The special financial assistance (SFA) program established by the American Rescue Plan Act of 2021 has been a lifeline to eligible financially distressed multiemployer defined benefit (DB) plans, providing funds to help delay or avoid plan insolvency. The trustees of a plan that receives SFA (SFA plan) have the unique opportunity to consider other options going forward that may not have been feasible before receiving SFA. For some of these plans, a merger with a better funded plan may be an option. But many SFA plans must forge ahead on their own. In this article, we will discuss how a change to a variable plan design could point an SFA plan in a new direction that could not only help keep the plan solvent, but prospectively address some of the challenges that led to its current funding state. We will also highlight a few design features that may need to be addressed if an SFA plan decides to implement a variable plan design.
There are two important questions to answer before an SFA plan should consider changing its plan design.
Question 1: Will a plan redesign materially change the risk profile of the plan within a reasonable amount of time? There is no bright-line metric that answers this question. It is a combination of many factors such as the comparative benefit levels over time and the relative size of the future active population to the total plan size. Actuarial modeling is needed so that trustees understand the timing and expectations of the change.
Question 2: Can the plan solve its funding problems? SFA plans receive funds that are expected to extend their plan solvency until 2051.1 For many of these plans, there are opportunities to invest plan assets in such a way that plan solvency is projected well beyond 2051. These plans are more attractive candidates to consider a change in design.
Why consider a variable plan design?
Traditional benefit designs provide fixed benefit accruals, which continue to grow over time regardless of how assets perform. Most SFA plans are very mature, meaning a larger portion of the plan is comprised of inactive members and inactive liability. When these plans were certified in critical status and rehabilitation plans were adopted, nearly all found themselves with higher contributions and lower benefits for active members. This is neither desirable nor sustainable in the long term.
SFA amounts received should help these plans reconfigure (at least to some degree) that undesirable relationship of large contributions buying small benefits, but SFA will not address the underlying risks that led these plans to their current financial situations. SFA plans found themselves in dire financial straits due to a variety of factors. While those factors differ by plan and industry, what is commonly seen with nearly all these plans is that their mature demographic populations exacerbated the difficulties for their traditional benefit designs to operate in a sustainable way during highly volatile markets, such as the 2008 global financial crisis. For many, their mature risk profile made it impossible to rebound.
At a high level, variable plans fundamentally change a plan’s risk-sharing profile. Investment performance (both good and bad) is passed directly on to plan members by adjusting accrued variable pension benefits.2 These plans therefore do not generate underfunding from investment performance. However, longevity (mortality) risk is retained by the plan because variable benefit plans still provide lifelong pension income, thereby shielding plan members from a risk that could be extremely difficult for them to manage on their own.
Molding the plan into something new takes time
Modifying a plan’s design is a long-term forward-looking process. There are no changes that can magically fix what happened in the past and eliminate existing underfunding. Elements of a plan’s design can be altered on a prospective basis to make the plan more attractive to both employers and employees. But the impact of those changes will happen gradually over time. The funds received by SFA plans are expected to extend plan solvency until 2051 but amending a plan design now could have a significant impact on the plan down the road and may keep it solvent well beyond 2051.
Figure 1 shows a sample projection for a mature plan (like an SFA plan) changing to a variable plan design. In this case, about 33% of the plan’s liability is projected to be under the new sustainable design after 20 years, and about half of the liability is under the new design after 30 years. A plan that is less (more) mature may see a faster (slower) progression to the new sustainable design.
Figure 1: Projected liability for sample plan
Addressing the challenges of the traditional plan design
The chart in Figure 2 outlines some of the significant challenges facing not only SFA plans but multiemployer DB plans in general and how changing to a variable design can transform them into sustainable retirement plans. Once the new design has been in place long enough, it can ultimately provide the intended lifelong benefits to members with significantly less worry of underfunding, potential plan insolvency, and intergenerational inequities in the level of benefits provided.
