PRACTICE NOTE

Hibernation versus termination Evaluating the choice for a frozen pension plan

James Gannon, EA, FSA, CFA, Director, Asset Allocation and Risk Management

ISSUE:

As a frozen corporate defined benefit pension plan matures, the need for a decision on exit strategy gets closer. One option is plan termination (the purchase of annuities for all participants, which effectively transfers all liabilities to an insurance company). Another strategy is “hibernation,” whereby the sponsor continues to manage the plan at a low level of cost and with some uncertainty about future cost. What are the primary considerations for plan sponsors evaluating the latter option? RESPONSE:

If a long view is taken, the choice to terminate a frozen pension plan is a “when, not if” decision. Meaning, at some point the sponsor of every frozen pension plan will purchase annuities for the remaining participants and then terminate the plan. It is unlikely that a plan sponsor would elect to manage the plan until the last benefit is paid to the last surviving participant. Knowing this, plan sponsors should first evaluate the relative cost of each decision – termination versus hibernation. In this paper we present a framework for comparing the upfront fixed costs of plan termination to the long-term costs, and uncertain outcomes, of plan hibernation. This paper shows that plan hibernation may be a sound strategy for many pension plan sponsors; however, Russell believes sponsors should enter into this phase only after fully evaluating and understanding the costs of hibernation relative to costs of an immediate plan termination.

Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan Frank Russell Company owns the Russell trademarks used in this material. See “Important information” for details.

MARCH 2014

Background When sponsors freeze their pension plans they have set out on a course toward eventual plan termination. If a sponsor chooses not to terminate a frozen plan immediately, the plan enters into a state called hibernation.1 The goal of this hibernation phase is to operate the plan at an ongoing cost that is lower than the cost of immediate termination, but sponsors need to be aware that they are accepting a level of uncertainty about future cost – i.e., that the cost of plan hibernation may ultimately exceed that of immediate plan termination. This paper will compare the costs of immediate termination and hibernation exit strategies, on a net present value (NPV) basis, to help clients determine their preferred strategy. Under our evaluation methodology, the following decision rules will guide us.  Termination is preferable when the NPV of all costs under the hibernation strategy is greater than the sponsor’s outlay to fund the immediate termination premium.  Hibernation is preferable when the NPV of all costs under the hibernation strategy is less than the plan sponsor’s outlay to fund the immediate termination premium. To illustrate the decision-making process: We set the definitions for the variables used in the evaluation; we show two simplified examples, and then transition to a more detailed stochastic analysis; and finally, we suggest possible sensitivity testing on the various variables used in the analysis. The stochastic analysis and the sensitivity testing will be where we examine more closely the uncertainty about the costs of the hibernation strategy. Further, in the following text, when we say “the plan sponsor should…,” what we are actually saying is “based on an NPV analysis, the plan sponsor should…”. We realize that some of these decisions have other inputs (such as employee relations, need for flexibility for the future, ability to fund the initial termination cost, internal resource constraints, and fiduciary or regulatory concerns) and therefore a plan sponsor may, quite reasonably, not always do what a pure financial analysis may point toward.

Definitions Termination – An immediate purchase of annuities for all plan participants.2 After this action the plan will cease to exist, and an insurance company3 will inherit the obligation for all future payments. A benefit of the termination strategy is that the cost is known in advance and comes with no variability.

Costs of plan termination  Termination premium – There is typically a “premium” over the corporation’s stated liability value (projected benefit obligation or PBO, for instance) for transferring the responsibility of ongoing payments to an insurance company.4 For example, to purchase annuities for $100 of stated plan liability, $110 of assets might need to be transferred to the insurance company.5 In such case, the termination premium is $10. This additional $10 covers a more conservative (and perhaps more accurate) valuation of the plan liabilities6 and expenses and the profits of the insurance company, to name a few of these items. The sponsor should include costs paid to various vendors who facilitate the plan termination process in the NPV analysis.  Surplus or deficit – Surplus is any assets in the plan in excess of the stated plan liability, and deficit is any amount less than the stated liability. Surplus can be used to offset the termination premium and therefore lower the sponsor’s cost of plan termination. Any deficit would add to the cost of the termination premium and increase the cost of terminating. For example, if the above plan has $105 in assets compared to its $100 in stated liability, then the cost of plan termination would not be the full termination premium, but rather the additional $5 necessary to raise the assets to $110 (the amount needed by the insurance company).

