Frozen pension plans: Is immunization or termination the right choice?

Vanguard research

Executive summary. Is it more cost-effective to immunize or terminate a frozen defined benefit pension plan? The answer depends on a number of quantifiable variables, such as the cost of purchasing a group annuity for plan participants, as well as on qualitative considerations such as the plan’s demand on management’s time. We identify the primary variables affecting the cost of immunization and termination, and we analyze three hypothetical pension plans to show how changes in these variables would affect the relative cost of each strategy. We find that, in general, the costs of immunization and termination are similar. Given the similarity in costs, the qualitative aspects such as management time spent on the frozen plan and administrative or legal hassles will likely steer more plan sponsors toward termination. We recommend that plan sponsors review each of the cost variables as well as the qualitative considerations unique to their case when deciding whether to immunize or terminate a frozen pension plan.

Connect with Vanguard > vanguard.com

November 2012

Authors Nathan Zahm, FSA R. Evan Inglis, FSA, CFA

It’s no secret that traditional defined benefit (DB) pension plans have been on the decline over the past two decades, with approximately 50% of private-sector plans now closed to new entrants or frozen entirely as of 2010.1 For frozen plans, a sponsor may wonder whether the plan should be terminated or whether it is possible to immunize the plan’s liabilities with matching assets at a lower cost (see accompanying box defining immunization and termination for pension plans).

As a result of the federal Pension Protection Act (PPA) of 2006, it is now less expensive to terminate a plan. Prior to the PPA, plan sponsors used 30-year U.S. Treasury bond rates to calculate the value of lump-sum payments to participants. The 2006 law allows sponsors to use the higher corporate bond rates to calculate the value of these lump sums. The use of these higher discount rates reduces the cost of lump sums and, consequently, terminations. This paper compares the cost of setting up a portfolio of assets to fully (or almost fully) hedge a plan’s liabilities with the cost of terminating the plan. We conclude that the cost of the two approaches is very similar, but that considerations that are hard to quantify tend to favor plan termination. A variety of factors, including asset performance, the size of the plan, annuity costs, and the level of lump-sum acceptances by participants affect the cost of each approach. Plan sponsors with a frozen pension plan may want to consider how the factors discussed in this paper apply to their particular situation.

Factors affecting the decision Plan demographics are a significant driver of cost differences between immunization and termination. To account for the influence of demographics, we look at three hypothetical plans representing different types of plan populations (see Figure 1). • A typical plan: The total liability is split equally among active, terminated-vested, and retired participants. • A mature plan: The majority of the liability is associated with retirees “in-payment.” • A young plan: The plan predominantly has liability for active participants and only small obligations for terminated-vested participants and retirees. This situation often occurs in plans that have offered a lump-sum option for many years. Because plan sponsors consider their pension liability the way it’s defined in accounting or PPA rules, we use that definition of liability as a reference point in this paper. We refer to it as the “ongoing plan liability.” This liability can be thought of as a typical pension benefit obligation (PBO) accounting measurement for a pension plan or a PPA Funding Target (FT) measurement calculated according to the yield curve parameters established in the law. By using these common methods to determine a market-consistent pension liability as a reference point, we can better illuminate a concept critical to the comparison—bond underperformance relative to the liability.

Immunization versus termination For purposes of this paper, we define immunization and termination as follows: Immunization. The purchase of assets that secure the benefit payments. These assets typically include long-duration, high-quality corporate bonds and U.S. Treasury securities whose cash flows match those of the pension plan in timing and amount. Termination. The complete wind-down of the pension plan and the distribution of all plan assets to participants through lump-sum payments and/or the purchase of a group annuity contract that provides annuity payments to plan participants. Termination eliminates all future obligations for the employer.

1 Source: Information from Pension Benefit Guaranty Corporation filings made available to Vanguard.

2

Figure 1.

Ongoing liability for three hypothetical frozen plans: Typical, mature, and young

(Liability in $ millions)

Plan type



Typical Mature Young

Active liability

$166.7

$100

$400

Terminated-vested liability

166.7

100

50

Retiree liability

166.7

300

50

Total liability

500.0

500

500

Source: Vanguard.

In addition to the plan demographics, a variety of quantitative and qualitative factors affect a cost comparison of the two approaches. We identified and tested the impact of seven critical variables, three for immunization scenarios and four for termination scenarios. For immunization scenarios, we examined: • Bond underperformance relative to the liability. • Administrative expenses. • The offering of lump sums. For termination scenarios, we examined: • The cost of a group annuity. • Lump-sum acceptance rates. • Enhanced early retirement benefits and other plan designs. • The offering of lump sums for retirees. For each variable, we started with a baseline estimate based on empirical values—historical measures of bond underperformance, for example. We then altered the variable to assess how changes in its value would affect the cost of immunization or termination. We also considered the size of a plan, which affects the relative level of administrative and other expenses. The costs for a large plan are smaller as a percentage of liability than for smaller plans.