Figure 2: Challenges of traditional designs and benefits of a variable design
| Current (traditional) plan design: Current challenges |
New (variable) plan design: Ultimate benefits3 |
|
|---|---|---|
| Funded status | Benefits grow with new contributions while assets fluctuate with the market, which can lead to underfunding. | Benefits grow with new contributions, and both benefits and assets move in tandem, keeping the funded status at or near 100%. No underfunding from investment performance. |
| Zone status | Plan could drop out of the Green Zone. | Plan is highly likely to remain in the Green Zone by design. |
| Contributions | Bargaining parties may have to negotiate additional contributions to satisfy the rehabilitation plan requirements. A sizable portion of contributions may need to be used to pay down existing underfunding for benefits earned by current active, terminated vested, and retired members. Contribution increases can impact pay for active members and the ability for employers to remain competitive in the industry. |
Because the new ultimate plan would not be expected to become underfunded, all contributions are expected to go solely toward benefit accruals for active members, plan administration expenses, and modest reserve building within the plan to account for adverse demographic experience. Importantly, contributions made on behalf of active members are not ever expected to be directed toward the underfunding of a prior generation. Contributions are expected to remain stable, providing predictability for employers and allowing them to remain competitive in the industry. |
| Withdrawal liability | Underfunding can lead to employer withdrawals and potentially high withdrawal liability. Employers may not want to participate due to the risk of withdrawal liability. Bankruptcies and/or withdrawals in overfunded periods may lead to increased liability exposure for remaining employers. |
By design, these plans are expected to remain fully funded, and therefore have no withdrawal liability. More employers would likely participate because withdrawal liability risk is low. |
| Inflation-protected benefits | Benefits remain fixed when earned and after retirement (absent cost-of-living adjustments) and lose value over time due to inflation. | Benefits adjust with investment returns, which can better keep up with inflation, both pre- and post-retirement. |
Some of the risks that can be effectively addressed through a variable plan design are listed below.
Investment risk
This is the risk that investment returns will be different from expected. This is generally the most significant and impactful risk traditional plans face. Plan designs that include fixed benefits can become underfunded if assets do not perform as expected. Conversely, variable plan designs that adjust benefits in line with investment returns will be more likely to remain at or near 100% funded regardless of how the market performs.
Zone status risk
This is the risk that the plan’s zone status deteriorates, requiring the trustees to take action to correct funding shortfalls. SFA plans will be in critical status for about 30 years, but a change to a variable plan design going forward may put the plan on a better path to emerge from critical status after 2051 and reduce its chances of becoming critical again.
Contribution risk
This is the risk that future contribution levels will be significantly less than expected to keep the plan funded. SFA plans likely require supplemental contributions under their rehabilitation plans, contributions well above what the bargaining parties originally negotiated. These supplemental contributions ultimately impact employers’ businesses and their ability to offer competitive wage packages to members. A variable plan design that minimizes or eliminates underfunding can keep contributions stable and predictable for employers and providing valuable lifelong benefits for members.
Withdrawal and/or bankruptcy risk
This is the risk that an employer or a group of employers withdraws that meaningfully reduces the plan’s future covered employment levels. SFA plans, like other plans in critical status, struggle to keep their current employers or to enlist new employers. Sometimes withdrawals involve bankrupt employers unable to pay their withdrawal liability. If employers withdraw from a plan that is underfunded, additional pressure is added to the remaining employers of the plan (and by extension the active members) to make up for the shortfalls over time. A variable plan design that minimizes or eliminates underfunding can keep potential withdrawal liability low, reducing the concern about collecting withdrawal liability payments. This can help retain current employers and create the ability to recruit new employers that will not be concerned about joining a plan where unfunded liability is expected to be small or zero.
Inflation risk
This is the risk that benefits lose value over time due to inflation. Members of traditional DB plans carry this risk. Consider a retiree receiving a $10,000 fixed annual pension. After 10, 20, and 30 years (assuming 3% annual inflation), that fixed amount will not be worth as much in the future and will only be expected to buy the equivalent amount of about $7,450, or $5,500, or $4,100 in goods and services at those future points in time, respectively. A variable plan design that adjusts benefits (even in retirement) can help offset the effects of inflation for members.
