Add to favorites
The calculation of expected benefits for plan participants of pension and OPEB plans requires the use of assumptions about the demographics of plan participants. Because the nature of benefits provided under pension plans differs from that of benefits provided under OPEB plans, the data needed may be slightly different between the two types of plans; however, many data points will be consistent. Examples of such demographic assumptions include estimates of:
  • Life expectancy or mortality
  • Expected retirement age
  • Employee turnover
  • Other factors such as probability of disability, marital status, and dependency status.
ASC 715 requires each significant assumption to reflect the best estimate of that particular future event (an "explicit" approach). Thus, employers may not apply an approach that looks to the aggregate effect of two or more assumptions, even if their aggregate effect may be approximately the same as that of an explicit approach.

2.3.1 Demographic data used in plan assumptions

The demographic assumptions of plan participants rely on categorical employee information. Therefore, employers will typically accumulate various information about individuals expected to be eligible for benefits under the plan, including active employees eligible or expected to become eligible, former employees who are eligible for benefits under the retiree plan (including disabled individuals), and retirees. Some of this so-called “census” data may include:
  • Date of birth
  • Gender
  • Date of hire
  • Business unit
  • Hourly/salaried employee
  • Marital status and spouse's date of birth (if spousal coverage is provided)
  • Retirement date
  • Salary (if plan is pay-related)

Data that may be required and would be unique to OPEB include:
  • Plan option selected (e.g., indemnity plan, HMO, or PPO)
  • Coverage offered through spouse's or other plan, including Medicare
  • Dependents and their dates of birth (if dependent coverage is provided)
  • Geographic location of employees and retirees

Some of the OPEB census data may not be available. For example, companies may not have updated information on dependents or their birth dates, or the existence of coverage through a spouse's plan. Depending on the nature and level of OPEB coverage, such data could be significant to the OPEB calculation. If not currently accessible, actuarial estimates of this data should be made.

2.3.2 Life expectancy or mortality assumptions

A key assumption in the calculation of expected plan costs is life expectancy or mortality—i.e., for how long will the plan be required to pay benefits as most arrangements provide benefits for life. As with any other assumption underlying an accounting estimate, employers should use the best available information to inform their selection of a mortality assumption. Generally, the most current published mortality tables should be used unless (1) an employer can demonstrate through an analysis of actual experience the need for higher mortality rates (shorter life expectancy) that may be reflected in older tables or (2) appropriate actuarial adjustments are made to the older tables to reflect trends towards the more current (generally lower) mortality rates. The expected mortality rates reflected in published mortality tables are based on historical mortality experience through the year of the mortality study (i.e., the base year). Adjustments to these tables are typically needed to reflect projections of expected future mortality improvements after the base year. Examples of typical adjustments made to a base mortality table include (1) projecting future mortality improvements after the base year and (2) blending separate male and female mortality tables into a unisex table in proportion to the male and female percentages of a company’s plan participant population.
ASC 715 does not prescribe use of a specific mortality table or mortality improvement scale. The mortality assumption should represent management’s best estimate of the expected duration of future benefit payments at the measurement date. Employers can consider using mortality tables that reflect the nature of their respective industries. For example, it would generally be inappropriate for an employer in a service-based industry to use mortality tables that were developed from employee data derived from the manufacturing industry. The estimate should be based on the facts and circumstances for each plan (e.g., population, demographics) and consider all available and relevant information at the date the reporting entity’s financial statements are available to be issued. The mortality assumptions for measuring defined benefit obligations under GAAP should be assessed independent of those specified for pension funding requirements as set by the IRS (for tax-qualified plans in the US).
Actuarial Standards of Practice 35, Selection of Demographic and Other Noneconomic Assumptions for Measuring Pension Obligations, requires actuaries to consider mortality improvements since the mortality table’s base year through the measurement date as well as projected improvements beyond the measurement date.
The Society of Actuaries (SOA) publishes mortality tables from time to time, which are generally considered to be reliable sources for developing mortality assumptions. These tables are based on observed experience over a defined study period and considering past trends, and reflect more recent data and the latest actuarial techniques.
The SOA also issues mortality improvement scales to reflect additional years of mortality data from the Social Security Administration between full studies leading to new mortality tables. Since 2015, mortality data shows a continued (although smaller) downward trend in mortality—i.e., increases (or improvements) in life expectancy. Companies should consider any published new mortality data for their plans in relation to their plan-specific mortality experience and future expectations.
Mortality assumptions are a long-term estimate of future experience, so we would not expect mortality improvements to be projected only through the current measurement date each year. That would be, in effect, phasing in a long-term assumption in annual increments over a long period of time, and inconsistent with the use of the current best estimate of future mortality experience. Further projections of mortality should reflect improvements (both from the base year of the mortality study to the measurement date and for periods after the measurement date) on a generational basis rather than using a static table projection (although a static projected table may be acceptable with significant additional support and analysis). Alternative custom-base tables may be supportable, for example, adjusting a standard published Society of Actuaries table based on company-specific recent historical mortality experience, to the extent there is sufficient data to be statistically credible. Regardless of base tables used, the improvement scale applied should generally be in line with most recent improvement scales available.
In addition to the published SOA standard improvement scales, other custom-designed mortality improvement scales may be supportable, such as those that incorporate data published by the Social Security Administration. Mortality tables for use in lump sum pension calculations

