Definition of Retirement Plan Resilience
Measuring the fiscal health of a retirement system is a complex project. In our work at Equable, we define financially healthy, sustainable retirement systems as those that are both resilient and affordable. Retirement systems also need to provide benefits in a transparent and responsible way (a set of activities we broadly define under the rubric of “accountability”). Plus, the benefits should to provide a path to retirement income security for all participants in the system. But without resilience, there are no funds to manage in an accountable way or use for providing retirement income.
Each year Equable Institute publishes an update on the fiscal trends and financial status of statewide retirement plans called State of Pensions. Throughout that report we make reference to resilience and fragility. In this article, we lay out a more complete definition of resilience for public retirement systems and provide a rubric for thinking about both measuring the status quo and identifying areas for improvement.
We care about resilience because for governments to thrive and be effective they need their programs and services to be prepared for negative shocks and tail risk events. Public sector retirement systems should not be prone to significant destabilization due to external factors, mismanagement, or political whims. Further, retirement systems should be able to adapt to changing economic and demographic conditions in both the short- and long-term. They should not be designed in ways that allow threats to build over time, to threaten the retirement security of members, or to impede the fiscal stability of their sponsoring government’s budgets.
The resilience of a retirement plan is more than just how much benefits cost. Of course affordability matters. But insofar as affordability is measured by comparing normal benefit costs (as opposed to considering normal costs plus unfunded liability payments), most decisions are a matter of policy choices that state or local leaders can make depending on their workforce goals and political preferences. What matters for affordability is that costs are known to all stakeholders and that they are paid for within existing budgetary resources.
Managing risks and designing governance systems to avoid politically misaligned incentives involves a range of cost and policy trade-offs. However, for states and cities it is worth reviewing carefully the fragility of their retirement systems. We believe states should work to reduce tail risks while improving their ability to respond to changing circumstances and be resilient in the face of negative events (financial, demographic, or ecological) when they come.
What Resilience Means
The dictionary definitions for resilience emphasize the ability to absorb a blow and recover quickly. Something that is resilient will bend, but not break. In some respects it is the opposite of fragility.
When thinking about the finances for a public retirement system, resilience is a way to think about how that system responds to external negative stressors, such as the shock of a financial market downfall, or a considerable change in demographic outcomes, or even just missing out slightly on one or two actuarial assumptions.
It is reasonable that large, institutional investors like pension funds with billions (and in some cases hundreds of billions) in assets under management will suffer financial downturns. Markets are unpredictable and the global financial landscape is more interconnected than ever. But it is also reasonable to expect public fund managers to develop counter-cyclical asset management strategies that aim to match the long-term duration of liabilities in defined benefit plans.
To the degree that asset management techniques can only be part of the approach to establishing resilience, it is incumbent on policymakers and trustees to ensure there are adequate risk management tools available to address sustainability problems when they arise. Such risk-sharing tools have become increasingly popular in recent years, ranging from automatic contribution rate increase programs to benefit adjustment provisions.
Ultimately, whatever the mix of techniques, tools, and personnel that are utilized to develop a plan for negative events, the key is that public retirement plans should be able to absorb whatever blows come their way and recover quickly.
How Resilience Should Be Measured
Measuring pension plan sustainability means looking at both solvency levels over time (funded ratios and unfunded liability levels), as well as the costs of providing the retirement plan relative to existing tax revenues. The larger required pension payments are relative to the size of state budgets, the harder it is for the state to ensure responsible funding policies because of the higher cost burden. If such costs are growing because of consciously made policies that set benefit levels in a known way with clearly allocated resources to pay them, this might not be a problem. But when larger payments are required because unfunded liabilities are growing, that represents a funding threat.
Given that measuring resilience should not be dependent on a single metric, here is a multi-factor framework for determining which plans currently have a resilient financial condition, which are fragile, and which are broken and in some kind of distress. There are five factors that we consider in setting this measurement:
- Funded Ratio (Funded Status)
- Unfunded Liability as a % of GDP (Ability to Pay)
- Assumed Rate of Return (Underperformance Risk)
- Share of Required Contributions Received (Willingness to Pay)
- Risk-Sharing Tools (Future Flexibility)
For details about each factor jump to notes at the end.
Beyond a Simple Framework
The framework we’ve articulated here is only a starting point for thinking about the resilience of a retirement system. It may be reasonable to weight some of the factors on importance, or to consider how some might off-set each other (like how risk-management tools might allow for slightly higher assumed returns). And there are other considerations beyond these factors too.
For example, complex demographic and economic conditions in each state that create important context: Are there incentives that will lead to a wave of retirements that could exacerbate cash flow concerns? Are there other economic pressures that have indebted a state that might constrain their ability to provide needed contributions? Are recently adopted tax laws projected to dramatically improve local revenues?
