar).
Improving benefits should always be approached with caution, because once improvements have been made to earned benefits, they are irrevocable. In general, improving benefits in a traditional pension plan will also increase the plan’s risk level. This relationship exists because larger benefits will lead to a larger measured liability. The gains and losses on the larger assets and larger measured liabilities will have greater swings when actuarial assumptions are not met.
When answering the question, “is it finally time to improve benefits?” risk should be considered. As stated earlier, there is a constant tension between more benefits, less contributions, and less risk. There are many ways to reduce or manage risk in a pension plan. The de-risking process either reduces the probability of adverse events, reduces the sensitivity to adverse events, increases the ability to withstand adverse events, or reallocates risks to where they can be best handled. Based on a plan’s funding and benefits policy, de-risking can be a consideration either before, in conjunction with, or after benefit improvements.
Reserve building: As discussed above, in the case of an explicit reserve, a target funded percentage is chosen above 100%. As the reserve “fills” it provides a cushion against adverse experience. Alternatively, an implicit reserve could be based on conservative assumptions. The purpose of a reserve, to protect against adverse experience, needs to be communicated to all stakeholders. Without an understanding of the risks the plan faces, funding over 100% could be incorrectly perceived as “excess” or “surplus” funding.
Asset allocation: Trustees can reduce exposure to market risks by updating the plan’s asset allocation to invest in less risky assets (or exploring other options such as purchasing annuities from insurance companies or matching high quality bond payments to retiree payments). Unfortunately, these approaches often reduce expected future returns.
Plan design: Plan benefits can be redesigned such that the risks associated with a retirement plan are minimized and shared in a more rational way. Designs such as the Variable Annuity Pension Plan (VAPP) or Sustainable Income Plan (SIP) pool longevity risk for the members’ benefits, while tying benefit levels directly to actual investment returns. Therefore, members are guaranteed lifelong benefits and the plan stays 100% funded with stable contributions in all market conditions (see sidebar).
Is it finally time to improve benefits? After a decade of lowered benefit accruals and increased contributions, it’s not hard to see why there is high demand for benefit increases. As trustees consider their options, it’s important to consider all factors including the long-term health of the pension plan. Depleting the entire reserve for a large one-time benefit increase followed by a market downturn can negatively impact a plan for a generation or more. Carefully considered guidelines such as a funding and benefits policy can establish a prudent path for managing the tension between more benefits, less contributions, and less risk (see sidebar).
Carefully considered guidelines such as a funding and benefits policy will ideally:
One plan design that many trustees are taking a fresh look at is the Variable Annuity Pension Plan (VAPP). A VAPP adjusts benefits each year based on actual investment returns. If an investment return for a given year is over a threshold, called a hurdle rate, then VAPP benefits increase. If investment returns are below the hurdle rate, then VAPP benefits will go down. One concern with VAPPs is that when returns are below the hurdle rate all benefits go down, including retiree benefits. Many varieties of VAPPs respond to the possibility of retiree benefits going down in different ways. One variety of VAPP, the Milliman Sustainable Income PlanTM (SIP) addresses this by maintaining a “stabilization reserve” to “shore up” benefits that would otherwise go down in retirement. VAPPs and SIPs have the advantages of (a) staying 100% funded in all market conditions allowing stable contributions, while (b) still providing lifelong benefits like a traditional pension plan, and (c) providing some increases to offset inflation over the long term to the extent that returns exceed a conservative hurdle rate. Since these plan design changes affect only VAPP or SIP benefits, the funding risk of any traditional benefits earned in the past is not impacted. Risk is therefore slowly decreased as old traditional benefits are paid out and new VAPP or SIP benefits are earned. Although there are many complications, one possible way to decrease risk more rapidly would be to explore a conversion of prior benefits to the new plan design. For more information on alternate plan design see: https://www.milliman.com/en/retirement-and-benefits/alternate-retirement-plan-design