Other considerations Some important aspects of the comparison between immunization and termination are difficult to quantify. These factors will generally tilt the comparison favorably toward termination. For example, we did not quantify the level of demographic risk2 or reinvestment risk3 that would remain with a plan in an immunized approach but be transferred to participants and an insurance company in a plan termination. Nor did we attempt to quantify costs internal to the plan sponsor, such as time spent managing the plan and making decisions about it. We describe these factors and others at the end of this paper.

Baseline results When a pension plan terminates, active participants, terminated-vested participants, and, occasionally, retirees are offered the option to take a lump-sum benefit rather than the annuity benefit. Typically, active and terminated-vested participants are more likely to choose a lump sum, and retirees are more likely to keep their annuities. The liability is paid off through these lump-sum payments and the purchase of a group annuity from an insurance company to provide the annuity benefits. It’s this combination of methods for settling the pension obligation in a termination that makes the comparison of costs for immunization versus termination interesting.

2 Demographic risk is the risk that participants’ retirement behavior, benefit elections, longevity, and so forth, are different from the actuarial assumptions. 3 Reinvestment risk is the risk that the yield on a future investment will not be the same as the yield on a current investment that matures.

3

Figure 2.

Baseline costs of immunization versus termination for typical, mature, and young plans in 2012

(Liability in $ millions)

Plan type Typical Mature Young

Baseline immunization liability

$537.6

Baseline termination liability Termination/immunization ratio Lower-cost alternative

$536.9

$540.0

$532.8

$537.4

$530.3

99.1%

100.1%

98.2%

Neutral

Neutral

Termination

Source: Vanguard.

In 2012 (and going forward), the cost of terminating a plan is not significantly different from the cost of keeping a plan with an immunized portfolio (Figure 2). Although there are excess costs related to an insurance company’s need to earn a profit and be compensated for the risk of guaranteeing a group annuity contract, the ability to settle a large portion of the obligation with lump-sum payments offsets this excess cost. Lump-sum payments represent a savings relative to the cost of providing pension annuity benefits, for various reasons. One reason is that the corporate bond rates used to determine lump-sum amounts undervalue the liability—this is the result of the credit downgrade phenomenon described in the upcoming section on bond underperformance relative to the liability. Lump sums also eliminate longevity risk and ongoing plan expenses. Figure 2 shows the “immunization liability” and “termination liability” for the three different plans described in Figure 1. Each plan is assumed to have an ongoing liability of $500 million. An “immunization liability” and a “termination liability” are calculated by adjusting the ongoing liability. The baseline liability from Figure 2 is identified for each plan using estimates of typical values for the seven quantitative variables that drive immunization and termination costs. We provide more details on these variables in Figure 3. (Appendix Figures A-1 and A-2, on page 16, include additional details about the assumptions used for the liability calculations.)

4

To calculate the impact of changes in each variable on cost, we modified the discount rate used to calculate plan liabilities. If a plan’s ongoing administrative expenses rose, for example, we reduced the discount rate and thus raised the cost of plan immunization. Insurance companies generally calculate premiums for group annuity contracts this way. We compared the costs of immunization and termination on this present-value basis. We calculated a “termination/immunization ratio” by dividing the termination liability by the immunization liability. When the liabilities were within 1 percent of one another, we assumed the decision to immunize or terminate the pension plan was cost-neutral. Otherwise we identified the lower-cost alternative. Figure 2 shows that for a plan with “young” demographics, termination is likely to be the lowercost option, but for the “typical” and “mature” plans, the cost comparisons are very close. This makes sense because a plan with more active and terminated-vested participants will generally have more lump-sum acceptances when terminating, and immunizing those participants would be relatively more expensive because of the longer time horizon. Figure 3 describes the variables that affect these cost calculations and shows the baseline values used to develop the comparisons in Figure 2.

Figure 3.

Descriptions and baseline values and assumptions for variables affecting immunization and termination costs

Termination variables

Immunization variables

Variable

Baseline value/assumption

Description

Plan sponsor considerations

Bond underperformance relative to the liability

80-basis-point (bp) reduction in discount rate.

Neither corporate nor Treasury bonds will have long-term returns as high as the discount rates used to measure ongoing pension liabilities.

Bond underperformance generally occurs during waves of economic and credit stress, followed by longer periods of stability.