Cash flow is also important when assessing how distressed a plan might be: What is the ratio of cash to benefit payments? What is the ratio between contribution inflows and benefit payments? At what point will these ratios mean holding an undesirable portion of assets in liquid investments in order to meet benefit payments? At what ratio level should plans recognize that they are within just one or two bad investment years of insolvency?
Asset allocation strategy and investment risk matters too: How dependent is a public plan’s portfolio on particular asset classes? Does the fund’s strategy for using active asset managers allow trustees flexibility? Is a sufficient portion of the portfolio being used to provide counter-cyclical hedges? Is the strategy aligned with the duration of liabilities? How have investments performed compared to a common benchmark like a simple, balanced portfolio of stocks and bonds that isn’t actively managed?
And the details of funding policies also can’t be overlooked. The process of how a board of trustees receives information can influence decision making too. For example, a funding policy that uses “open amortization” over a multi-decade period of time with a high payroll growth assumption because it isn’t reconsidered every two years or less could be inappropriately minimizing their required contribution rate and squeezing into the fragile status category when distressed status is more appropriate.
Ultimately, this resilience framework should be used to identify plans that are in need of some kind of improvement to funding policy, risk-sharing policies, fund management, and/or decisions made by trustees. Stakeholders in public sector retirement systems should not be apathetic toward persistent levels of underfunding. Funding shortfalls make it harder to accumulate the returns needed to get back to full funding. The longer those shortfalls persist the more expensive they make funding the system, which creates political pressures.
Every retirement system has a lot of moving parts that should be reviewed carefully on their own, and then put together as a group. The purpose of this kind of national analysis that buckets off state plans is to identify who should get a closer look, rather than to be a final word on who is good or bad. And the reality is that even plans with resilient funded status can’t just sit back and become complacent. The world is constantly changing all plan leaders must be forever looking at how to adapt to the future.
 In separate articles we will more fully define how we break out the definitions of each of these terms. In short, we think about plans through the prism of “sustainability” (that is resilience and affordability), “accountability” (that is transparent governance and fiduciary responsibility), and “retirement security” (that is the plan is accomplishing its goals and providing a path to retirement income security for all participants).
 Here is some of the logic behind where we’ve set each threshold and why each factor matters. There is no magic to using round numbers. In practice setting the upper bound for the funded ratio at 92% or 88% might be as reasonable as 90%. (ERISA sets higher funding rules for private pension plans that dip below 90%, so we’re not alone in landing on this round number logic.) But the degree of where each threshold is set does matter.
(1) Funded Ratio
→ Resilient: If the funded ratio has been 90% or above for three years in a row
→ Fragile: If the funded ratio is above 60% but never consistently above 90%
→ Distressed: If the funded ratio is 60% or below
Logic: Plans that have had a funded ratio above 90% consistently are going to be in a strong position to recover from financial downturns, as funding policy improvements are easier to make when the plan’s finances are stable. There isn’t particular magic to this needing to be three years in a row, a range from two to four years consistently is reasonable. The important aspect is that the funded ratio is around 100% funded consistently.
By contrast, pension plans with a funded ratio consistently between 60% and 90% are in a more fragile condition. While these plans aren’t going insolvent any time soon, they will be building up unfunded liabilities that will gradually become a strain on budgets and government revenues. A plan that is 85% funded for several years in row is healthier than one 65% funded, but it is still exposed to risk. One or two severe asset shocks could send the plan into a downward spiral.
Pension systems with funding levels below 60% should be looking to make immediate steps toward fixing their problems. It shouldn’t matter how many years they are below this threshold because they either suffered a massive shock or they’ve been in a fragile condition for a number of years already. While the specific threshold for when these plans may spiral into insolvency will vary, at a certain point it is nearly impossible to return to fiscal health without an injection of capital from the government plan sponsor.
(2) Unfunded Liability as a % of GDP
→ Resilient: A ratio of 5% or less
→ Fragile: A ratio between 5% and 10%
→ Distressed: A ratio of 10% or higher
Logic: The size of a retirement plan relative to the state matters when considering just how fragile or distressed that plan might be. One way to measure this is comparing the dollar amount of unfunded liabilities to a state’s economy, which is a rough metric of its long-term capacity to pay down that funding shortfall. Unfunded liabilities as a share of a state’s economy gives a sense of what scale of resources will be needed from a local tax base to improve funded status.
For example, a large unfunded liability in a state with a growing economy might be less concerning than a smaller dollar amount of unfunded liabilities in a state with a contracting taxpayer base.
If the dollar amount of unfunded liabilities is relatively small compared to a state’s — or sponsoring local government’s — capacity to pay down that shortfall, then the underlying system has a greater probability of surviving a market shock. By contrast, an unfunded liability that is a large percentage of a state’s capacity to pay it down is a particularly large problem not only because of the way it might pose a threat to the sponsoring governments finances but because the retirement system itself is less likely to receive the kind of contributions needed to help with recovery from a market shock.