How do variable plans work?
As mentioned earlier, variable plan designs alter a plan’s risk-sharing profile by transferring investment risk to plan members while retaining longevity (mortality) risk in the plan. Variable benefits accrue the same as traditional benefits and then adjust up and down based on investment performance, even in retirement. Members will receive lifelong pension income, eliminating the need for them to manage this portion of their retirement assets over their expected lifetime.
This concept of adjusting benefits could be a fundamental shift for some trustees who believe that, once benefits are earned, they should remain fixed. However, one of the strengths of the variable plan design is that it does a good job providing intergenerational equity, sharing all experience equally with all members, active and retired. This can help avoid the situation many plans find themselves in where different cohorts (tiers) of members receive materially different benefit formulas throughout time, as is happening now for most plans receiving SFA assets.
The benefit formula for a variable plan can be based on pay or benefit multipliers per year of service just like traditional plans. However, without “new” money for SFA plans, the benefit accrual rate in a variable plan design will likely be lower initially than the current traditional benefit accrual rate. Some trustees may feel resistant to this change, especially if it is perceived as a benefit reduction. The reality though is that the initial benefits earned by the member stemming from a lower accrual rate are expected to grow over time and will ultimately be much larger. By design, at some point in the future, the variable benefit is expected to increase beyond what the traditional, fixed benefit would have been. Education and modeling are crucial in understanding this feature.
A variable plan design that adjusts benefits up or down based on actual asset returns with no other provisions is a pure variable plan design. It is just one of many in a spectrum of variable designs that trustees can consider. Some features that could be added to a variable design are as follows:
- Increasing the reserve in the plan to be used to help avoid benefit reductions in years when assets underperform.
- Adjusting the benefit before retirement by either fixing the benefit in retirement or allowing members to choose a variable or fixed benefit at retirement.
- Adding a floor benefit that the adjustable benefit will not drop below.
- Accruals that are split between a traditional formula and a variable formula.
Keep in mind that adding features that fix or guarantee certain benefits (such as options 2, 3, and 4 above) will reintroduce the risk that those benefits can become underfunded, shifting the plan’s risk profile somewhat back toward that of a traditional plan design.
Unique design considerations for SFA plans
For readers with a deeper understanding of variable plan designs, we highlight three design considerations SFA plans will face that other plans considering a variable design will not.
Defining the investment return that determines annual variable benefit adjustments
SFA plans have certain conditions and restrictions that apply to them following receipt of SFA. One of the restrictions is that only 33% of the SFA money received may be invested in permissible “return-seeking assets,” such as common stock and certain equity securities. The remaining SFA money received must be invested in permissible “investment-grade fixed income,” such as cash, U.S. Treasuries, and certain high-quality bonds.
If the investment return that determines annual benefit adjustments is based on the investment return on the plan’s total assets, including SFA assets, it could result in smaller benefit adjustments due to the likely lower returns driven by the SFA assets. If future contributions expected to fund the variable accruals are deposited into the non-SFA portion of the plan’s assets, then the plan’s investment return that determines annual benefit adjustments might be best defined as the non-SFA investment return. In this way, the variable accruals and the assets associated with the funding of those accruals will move in tandem. Further, this approach disentangles the impact of likely lower expected asset returns on the restricted SFA portion from influencing the adjustment in benefits.
How additional reserves are built (if applicable)
In designs where benefits are protected from being reduced in years where they would otherwise go down (aka “shore-up” the benefits), additional reserves need to be built for this purpose. Plans typically build reserves with variable benefit modifications or contributions. Because SFA plans are deemed to be in critical status through the plan year ending in 2051, trustees may need to fund this reserve with new contributions not contemplated by the rehabilitation plan. Careful analysis and discussion with plan counsel are important to make sure any “shore-up” design is consistent with regulations.
Many variable plans use a capped investment return, whereby actual plan returns in excess of the cap do not increase benefits and instead are used to build up the reserve. In years of asset underperformance, variable benefit amounts can be maintained at current levels if there are enough reserves. If reserves are not sufficient, variable benefits are subject to decrease.