For qualified pension plans in the US that offer a lump sum benefit payment rather than an annuity, the IRS prescribes the actuarially equivalent interest rate and mortality table for determining the minimum statutory required lump sum payment. Pension plans typically require the use of the then-current IRS tables (or the "greater of" the amount calculated using such tables and the amount calculated using another specified set of tables) to convert the participant's stream of future benefits into a lump sum amount. For actuarial valuations of pension plans that offer lump sum payments, management is required to choose a mortality table and projection scale assumption for accounting purposes to determine the lump sum amounts expected to be paid at each expected payment date in the future.
Under current law, for payment of lump sums, mortality is required to be based on US retirement plan experience, and the Secretary of the Treasury is required to consider updating the lump sum tables at least every 10 years. However, there are no formulaic or automatic provisions required as part of that consideration, and the Secretary of the Treasury has latitude in implementing the updates. Thus, applicable mortality tables for subsequent years will be determined in the future based on the IRS’ approach of updating the tables when considered appropriate.
This discretion by Treasury gives rise to a question of how to project future lump sum payments over the life of the plan. We believe there are two acceptable approaches under ASC 715:
  • View 1: Future anticipated updates by the IRS to its lump sum mortality tables can be anticipated and reasonably estimated prior to future IRS regulations being promulgated (much like annual cost of living increases). This would be consistent with the general approach in ASC 715 to reflect the employer’s best estimate of future benefit payments, incorporating a wide variety of projections of future events. Management should reflect its best estimate of what it believes the IRS-prescribed mortality table will be in future years when the expected lump sum amounts are to be calculated and paid.
  • View 2: Future updates by the IRS to lump sum mortality tables is equivalent to a change in law, which should not be anticipated under ASC 715-30-35-31 and ASC 715-60-35-102. Management should assume that mortality rates used to calculate lump sum amounts paid in all future years will be based on the currently promulgated IRS tables. When the IRS updates their lump sum mortality tables, the impact on the projection of future lump sum payments would be reflected in the measurement of the obligation at the next measurement date following the IRS update. Mortality tables for promised future annuity purchases

Rather than paying benefits directly, some pension plans may promise to purchase annuities from an insurance company to provide the retirement benefit. In those cases, the mortality tables expected to be inherent in insurance companies' rates should be incorporated into the measurement of the plan’s benefit obligation. That would be true not only for special valuations, such as lump sum windows, retiree annuity purchases, and plan terminations, but also for ongoing annual valuations when lump sums or annuity purchases are assumed in the future.

2.3.3 Expected retirement age assumption in pensions

The assumption of retirement age—i.e., when benefits will begin versus when they will end (mortality)—can be more critical to the measurement of an OPEB obligation than to a pension obligation. Pension plans often provide for a reduced benefit if early retirement is elected to reflect the longer period over which the retiree is expected to receive a defined benefit. OPEB plans often do not have a similar provision. In addition, prior to age 65, there would generally be no Medicare reimbursement, resulting in a substantially higher annual per capita cost for OPEB plans. In general, the expected retirement age assumption should reflect an employer’s historical experience relative to its plan’s demographics and general economic trends.

2.3.4 Employee turnover assumption in pensions

Turnover represents the rate at which employees participating in the plan are expected to leave before becoming eligible for benefits (i.e., forfeiture rate). This assumption will have a significant impact in OPEB plans that specify that employees who terminate before the date they are eligible for benefits lose all OPEB benefits earned to date. Because pension plans may provide for vesting of benefits over relatively short periods of service, the measurement of costs and obligations for pension purposes is generally less sensitive to the turnover assumption than they are for OPEB purposes. See PEB 2.5 for a discussion of the impact of vesting provisions on the attribution period of service cost.

2.3.5 Disability assumption in pensions

Many employers provide lifetime health care coverage to employees who retire due to disability. For such plans, rates of disability and recovery are required to project benefit payouts. It may also be necessary to use per capita claims costs specific to disabled retirees if they are greater than for other retirees.