Plan expenses

10-bp reduction in discount rate.

Administration, actuarial fees, Pension Benefit Guaranty Corporation (PBGC) premiums, etc., are included. This factor tends to vary by plan size.

Plan expenses are generally a higher percentage of the liability for smaller plans.

Lump sums

Offered to active and terminated-vested participants in both immunization and termination alternatives.

Lump sums are the least expensive way to settle the pension obligation, so they are assumed for both termination and immunization.

Offering lump sums in plans that aren’t terminating is a way to reduce liability, headcount, and expenses.

Annuity purchase costs

110-bp reduction in discount rate for retirees and 150-bp reduction in discount rate for active and terminated-vested participants.

Group annuity insurers capture the cost of risk, administration, and profit by adjusting the discount rate. Retirees are less expensive because the amount of payments is more predictable.

The annuity purchase is the most expensive part of terminating a plan. Sponsors should receive quotes for their plan, as pricing can vary based on the insurance company, economic environment, and plan specifics.

Lump-sum acceptance rates (affect both immunization and termination)

75% of active and terminated-vested participants elect lump sum.

Most participants choose a lumpsum option when it is offered.

Plan sponsors should use historical or peer data to project potential lump-sum acceptance rates for their plan.

Enhanced early retirement benefits (affect immunization and termination)

No enhanced early retirement benefits.

Enhanced early retirement benefits increase the cost of annuities, but they can be eliminated from lumpsum payments.

If early retirement benefits are not included in the lump-sum value, then lump-sum acceptance rates may be lower.

Offering retiree lump sums

No lump sums are offered to retirees at termination.

Retirees already in-payment can be offered a lump-sum option at termination.

Offering lump sums to retirees may increase the cost of the group annuity purchase.

Source: Vanguard.

5

Figure 4.

Ratio of termination cost to immunization cost and impact of key variables

Baseline comparison

Plan type Typical Mature Young 99.1%

100.1%

98.2%

Immunization variable changes to baseline ratio Annual bond underperformance relative to liability from 80 bps to 100 bps

(1.6%)

(1.6%)

(1.7%)

Ongoing plan expenses from 10 bps to 30 bps

(1.6)

(1.6)

(1.7)

Lump sums not offered during immunization

(2.9)

(1.8)

(4.0)

Termination variable changes to baseline ratio Group annuity discount rate decreased by 35 bps

1.8%

2.2%

1.5%

Lump sum acceptances decrease from 75% to 25%

4.7

2.8

6.6

Plan contains enhanced early-retirement benefits

1.6

1.0

3.9

Retirees offered lump sum as part of termination

1.5

2.6

0.4

Source: Vanguard.

The rest of this paper examines the sensitivity of these cost comparisons to changes in the value of the variables. Figure 4 summarizes the results from our tests of each variable. Negative changes indicate a decrease in the cost of termination relative to immunization. Positive changes indicate an increase in the cost of termination relative to immunization. Figures 7–9 focus on the variables affecting the cost of immunization. Figures 10–13 focus on the variables primarily affecting the cost of termination. We list additional factors that largely resist quantitative analysis at the end of the paper.

Variables affecting the cost of immunization Bond underperformance relative to liability Bond underperformance relative to changes in the value of the pension liability is a critical factor in determining the cost of immunization for a frozen pension plan. It is important that plan sponsors realize that the true cost of their plan is higher than the ongoing plan liability measured under PPA and accounting rules, largely because of a mismatch

between the discount rates assumed in the regulations and the real-world investment opportunities in the bond market. The assets available for an immunization strategy are corporate bonds and Treasury securities.4 Corporate bonds can experience credit downgrades and defaults, which raise their yields and decrease their returns. The discount rate used to calculate the plan’s liability is subject to no similar changes in yield (see Figure 5, for an example). Treasury securities have lower yields than the corporate bond yields used to measure the ongoing liability. In addition, there is a cost for managing the immunized asset portfolio. For these reasons, a portfolio of assets dedicated to immunizing the pension liability will underperform the “return” on—or changes in the value of—the liability. This underperformance reflects the fact that plan liabilities are calculated using the corporate bond discount rate with no downgrades or costs. We define this effect as “bond underperformance relative to the liability.” (For additional information on bond performance and allocation compared to liability return, see Bosse and Inglis, forthcoming, 2012).

4 We don’t include other types of bonds, such as mortgage-backed securities (MBS), because their behavior in changing interest rate environments and other characteristics do not make them a suitable asset for matching a pension liability.

6

Figure 5.