Because SFA plans may face challenges in achieving superior asset returns when considering the restricted asset pool (SFA money), they may want to consider comparing the non-SFA asset return to the cap when developing the reserve. Careful thought should also be given to a potential situation where an SFA plan is taking substantially more risk with non-SFA assets than with SFA assets. Such an SFA plan may have an overall investment risk profile similar to that of a typical multiemployer DB plan but, when focusing solely on the non-SFA portion of assets, the risk profile may be substantially riskier, perhaps leading to increased volatility in reserves and benefits within the variable construct.
An alternative approach to building the reserve could be to allocate a portion of each percentage point return above a threshold to the reserve. For example, the design could direct one-quarter of the benefit increase in good return years to the reserve while taking nothing for the reserve when benefits are going down. In those years a portion of the reserve would be spent on preventing benefit declines. Such an approach could help build reserves without requiring superior returns.
Reserves could also be built using contributions, which could be a challenge for SFA plans. If a portion of current contributions going to accrue benefits is reassigned to build reserve, this would result in smaller accruals. Asking for additional contributions to build reserve may be a big ask because SFA plans are likely receiving a high level of contributions already due to the supplemental contributions required by their rehabilitation plans. For SFA plans that are expected to emerge from critical status by 2051 and whose supplemental contributions are more than sufficient to ensure that this happens, the level of supplemental contributions required by the rehabilitation plan could be reduced (the reason to do this is described below), and the former supplemental contributions could be used to build reserves. This change would not reduce the overall contribution rate, because one of the conditions for SFA plans is that contribution rates in effect on March 11, 2021, are not decreased unless certain requirements are met.
How the additional reserve is used (if applicable)
If additional reserves are built to shore up benefits, how that reserve is used in years when benefits would otherwise adjust downward necessitates thoughtful planning. SFA plans are deemed to be in critical status through the plan year ending in 2051, and therefore are subject to restrictions on benefit increases. Plans in critical status may not increase benefits unless the actuary certifies that the increase is paid from additional contributions not contemplated by the rehabilitation plan, and the plan is reasonably expected to emerge from critical status on the schedule contemplated by the rehabilitation plan.
If the reserve is developed with new contributions or former supplemental contributions, it is likely to be available for shore-ups. However, if the reserve is not created from those sources, for example if it was built solely from superior investment returns, then more analysis and discussion with plan counsel will be needed.
Charting a new path forward
Answering the two key questions discussed at the start of this article is crucial in determining whether a variable design is right for your SFA plan. Sufficient and robust modeling is recommended and will help answer those fundamental questions.
Ultimately, the journey to a new plan design will likely require several months of special meetings with your actuary and other plan professionals. Meetings will involve education on the various types of variable plan designs, deciding on a new benefit structure while staying within the bounds of critical status and SFA conditions and restrictions, and analyzing the plan’s projected funded status to assess the new structure’s resilience and long-term viability under various market scenarios.
If the trustees decide to adopt a variable plan design, then the rollout to members is the final key step in the process. Remember, members have received numerous notices explaining how the plan was in critical status and that certain benefits were being reduced. Such notices could have been sent multiple times over the years if the plan made subsequent benefit reductions or dropped into critical status more than once. Members also know the plan received SFA, which may give them the false sense that their benefits will now be secure forever. Now, you will be sending a new notice explaining a full benefit change going forward, with benefits that adjust with plan asset performance. It may be difficult for members to comprehend yet another change with a single correspondence. A communication campaign comprised of mailings, videos, personalized statements, and in-person meetings may be needed to allow members to fully understand the change and appreciate why it is needed.
Please visit our website to learn more about alterative plan designs and contact your Milliman consultant to discuss how a change in plan design can impact your plan.
1 If all assumptions from the SFA application are met.
2 Mechanisms can be put into place in a variable design to limit the chance that benefits ever decline.
3 These are the long-term, ultimate benefits that exist in a variable defined benefit plan design once all liability is associated with variable benefits. Until that time, traditional benefits will continue to be subject to the risks and challenges listed under the current plan design column.