2.3.6 Marital and dependency status assumption in pensions

In addition to providing benefits to retirees, many OPEB plans provide health care benefits to retirees' spouses and dependents, often with no reduction in benefit levels. For these plans, spousal and dependent coverage can significantly increase the OPEB obligation otherwise payable for retiree-only coverage. In contrast, although pension plans may also provide for spousal benefits, they generally do so through a surviving spouse option, which typically provides reduced benefits to the retirees to take into account the additional payments expected to be made to the spouse subsequent to the retiree's death. Regardless, ASC 715 requires that an actuarial assumption be made about employees' expected marital status and number of dependents during retirement. The starting point for any such assumption is the accumulation of demographic data for current spouses and dependents.

2.3.7 OPEB plan assumptions

Additional assumptions used in the calculation of expected benefits for plan participants specific to OPEB plans may include:
  • Opting into Medicare coverage
  • Per capita medical claims costs
  • Healthcare cost trend rates Opting into Medicare

Under existing law, prescription drug benefits and certain health care benefits are available to individuals age 65 and older through the federal Medicare program. The estimation of the employer's future cost, therefore, includes a reduction of the assumed gross per capita claims cost to the extent that plan benefits are reduced for amounts expected to be either paid through Medicare or subsidized by a federal government program when determined to be "actuarially equivalent" to Medicare. Similarly, the assumed gross per capita claims cost is reduced by expected reimbursements of costs by others, for example, by a retiree’s spouse's OPEB plan sponsored by another enterprise. The Medicare and other provider reimbursement amounts must be estimated for each future year using the level of benefit coverage provided under the present law and/or existing provisions of the other plan. Only enacted changes in the law or the existing terms of other providers' plans, including amendments that are approved and that take effect in future periods, are to be considered in current period measurements for benefits expected to be provided in those future periods. Future changes in the law or other provider plans cannot be anticipated. Per capita medical claims costs

For benefits defined in terms of health care coverage, future benefit payments to be made by an employer to or on behalf of each plan participant must be estimated. These future benefit payments are referred to in ASC 715-60 as the assumed "net incurred claims cost at each age" and are derived by estimating the assumed (gross) per capita claims costs by age and reducing it by the effects of Medicare and coverage by other providers and the effects of the plan’s cost-sharing provisions (i.e., employee/retiree contributions, co-payments, deductibles).
Separate base period gross per capita claims costs should be developed at each age, and possibly also by gender, geographic location, and type of medical service (e.g., hospital care, physician services, prescription drugs).
Trend rates (see PEB are estimated for each future year through the last year the youngest current plan participant or dependent is expected to receive benefits. These rates are then applied to the base period cost for each of the applicable years during retirement to estimate the aggregate future claims cost during retirement. For example, to estimate the future cost of retiree health care coverage at age 70 for an employee who is 45 years old in 2019, the annual trend rates for 2019 through 2044 would be applied to the base period 2019 cost to estimate the cost for a 70 year old retiree. Some plans incorporate "caps" that limit the maximum OPEB benefit payable. See discussion of capped plans in PEB 2.4.4.
Participant cost-sharing amounts would also reduce the assumed gross per capita claims cost. These components of per capita claims cost are calculated by applying the cost-sharing provisions embodied in the employer's substantive plan to each year's projected assumed gross per capita claims cost. Certain plans require contributions by active employees toward their postretirement benefit coverage. In those cases, the actuarial present value of such contributions would reduce the actuarial present value of the aggregate assumed incurred claims cost.
The assumed net incurred claims cost at each age during the retirement period is estimated for each employee/retiree participating in the plan and is then applied, along with actuarial assumptions similar to those used in pension calculations, to estimate benefit payments for the entire participant group.
Internal and external costs directly associated with administering the plan would also be included as a component of the estimated gross claims cost, if significant. ASC 715 does not elaborate on the types of direct costs to be considered, but we believe they should be limited to incremental costs. Therefore, employee salaries and general and administrative expenses that would have been incurred even if no OPEB plan existed should not be included in assumed gross claims cost. Health care cost trend rates