Impact of bond-rating downgrade on pension funding Bonds A and B make up liability universe

Market sees more credit risk in Bond B

Bond B is downgraded out of the liability universe

Bond B

6.0%

Bond B

Yield Bond A

5.0%

Bond B

Discount rate = 5.0%

Bond A

Yield increases

Bond A

Discount rate = 5.5%

Discount rate = 5.0%

Assets: Lose value Liability: Decreases

Assets: No change in value Liability: Increases

Source: Vanguard.

We estimated the average annual loss of an immunization portfolio relative to a pension liability to be roughly 65 basis points (bps) per year for a typical plan. This estimate was based on the perfor­ mance of a hypothetical immunization portfolio for the period March 1997–March 2012 (see Figure 6, on page 8).5 The impact of defaults and downgrades alone for a portfolio of 100% investment-grade corporate bonds has been estimated at roughly 70 bps per year (Ransenberg and Hobbs, 2011). We added investment fees (15 bps) to our 65 bps estimate, resulting in assets underperforming liabilities by a total of 80 bps annually. The level of underperformance will vary over time, and the cost of an immunized approach will differ depending on the solution and asset manager. We examined an alternative scenario in which bond underperformance was 100 bps per year instead of 80 bps. Figure 7, on page 8, shows the change in the immunization liability with this alternative assumption. Termination looks even more favorable for the young plan and is also now the lower-cost option for the typical and mature plans, based on this new assumption for bond underperformance.

Ongoing plan expenses In addition to asset performance and investment management expenses, an immunization approach must also bear the costs of maintaining the pension plan in the future. These expenses include items such as actuarial fees, administration costs, and Pension Benefit Guaranty Corporation (PBGC) premiums. The pension benefit liability does not typically recognize these items, but the assets supporting the plan’s liabilities must be used to pay these additional expenses. In the case of a group annuity, an insurance company captures these costs in the discount rate it uses to price these contracts. For purposes of our comparison, we have translated anticipated expenses for an immunized plan into a discount rate factor.

The baseline assumption, a 0.10% decrease in discount rate, is based on a plan with $500 million in assets and 5,000 participants and equates to annual operating costs of approximately $250,000. Administrative costs may be higher for various reasons, but as a percentage of the liability, the size of the plan will be a major influence on the size of these costs. In general, costs as a percentage of the plan liability will be higher for a smaller plan (because

5 In Figure 6, the immunized portfolio loses approximately 9.15% of funding over 15 years, for an average yearly drop of 61 basis points (1 basis point = 1/100 of 1%).

7

Figure 6.

Bond downgrades and defaults have contributed to a decline in funded status for a hypothetical immunized portfolio: March 1997–March 2012

Quarterly pension funded status based on immunization portfolio 120%

110

100

90

80 Mar. Dec. Sept. June Mar. Dec. Sept. June Mar. Dec. Sept. June Mar. Dec. Sept. June Mar. Dec. Sept. June 1997 1997 1998 1999 2000 2000 2001 2002 2003 2003 2004 2005 2006 2006 2007 2008 2009 2009 2010 2011

Mar. 2012

LDI Strategy

Note: The liability is represented by the Citigroup Pension Liability Index, and the asset portfolio is represented by 10% Barclays U.S. Aggregate Bond Index, 55% Barclays U.S. Long Credit Bond Index, and 35% Barclays U.S. 20–30 Year Treasury STRIPS Bond Index. Source: Vanguard.

Figure 7.

As bond underperformance increases, termination becomes more attractive

(Annual bond underperformance increases from 0.8% to 1.0%) (Liability in $ millions) Baseline immunization liability

Change in immunization liability

Typical Mature Young $537.6

$536.9

$540.0

$8.8

$8.6

$9.4

Adjusted immunization liability

$546.4

$545.5

$549.4

Baseline termination liability

$532.8

$537.4

$530.3

Termination/immunization ratio Lower-cost alternative Change in termination/immunization ratio Source: Vanguard.

8

Plan type

97.5% 98.5% 96.5% Termination

Termination

Termination

(1.6%)

(1.6%)

(1.7%)

Figure 8.

