In recent decades, there has been a dramatic shift in US corporate retirement programs from defined benefit (DB) to defined contribution (DC)¹. Under the pretense of risk management and cost management for sponsors, this has shifted significant risk to employees and added cost for retirees who wish to secure a portion of their retirement income by purchasing annuities².

As employees and retirees feel the impact of this shift in risk, and those approaching or entering retirement feel less secure in their outlook³, many sponsors are re-examining their retirement programs.

Policymakers recognize the challenges associated with this shift in risk and, as a result, have passed legislation through the SECURE Act and SECURE 2.0 to empower sponsors to improve retirement programs to address the key risks participants face.

In this paper, we address the various approaches across DB and DC plans for providing a guaranteed (annuity-backed) income stream through the lenses of plan sponsor costs, participant outcomes, and the balance of risk. In doing so, we aim to help plan sponsors understand how longevity, market, and inflation risk can be reduced for participants.

Many factors will ultimately influence finding an optimal solution for a given participant population and will be dependent on evaluating specific sponsor and participant situations. A comprehensive evaluation should include assessment of the financial impact on the plan sponsor, fiduciary considerations, and participant retirement outcomes.

We considered a range of program types, as summarized in Table 1, ranging from a full DC program invested in a moderate target date fund glide path and no guaranteed income beyond Social Security, to various program structures that provide guaranteed income post-retirement. These “hybrid” designs can incorporate DC-type market returns or interest credits on notional accounts in DB plans (cash balance), or add annuity options to DC plans. It is well understood, and our analysis supports, that a component of guaranteed lifetime income (an annuity-backed income stream, as part of a broader retirement income program) in retirement significantly decreases the likelihood of outliving assets for long-term retirees. For DC plans, we have seen more annuity-backed solutions launched and funded in the past few years, from target date funds with an annuity component to institutional annuity purchase platforms available upon withdrawal. For DB plans, sponsors may be considering how to effectively use a surplus (if available), or how to manage the design of a plan that aligns with the level of balance sheet risk they are willing to accept. Against this backdrop, many plan sponsors wish to understand the optimal way for their organization to incorporate guaranteed lifetime income in a comprehensive retirement program, whether it is provided in a DC or DB construct. The analysis that follows can be customized to sponsor-specific situations, retirement program designs, and participant populations to assist in this important decision.

_{¹US Department of Labor, Employee Benefits Security Administration (EBSA, Private Pension Plan Bulletin Historical Tables and Graphs: 1975-2019, September 2021}

_{²Cumming, D., Lu, F., Xu, L et al. Are Companies offloading Risk onto Employees in Times of Uncertainty? Insights from Corporate Pension Plans. J Bus Ethics (2024)}

_{³Retirement Insecurity 2024: Americans Views of Retirement (Greenwald Research)}

Plan Sponsor Costs

There are a number of costs to the plan sponsor while running a retirement plan program which may include various administrative expenses, plan sponsor contributions, fiduciary monitoring oversight costs, and third-party fees such as consulting and legal. In order to assess the different program types in a consistent fashion, we have reviewed and included only the hypothetical sponsor costs that would impact the company income statement and balance sheet.

In a DC program, for instance, the sponsor cost impact will align with the level of employer contribution.

A market-return cash balance is typically invested in a portfolio that aligns with the target allocation for the cash balance interest credit, and the employer cost is therefore aligned with the notional crediting rate.

For cash balance plans that credit interest using a Treasury-based rate, the employer cost can vary from the notional contribution. In this case, a higher notional credit rate can be used, and the employer can defray some of the cost by earning a higher return – with corresponding balance sheet risk for the plan sponsor. A Treasury-based interest credit results in market risk for the plan sponsor, given that crediting a notional return above cash (either a fixed spread, or a longer-duration yield) cannot be directly hedged. Over the long term, sponsors can expect to out-earn the Treasury-based interest credit with a high likelihood (we have assumed sponsor returns 2% above the crediting rate), but short-term results may vary and market risk is an important consideration for sponsors.

The five program designs below are intended to be cost-neutral to employers under median long-term economic projections

Table 1: Details of Retirement Program Types Reviewed

_{¹ Administrative expenses for different program types will vary by sponsor size. For this analysis, the administrative expense across programs is assumed to be consistent; sponsor-specific analysis would incorporate expenses in assessment of retirement program alternatives.}

_{² Assumed participant deferrals are consistent across the retirement program types, ranging from 6% in early career (age under 30) to 10% in late career (age 60+).}

_{³ Sponsor cost is lower than the crediting rate (assumes pay credit is 4.5%) due to excess investment returns. The sponsor takes on investment risk in this structure.}

Aiming to Improve Participant Outcomes

The objective of an effective retirement program is to allow employees to fund a reasonable retirement budget at reasonable cost. Active participation in the retirement program, through sustained employee savings over a career, will be a key factor in achieving a good level of income replacement. Such savings can be encouraged by sponsors through providing an employer match, automatic deferrals, auto-escalation and good participant communications, but meeting retirement objectives will ultimately depend largely on employee behaviors.