The health care cost trend rates to be applied to the base period gross per capita claims costs represent the expected annual rates of change in the gross cost of the specific health care benefits provided under the plan. The health care cost trend rate assumption generally contains three components:
  • The initial trend rate, which is used to project current per capita claims costs to the next year
  • The ultimate rate, which is the rate at which health care cost trends will level off in some future year
  • The number of years and pattern of change between the initial and ultimate rates
ASC 715 requires that these rates be developed using past and present health care cost trends, which would implicitly consider estimates of health care inflation, changes in health care utilization and delivery patterns, technological advances, and changes in the health status of plan participants. It also notes that different types of services, for example, hospital and dental care, may require different trend rates.
The initial trend rate assumption should reflect the employer's recent retiree-specific health care cost experience, adjusted as appropriate for expectations of next year's costs. The ultimate trend rate assumption should reflect long-term expectations of future general inflation, plus some additional amount to reflect that retiree health care costs are expected to continue to rise at a greater rate than general inflation. The ultimate rate is limited by the general expectation that health care spending will not continue to rise at current levels forever. The pattern of decline in inflation rates and the number of years between the initial and ultimate rates is generally the most subjective component of the health care cost trend rate assumption.
Developing future trend rates begins with an analysis of past years' actual experience. If estimates of the assumed per capita claims costs are made for various categories of health care services, for example, hospital care and dental care, separate trend rates would be developed for each such category. This historical analysis would typically be developed from the same source as the data used for developing the base period gross per capita claims cost discussed in PEB The extent of the historical periods necessary to develop an appropriate estimate is a matter of judgment. It is also important to consider any recent plan changes that may reduce the usefulness of historical data.
When historical data is not available, or is not considered reliable or indicative of the plan's expected experience, trend rates can be developed from per capita costs of other employers, adjusted to best reflect the terms of the employer's plan and the demographics of the participants.
Estimates of future health care trends should not rely on history alone. The analysis of past trends is supplemented by assumptions about the magnitude and direction of changes in future trend rates from present rates. In developing estimates of future trends, it is important that preliminary estimates be tested against general inflation and productivity estimates to ensure that health care remains logically correlated to other economic factors. This approach is similar to that used in developing projections of the Consumer Price Index.
ASC 715-60-35-99 through ASC 715-60-35-101 provides general guidance for developing the trend rates assumption.

ASC 715-60-35-99

The assumption about health care cost trend rates represents the expected annual rates of change in the cost of health care benefits currently provided by the postretirement benefit plan, due to factors other than changes in the demographics of the plan participants, for each year from the measurement date until the end of the period in which benefits are expected to be paid. Past and current health care cost trends shall be used in developing an employer's assumed health care cost trend rates, which implicitly consider estimates of health care inflation, changes in health care utilization or delivery patterns, technological advances, and changes in the health status of plan participants.

ASC 715-60-35-100

Differing services, such as hospital care and dental care, may require the use of different health care cost trend rates. It is appropriate for that assumption to reflect changes in health care cost trend rates over time. For example, the health care cost trend rates may be assumed to continue at the present level for the near term, or increase for a period of time, and then grade down over time to an estimated health care cost trend rate ultimately expected to prevail.

ASC 715-60-35-101

An assumption about changes in the health status of plan participants considers, for example, the probability that certain claims costs will be incurred based on expectations of future events, such as the likelihood that some retirees will incur claims requiring technology currently being developed or that historical claims experience for certain medical needs may be reduced as a result of participation in a wellness program.

As in any actuarial valuation, a number of alternative scenarios may need to be considered before the most probable assumptions are identified. These may vary by the length of time the near-term trend pattern is expected to continue and the timing and degree of the grading down to the ultimate trend rate. It may not be unrealistic to assume that, at some point, the trend rate would approach the forecasted general inflation rate.
In the periodic reporting of OPEB cost, the difference between actual versus estimated rates of change in per capita claims cost, and between actual versus expected benefit payments in the year just completed, will result in actuarial gains or losses, as discussed in PEB 3. Accounting for these gains or losses is discussed at PEB 3.2.7. In addition, the most recent trend rate experience should be considered in the development of the trend rate estimates to be used in measurement of the following period's OPEB cost and obligation.
Rates of increase in health care costs experienced in the past and estimated to occur in the future will vary from employer to employer, and even from plan to plan within each reporting entity, depending on a number of factors:
  • Actual retiree health care inflation experienced in prior years
  • Expected cost increases—e.g., premium increases on the part of the insurance carrier
  • Type of health care coverage offered—traditional indemnity, health maintenance organization (HMO), or preferred provider organization (PPO)
  • Specific categories of services covered—major medical or basic medical, prescription drugs, etc.
  • Demographics of the covered group—age, gender, and geography
  • Effectiveness of cost and utilization controls, such as contracts with specified providers, catastrophic case management, and hospital pre-admission reviews
  • Regulatory changes that may affect the cost of providing health care (e.g., the Affordable Care Act)
Accordingly, and consistent with the base period per capita claims cost assumption, the health care cost trend rates assumption must be developed on a plan-specific basis.
As health care costs continue to rise, assumed health care cost trend rates have a significant impact on postretirement benefit obligations and the periodic expense related to those obligations. Employers should reevaluate the health care cost trends rates as part of each actuarial valuation to ensure they reflect the best available information as of the measurement date.

Welcome to Viewpoint, the new platform that replaces Inform. Once you have viewed this piece of content, to ensure you can access the content most relevant to you, please confirm your territory.

signin option menu option suggested option contentmouse option displaycontent option contentpage option relatedlink option prevandafter option trending option searchicon option search option feedback option end slide