As a plan’s ongoing expenses increase, termination becomes more attractive

(Ongoing plan expenses increase from 0.1% to 0.3%) (Liability in $ millions) Baseline immunization liability

Change in immunization liability

Plan type Typical Mature Young $537.6

$536.9

$540.0

$8.8

$8.6

$9.4

Adjusted immunization liability

$546.4

$545.5

$549.4

Baseline termination liability

$532.8

$537.4

$530.3

Termination/immunization ratio Lower-cost alternative Change in termination/immunization ratio

97.5% 98.5% 96.5% Termination

Termination

Termination

(1.6%)

(1.6%)

(1.7%)

Source: Vanguard.

relatively fixed costs such as actuarial fees and administrative costs are borne by a smaller asset base) and smaller for a bigger plan. Therefore, immunization may be more cost-effective for larger plans than for smaller plans. Note that a group annuity contract will also recognize administrative economies of scale, but the impact will be more significant for ongoing plans. Figure 8 examines the impact of roughly tripling the plan expenses (as a percentage of plan liability) from our baseline case, as might be the case for a smaller plan. The cost of immunization rises and termination becomes more cost-effective. The low-cost option changes for both the typical and mature plan, while termination becomes even more favorable for the young plan.

Offering lump sums Lump sums provide the most cost-effective way of settling corporate pension liabilities. Typically during a plan termination, lump sums are offered to active and terminated-vested participants to reduce the cost of termination (Inglis and Sparling, 2011). Our baseline comparison assumed that the ongoing immunization option would also offer lump-sum payments to participants (paid at termination or retirement), which gave the immunization option some of the same cost benefit of offering lump sums on termination. The cost benefit for the ongoing plan is not the same, because participants do not immediately elect lump sums as they would in a plan termination.

9

Figure 9.

When lump-sum payouts are not offered in immunization, termination becomes a more attractive option

(Liability in $ millions) Baseline immunization liability

Plan type Typical Mature Young $537.6

$536.9

$540.0

$16.1

$9.7

$23.0

Adjusted immunization liability

$553.7

$546.6

$563.0

Baseline termination liability

$532.8

537.4

$530.3



Change in immunization liability

Termination/immunization ratio Lower-cost alternative Change in termination/immunization ratio

96.2%

98.3%

94.2%

Termination

Termination

Termination

(2.9%)

(1.8%)

(4.0%)

Source: Vanguard.

However, plan sponsors that do not terminate their plan would not necessarily decide to offer lump sums. The alternative results shown in Figure 9 illustrate how the cost of immunization changes when lump sums are no longer offered at the point of retirement or termination of employment. (We still assumed that lump sums are offered in the plan termination alternative.) The larger the number of active participants, the bigger the impact of offering lump sums during immunization. The cost of the “young” plan increases by more than 4% if a lumpsum option is not offered. Variations on lump-sum offerings exist. For example, a plan sponsor can set limits on the amount of a lump-sum payout or require that certain eligibility criteria be met. The cost savings associated with these in-between approaches will lie somewhere between the baseline case and the results shown in Figure 9 on a percentage basis.

10 

Variables affecting the cost of plan termination Cost of a group annuity purchase For reasons described previously, purchasing annuities in a group annuity contract is relatively expensive. The primary pricing lever used for group annuity contracts is the discount rate. We have used survey information from group annuity brokers and compared their price quotes with data from the Citigroup Pension Liability Index (a widely used pension discount rate benchmark) and the IRS non-averaged three-segment rates (used for lumpsum calculations) to estimate the typical difference between the discount rate used by insurance companies to price a group annuity contract and the discount rate used to measure plan liabilities in ongoing pension plans.

Figure 10.

As group annuity costs increase, the case for immunization improves

(Group annuity discount rate decreases by 35 basis points) (Liability in $ millions)

Plan type



Typical Mature Young

Baseline termination liability

$532.8

$537.4

$530.3

$9.7

$11.8

$8.3

Adjusted termination liability

$542.5

$549.2

$538.6

Baseline immunization liability

$537.6

$536.9

$540.0



Change in termination liability

Termination/immunization ratio

100.9%

102.3%

99.7%

Lower-cost alternative

Neutral

Immunization

Neutral

1.8%

2.2%

1.5%

Change in termination/immunization ratio Source: Vanguard.

We estimated the typical difference between the discount rates for ongoing pension plans and those used to price group annuities to be approximately 1.10% for a population of retirees and 1.50% for a group of active or terminated-vested participants (for details on group annuity costs, see appendix Figure A-3).These were the levels used in the baseline calculations. The larger reduction in the discount rate for active participants represents the additional risk associated with active participants, such as uncertain benefit amounts, the timing of retirements, and the reinvestment of assets in different interest rate environments. Note that the decrease in the discount rate affects the active, terminated-vested and retiree liabilities differently because of the different durations of the liabilities associated with these populations. Figure 10 reduces the discount rate by an additional 0.35% (that is, makes annuities more expensive) for both the retiree and active/terminated-vested group annuity purchase. We modeled this as a parallel shift of 0.35% in the entire yield curve of rates used by