Putting this in real terms: the $283 billion in California unfunded liabilities (as of 2020 reporting) are 11x the $41 billion funding shortfall for plans in Kentucky. But those same unfunded liabilities in Kentucky are harder for the state’s taxpayer base to pay (19.5% of GDP and 352.8% of state general fund expenditures) than they are in California (9.2% of GDP and 193.2% of state general fund expenditures).
Another similar metric is actuarially determined employer contributions as a percentage of state budgets (or own-source revenue). A state’s capacity to pay for retirement benefits out of existing revenues is not necessarily a key metric when measuring resilience, as there are many policy reasons why the value of benefits might be set at high levels (translating to high normal cost prices). If a government wants to provide large benefits that require large portions of their budget and these costs are known and accounted for, that could be an appropriately stable situation. However, a state’s — or sponsoring local government’s — capacity to pay for growing unfunded liability amortization payments is an important part of resilience because when contribution rates are pushed up due to funding shortfalls and this puts pressure on government budgets it makes it less likely that retirement systems will continue to receive necessary funds. And such potential underfunding of contributions can be both explicit skipping out on required payments, or implicit behavior — such as not adopting a reasonable investment assumption in order to avoid how that might trigger a higher required contribution rate.
(3) Assumed Rate of Return
→ Resilient: 6.5% or below
→ Fragile: Between 6.5% and 7.25%
→ Distressed: 7.25% or higher
Logic: The most important actuarial assumption is the assumed rate of return (or discount rate) used to determine contribution rates for a retirement plan. This rate is important with respect to resilience because it signals the reasonableness of a plan’s measured funded ratio and unfunded liabilities. Retirement plans that report 90%+ funded ratios consistently but that use 8% assumed returns are not appropriately measuring their liabilities and are likely not as to be reflecting the resilience or fragility of a retirement system. So while the assumed return on its own is no more a perfect measure of resilience than any other, it is a particularly good signal.
Exactly how to define a resilient assumed return is a matter of much debate. If assumed returns for public plans had declined at the same pace as interest rates over the past two decades, then the average today would be a 4.3% assumed return instead of 7%. That would also put it more in line with private pension assumed returns.
Capital market forecasts for the next decade suggest that 5.5% might be the best current benchmark for a reasonable assumed return, if you’re looking for the average expected return (e.g. 50th percentile) on a standard public pension plan’s asset allocation (as measured by the Horizon survey).
All of which is to say that 6.5% is arguably a tough high when thinking about what a resilient assumed return would be. It could easily be argued that resilience in this context should be taking stock of the current financial landscape and selecting a rate that is conservative within that outlook. Which could translate to setting this threshold for resilience below even a 6% assumed return.
Going forward, it would be appropriate to link this assumed return threshold to interest rates, with some flexibility to factor in capital market forecasts. Our current framework would allow for interest rates to rise another 150 basis points and still reduce the threshold for a resilient assumed return to 6%. But if interest rates were to jump even higher, then it might be appropriate to nudge that measurement rate up.
(4) Share of Required Contributions Received
→ Resilient: 100% or more of the ADEC paid for at least 7 of the last 10 years
→ Fragile: At least 85% or more the ADEC paid in the last decade
→ Distressed: Less than 85% of the ADEC paid in any year without otherwise satisfying the definitions for Resilient or Fragile
Logic: It is important that retirement systems have protection against political whims that might lead to underfunding behavior. States that do not consistently pay required contributions create more fragile retirement systems. For example, a pension fund that is 75% funded and at least getting the full ADEC each year is going to be in a better position to respond to a market shock than a similar pension fund with the same funded ratio but that isn’t getting paid the full ADEC.
(5) Risk-Sharing Tools
→ Resilient: When there is negative experience for the plan contribution rates and/or benefits are automatically adjusted based on a preset criteria; OR the board has the authority to increase contribution rates (for employers and members) and/or modify benefits (e.g., adjust COLAs, permanent benefit increases, variable benefit increases, or prospective changes for active members within the limits of legal protection)
→ Fragile/Distressed: No risk-sharing or risk-management tools
Logic: Sometimes exogenous experience can threaten even the best plans that are receiving all required contributions, have strong funded ratios, and use reasonable assumptions. In such cases, a hallmark of a resilient plan is one where there are tools available to try and remedy the situation. Such tools could automatically kick-in based on certain triggers or guardrails, or a board of trustees might be pre-authorized to make decisions as necessary to keep a plan in a resilient condition.
An example of such a tool would be Wisconsin Retirement System’s variable benefit program, which gives the board authority to make certain decisions when necessary to keep the plan at 100% funded ratio levels. Another example is Colorado Public Employees Retirement System’s automatic increases made to active member contribution rates, employer contribution rates, and cost-of-living adjustments any time actual contribution rates drop below actuarially determined rates.
Another benefit of having risk-sharing or risk-management tools defined ahead of time is the avoidance of political fights related to improving funded status. Making painful or difficult decisions that are in the best interests of the retirement plan and its members is not always politically viable. So having such tools already defined before a crisis has a lot of value.