For employees who save consistently at sufficient levels to enter retirement with a significant DC account balance, the uncertainty over market risk, longevity risk (the risk of outliving assets), and inflation risk can make retirement planning extremely challenging. While some will succeed, many employees may fall short of their retirement objectives as they underestimate the level of wealth needed to sustain a healthy and comfortable retirement. We find that there is opportunity to significantly enhance the likelihood of the retirement nest-egg lasting through a full retirement by incorporating a lifetime income component. The value of the income guarantee becomes apparent as the post-retirement time horizon is extended. A plan sponsor can expect many participants to live 30 or 40 years after retirement, or longer. The guaranteed income component can take multiple forms, but an annuity that offers a consistent income stream may help participants mitigate longevity risk and reduce the shortfall to budget under downside economic scenarios as compared to a participant with no guaranteed income beyond Social Security.  This is observed in the table below:

A driving factor in the ability to meet a real income budget over a long-term retirement is the level of equity allocation. Equities provide the strongest source of long-term growth in the accumulation phase and help to reduce the likelihood of outliving assets or depleting assets relative to a real spending budget post-retirement. It’s important to retain a meaningful equity allocation to limit the chance of asset depletion over a long retirement. When an annuity through an insurer or cash balance plan is provided, the remaining DC asset allocation should be more aggressive (in other words, the annuity can be considered a portion of the defensive or fixed income assets). As a result, we modeled the cash balance options with an “aggressive” target date fund glide path that lands at a terminal 60% equity allocation. While aggressive compared to off-the-shelf funds, a target date fund glide path customized to consider the presence of a cash balance plan would allow a sponsor to optimize the DC glide path within their overall retirement program. Similarly, the DC program with an immediate annuity at 65 is assumed to have a 50% equity allocation for the remaining assets from that point forward. The longevity insurance (deferred annuity at age 78) maintains the moderate glide path of 33% equity allocation from that point.

We observe that cash balance programs with a conservative interest crediting rate (e.g., T-bills plus 1%), or conservative market-based allocations, sacrifice returns in the accumulation phase and that can offset the benefit of lifetime income post-retirement. When the target date fund asset allocation considers the fixed income-like value of the cash balance and retains a more aggressive allocation, the overall retirement program provides better long-term outcomes than the DC program with no income guarantee. It is important for sponsors with a defined benefit or cash balance program to factor this income-like component into the target date fund asset allocation decision (be it custom or off-the-shelf ) in order to potentially enhance long-term participant outcomes.

Balancing Sponsor and Employee Risk

Ultimately, individual longevity risk in a DC program is likely to create large disparities in outcomes. Target date funds have become the prevalent qualified default investment alternative (QDIA), which can drive similarities in the market returns achieved. Individuals may experience inflation differently depending on their spending mix and blend of real versus nominal costs, but the overall inflation trend will drive similarity in experiences with unexpected inflation. In contrast to the broadly felt economic impacts of market returns and inflation, individual longevity may vary significantly, and predicting and managing over a retirement time horizon that could last less than 10 or more than 40 years is extremely difficult. An annuity option significantly improves the expected shortfall to budget under downside scenarios over the long-term and reduces the likelihood of outliving assets when compared to spending down the defined contribution assets. 

Spending rates across all scenarios are based on a consistent set budget level, growing with price inflation. "DC - no income guarantee" assumes no guaranteed income beyond Social Security benefits.

Balance of Market, Inflation and Longevity Risk by Program Type

Conclusion

Market risk, inflation risk, and longevity risk are three major risks impacting long-term retirement outcomes. We believe that the pendulum has swung too far in shifting risk from plan sponsors onto employees, creating concern over potential disparate participant outcomes. At the same time, reverting to traditional DB programs is likely to put excessive market risk and cost back on plan sponsors and may even reduce the incentive for employees to save.

There are multiple approaches gaining traction with plan sponsors to address these concerns and strike the right balance; helping participants manage some of this risk will create a better balance of risk-sharing between sponsors and participants. Plan sponsors can help vet annuity providers and cost structures for in-plan offerings, and leverage institutional risk pooling and pricing, rather than leaving participants on their own to wade through a potentially complex and murky individual annuity decision. Depending on the size and scale of the employee population, this pooling of risk may be better accomplished with an insured approach (e.g. target date funds with annuities, or in-plan institutional annuity purchasing platforms) or a defined benefit (cash balance providing an annuity option). Undertaking customized modeling of various options based on sponsor specifics can be useful in finding the most appropriate direction for a given population. Ultimately, this can be viewed similarly to the way employer health benefits are provided – where small to mid-sized employers may opt for a fully-insured plan (pooling risk with other organizations through an insurance intermediary), while large sponsors often have a sufficient risk pool to self-insure and minimize costs. Employers large enough to self-insure their health plans or utilize captive insurance for other risks may also be candidates for self-insuring their participants’ longevity risk with a defined benefit program – a DB plan remains an efficient way to help provide guaranteed lifetime income to participants.

With these approaches to guaranteed retirement income, efficient delivery of retirement programs combined with effective risk management can help to improve both expected and downside outcomes for participants. Employees who are more secure in their retirement planning help sponsors manage their workforce effectively, and potential disparities in outcomes can be meaningfully reduced by helping those who live the longest avoid outliving their retirement assets.