insurance providers to price their contracts. This equated roughly to a 1.50%–2.50% increase in the termination liability. (Note that only the non-lumpsum portion of the liability was affected.) The lower the group annuity discount rate, the more expensive the group annuity is, such that immunization compares more favorably. The size of a plan will have an impact on group annuity pricing, sometimes in countervailing ways. For example, a large plan may benefit from lower costs as a result of economies of scale for administration. Insurance providers may also be more aggressive about winning group annuity business depending on their circumstances (mix of desired business, available capital, view on interest rates, and so on). On the other hand, economies of scale can sometimes be hard to realize because insurers may not have the risk capacity to take on larger plans. Plan sponsors can work with brokers or consultants to follow the market and make good decisions about when to solicit group annuity bids.

11

Figure 11.

As fewer participants take lump sums, immunization becomes more attractive

(Lump-sum acceptance rate decreases from 75% to 25%) (Liability in $ millions)

Plan type



Typical Mature Young

Baseline termination liability

$532.8

$537.4

$36.2

$21.7



Change in termination liability

$530.3 $51.7

Adjusted termination liability

$569.0

$559.1

$582.0

Baseline immunization liability

$537.6

$536.9

$540.0



Change in immunization liability

Adjusted immunization liability Termination/immunization ratio Lower-cost alternative Change in termination/immunization ratio

$10.7

$6.4

$15.3

$548.3

$543.3

$555.4

103.8%

102.9%

104.8%

Immunization

Immunization

Immunization

4.7%

2.8%

6.6%

Source: Vanguard.

Lump-sum acceptances Another important factor in estimating the cost of terminating a plan is how many participants will elect a lump-sum payment, instead of an annuity. As described earlier, lump sums are a cost-effective way to settle a pension promise. The more participants electing a lump sum, the lower the cost of terminating a plan. While most active and terminated-vested participants are expected to take a lump sum, concerns about longevity or investing the money themselves will lead some participants to select the annuity.

In the baseline, 75% of active and terminatedvested participants are assumed to select a lump sum. Figure 11 compares that to an acceptance rate of 25%. We looked at this very low potential rate of lump-sum acceptance in recognition of situations where lump sums might be discouraged—some union environments, for example. This assumption was meant to represent the wide variability of

12 

acceptance rates from sponsor to sponsor, not the expectations for any particular sponsor. In general, however, uncertainty about the rate of lump-sum acceptance abounds. Our anecdotal experience suggests that acceptance rates for any one plan sponsor may vary by 20% or more from what is expected. Because the baseline assumes lump sums are paid during immunization and termination, we changed the acceptance rate for lump sums in both situations to examine this factor. The lump-sum acceptance rate has a bigger impact on the termination scenario because in this situation the lump sum is elected and paid immediately. Figure 11’s results show that a low acceptance rate significantly increases the cost of termination to the point where immunization looks more cost-effective for all three plans.

Figure 12.

When plans contain enhanced early-retirement benefits, immunization becomes more attractive

(Liability in $ millions)

Plan type



Typical Mature Young

Baseline termination liability

$532.8



Change in termination liability

$537.4

$530.3

($4.8)

($2.9)

($11.5)

Adjusted termination liability

$528.0

$534.5

$518.8

Baseline immunization liability

$537.6

$536.9

$540.0



($13.3)

($8.0)

Adjusted immunization liability

$524.3

$528.9

$508.1

Termination/immunization ratio

100.7%

101.1%

102.1%

Lower-cost alternative

Neutral

Immunization

Immunization

1.6%

1.0%

3.9%

Change in immunization liability

Change in termination/immunization ratio

($31.9)

Note: Terminated-vested participants are assumed to have forfeited their right to early-retirement benefits by quitting, and retirees are already in payment; thus no adjustment is required. Source: Vanguard.

Enhanced early retirement Many pension plans provide incentives to retire at certain ages through enhanced early-retirement benefits. These may be in the form of reduced benefits that start before normal retirement or supplements intended to help the retiree bridge the period until Social Security benefits can be paid.

An enhanced early-retirement benefit will change the termination versus immunization comparison. First, plan sponsors are not obligated to include the value of enhanced early-retirement benefits in lumpsum payments, so the cost of a plan termination may be reduced by excluding these benefits from lump-sum payments. On the other hand, a group annuity provider will charge for the extra risk associated with the uncertainty of participants’ retirement date and benefit amount.

In our baseline case, we assumed the plans had no enhanced early-retirement benefits. Figure 12 examined the impact if: • Ten percent of the active liability was due to enhanced early-retirement benefits. • Participants electing the lump sum forgo the 10% of their liability associated with their enhanced early-retirement benefit. • The group annuity for active participants has a 15% load added for the additional risk to the insurance company. Note that because lump sums are also assumed to be paid for the ongoing immunized plan, the presence of early-retirement benefits affects the cost of immunization as well. The immunized liability is not loaded for the risk associated with early-

13

Figure 13.

When retirees are offered a lump sum at termination, the financial impact is sensitive to the number of participants accepting the lump sum

(Liability in $ millions)

Plan type



Typical Mature Young

Baseline termination liability

$532.8



Change in termination liability

$7.7

$537.4

$530.3

$13.9

$2.3

Adjusted termination liability

$540.5

$551.3

$532.6

Baseline immunization liability

$537.6

$536.9

$540.0

Termination/immunization ratio

100.6%

102.7%

98.6%

Lower-cost alternative

Neutral

Immunization

Termination

1.5%

2.6%

0.4%

Change in termination/immunization ratio Source: Vanguard.

retirement benefits, because we have assumed the actuarial assumptions are realized, but plan sponsors should be aware of the risk that more participants than expected will take the early-retirement benefits and thus increase the cost of immunization. As shown in Figure 12, immunization looks more favorable for both the young and mature plans if the plan includes enhanced early-retirement benefits. Note that if a plan sponsor chooses to include the value of enhanced early-retirement benefits in lumpsum payments, the group annuity would be more expensive without an offsetting decrease in the value of lump sums, and the cost of immunization would look even more favorable. Retiree lump sums Typically, once a retiree elects a form of payment, the payment form cannot be changed to a lump sum.6 However, during a plan termination, a plan sponsor has the opportunity to allow retirees who have already elected an annuity to change their minds and elect lump sums.

The appeal of offering a lump sum to retirees already receiving an annuity is to avoid the expense of purchasing an annuity for as many participants as possible. However, a group annuity provider is also likely to increase the annuity premium for adverse selection (more healthy people choose the annuity) if a lump-sum offer is made or has been made in the recent past. These conflicting forces—reduced costs for paying lump sums and increased costs for adverse selection—leave the ultimate level of cost reduction uncertain. Based on the assumptions we used for retiree lump-sum acceptances and adverse selection costs, total termination costs actually increase, but this will vary based on actual plan experience. Figure 13 takes into account the impact of offering lump sums to retirees during a plan termination. We assumed that 50% of retirees would elect to take the lump sum. If the election percentage differs from 50%, then the cost of termination may look more or less favorable.

6 In an exception to this generalization, Ford Motor Company announced in April 2012 that it would offer lump sums to retirees already in payment, even though the company was not terminating its pension plan.

14 

Other factors

Conclusion

Figures 7–13 quantified the costs of immunization and termination in order to compare them on a numerical basis. Plan sponsors will also want to consider factors that are harder to quantify. In most situations, the costs of termination and immunization are very close, so the factors listed here will be a significant part of the decision.

Generally the cost of immunizing a portfolio and the cost of terminating a pension plan will be about the same, but the comparison does vary based on a variety of factors. Factors for which costs are difficult to quantify may also be significant and will affect the decision. Plan sponsors should be aware of the cost impact for each of these factors on their frozen pension plan when deciding whether to immunize or terminate.

Factors that may sway a plan sponsor toward immunization: • The plan sponsor desires that employees receive an annuity benefit and a lifetime source of income from the company. • The frozen pension benefit is still used as part of the larger retirement benefit for the sponsoring organization. • The plan sponsor feels a fiduciary responsibility to execute the plan. Factors that may sway a plan sponsor toward termination: • The administrative burden of the pension plan. • Management decisions related to the plan distract from the core business. • The frozen pension plan is confusing and misunderstood by employees. • Employees want access to the value of their frozen pension benefit.

References Bosse, Paul M. and R. Evan Inglis, forthcoming, 2012. For Better Pension Liability Matching, Add Treasuries to the Mix. Valley Forge, Pa.: The Vanguard Group. Brentwood Asset Advisors. December 2010– October 2011. Annuity rates; available at brentwooodllc.com. Dietrich and Associates. December 2009– October 2011. Monthly interest rate updates; available at www.dietrichassociates.com. Inglis, Evan R., and Jeff Sparling, 2011. Pension Plan Terminations: Minimizing Cost and Risk. Valley Forge, Pa.: The Vanguard Group. Ransenberg, Daniel, and Jonathan Hobbs, 2011. Overcoming Credit Downgrades: Four Ways to Improve Your Liability Hedge. New York: BlackRock.

• An immunization approach does not eliminate all risk, including demographic risks related to participants’ life expectancy and other factors that cannot be hedged effectively.

15

Appendix. Ongoing liability assumptions Figure A-1.

Hypothetical liability durations

Figure A-2.

Hypothetical liability discount rates prior to variable adjustment

Liability duration (years) Active liability

15

Ongoing liability discount rates

Terminated-vested liability

13

Active rate

Retiree liability

Figure A-3.

8

5.00%

Terminated-vested rate

4.50%

Retiree rate

4.00%

Comparing discount rates of a group annuity provider and a typical pension

Liability discount rates

Retiree Active/terminatedannuity discount vested annuity rate costs discount rate costs

Active/ Retiree Retiree Active Active Annuity purchase rates Retiree terminated- insurance insurance insurance insurance discount vested Dietrich Dietrich Brentwood Brentwood costs— costs— costs— costs— Date rate* rate** retiree active immediate deferred Dietrich Brentwood Dietrich Brentwood

December 31, 2009

5.32%

6.13%

4.50%

5.13%

0.82%

1.01%

January 31, 2010

5.31

6.14

4.00

4.38

1.31



1.76

February 28, 2010

5.42

6.29

4.25

4.75

1.17

1.54

March 31, 2010

5.31

6.21

4.25

4.63

1.06

1.58

April 30, 2010

5.27

6.12

4.00

4.63

1.27

1.50

May 31, 2010

5.11

5.92

3.50

4.13

1.61

1.80

June 30, 2010

5.05

5.91

3.50

4.13





1.55



1.79

July 31, 2010

4.80

5.71

3.50

4.00





1.30



1.71

August 31, 2010

4.51

5.45

3.13

3.88

1.39

1.58

September 30, 2010

4.45

5.52

3.13

3.88

1.32

1.64

October 31, 2010

4.41

5.62

3.13

3.88

1.29

1.75

November 30, 2010

4.58

5.88

3.38

4.38

1.21

1.50

December 31, 2010

4.91

6.00

3.63

4.38

4.30%

4.60%

1.28

0.61%

1.63





1.40%

January 31, 2011

4.92

5.95

3.63

4.38

4.25

4.60

1.29

0.67

1.58

1.35

February 28, 2011

5.01

6.06

3.63

4.38

4.25

4.60

1.38

0.76

1.69

1.46

March 31, 2011

4.90

5.97

3.88

4.38

4.25

4.60

1.02

0.65

1.59

1.37

April 30, 2011

4.90

5.96

3.63

4.38

4.10

4.60

1.28

0.80

1.58

1.36

May 31, 2011

4.67

5.73

3.63

4.38

4.00

4.60

1.05

0.67

1.35

1.13

June 30, 2011

4.66

5.79

3.63

4.38

4.10

4.85

1.04

0.56

1.42

0.94

July 31, 2011

4.65

5.76

3.13

3.88

3.80

4.35

1.52

0.85

1.88

1.41

August 31, 2011

4.34

5.46

3.13

3.88

3.35

4.10

1.22

0.99

1.59

1.36

September 30, 2011

4.24

5.28

3.38

3.88

3.50

3.85

0.86

0.74

1.40

1.43

October 31, 2011

4.31

5.12

3.13

3.63

3.60

4.10

1.19

0.71

1.50

1.02

November 30, 2011

4.22

4.94

3.13

3.88

N/A

N/A

1.10

N/A

1.07

N/A

December 31, 2011

4.21

4.92

3.13

3.63

3.35

3.85

1.09

0.86

1.30

1.07



Average— Average— retiree 1.07% active

Notes: The IRS publishes, on a monthly basis, three segment rates for plan sponsors to determine minimum lump-sum benefits. The first segment rate discounts payments due in the first five years; the second segment rate discounts payments due for the next 15 years; and the third segment rate discounts payments due for 20 years and beyond.   *Retiree discount rate is 10% of the first segment rate and 90% of the second segment rate as published by the IRS for each month. **Active/terminated-vested discount rate is 40% of the second segment rate and 60% of the third segment rate as published by the IRS for each month. Sources: Vanguard, using data from the U.S. Internal Revenue Service, Dietrich and Associates, and Brentwood Asset Advisors, as indicated in this table.

16 

1.46%

P.O. Box 2600 Valley Forge, PA 19482-2600

Connect with Vanguard® > vanguard.com

Vanguard research > Vanguard Center for Retirement Research Vanguard Investment Counseling & Research Vanguard Investment Strategy Group E-mail > [email protected] CFA ® is a trademark owned by CFA Institute.

© 2012 The Vanguard Group, Inc. All rights reserved.. ICRFPP 112012