Hibernation – Managing the plan for an extended period of time and then terminating it at a later date. When deciding on the hibernation strategy, plan sponsors must be confident that they can minimize the related costs so that they will be less than the costs of immediate plan termination – even as they bear in mind that the initially lower cost of hibernation may be accompanied by significant uncertainty. Thus we will define “cost” as the sum of all cash expenditures made with respect to the pension plan. Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

2

Costs of plan hibernation – For this paper we will assume that the following components are either costs or offsets to costs of the hibernation process and therefore should be included in any analysis:  Annual expenses – Investment management, actuarial, legal and staff fees; flat rate and variable PBGC (Pension Benefit Guaranty Corporation) premiums; other administrative expenses. Annual expenses will be deemed costs during the hibernation process. Therefore it is important to reflect all expenses of the pension plan when comparing the options.  Ongoing plan funding – The funding requirements of the plan during the hibernation period.  Termination premium – This is the same concept as above; however, the termination premium will be paid based on the liability structure at the end of the hibernation period. The sponsor should include costs paid to various vendors who facilitate the plan termination process in the NPV analysis.  Surplus or deficit – Any surplus of the pension plan generated during the hibernation process could be used as an offset to the above costs. Therefore, if plan assets earn amounts in excess of liability growth, those amounts (a surplus) would be used to offset plan expenses during hibernation or be applied to the termination premium at the end of hibernation. This recognizes that at some level, there may be a reward for taking investment risk during the hibernation process, but that the potential reward (as we will see later) is limited.7 Any deficit incurred during the hibernation period would either increase the cost of termination at the end of the hibernation period or extend the hibernation period, and thereby increase costs due to ongoing plan expenses. One implication of using an NPV approach is that if at any point in time (either at the outset or in the course of the hibernation period) the plan has a surplus that is at least equal to the termination premium, the analysis will show that the sponsor should terminate the plan, since the cost to the company of plan termination is zero (no additional contribution is needed to terminate the plan); ongoing plan hibernation would be more expensive, due to future plan expenses.

The ideal situation: the liability-matching asset SAMPLE PLAN FOR FURTHER ANALYSIS

Pension liability = $1 billion, frozen to new benefit accruals Pension assets = $1 billion Annual expenses = 50 basis points relative to plan assets, paid at end of year Liability discount rate = IRS full yield curve as of January 1, 2014 NPV discount rate = 5% The assumption is that this company commits to a hibernation period of 10 years.

We will weigh two choices: 1) plan termination today and 2) plan hibernation for 10 years. Let’s assume that this pension plan has the ability to invest in the perfect liability-matching asset, an asset whose movements exactly mimic the movements of the plan liability. Through the hibernation process, the plan will incur the following costs: annual plan expenses and a termination premium at end of the hibernation period. No additional contributions (beyond plan expenses) will be paid during the hibernation period, because the liability is perfectly matched; and none of the termination premium will be offset by any surplus, because no surplus is created when investment is in the liability-matching asset. The cost of hibernation, in each case study, will be compared to the immediate termination premium.

Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

3

Case Study 1: The stable pension plan Let’s start with a simple, but very unrealistic, plan and calculate the cost of its hibernation. In this case we will use a stable pension plan frozen to new benefit accruals, where the main attribute is that liability is flat from year to year. This is accomplished by setting the benefit payments equal to the interest cost. We would call this unrealistic, because a pension plan will eventually have to pay out benefits that will run the liability to zero. So although this is unrealistic, it lays a good foundation for the case studies that follow. In this case we have assumed that the immediate termination premium, based on the demographics of the stable population, is $100 million.

Exhibit 1 YEAR Liability (and asset value)

0

1

2

3

4

5

6

7

8

9

10 1,000

1,000

1,000

1,000

1,000

1,000

1,000

1,000

1,000

1,000

1,000

Benefit payments

50

50

50

50

50

50

50

50

50

50

Interest cost

50

50

50

50

50

50

50

50

50

50

5

5

5

5

5

5

5

5

5

Plan expenses (also contributions) Termination premium NPV cost of plan hibernation

5 100

100

In million $.

The cost of plan hibernation is the present value of costs paid by the company during the hibernation period. This will equal 10 annual payments of plan expenses plus a payment at Year 10 equal to the termination premium, which will still be $100, due to our stable population. Remember: there are no contributions related to deficit during the hibernation period and no offset to the termination premium from surplus, because this plan is invested in an asset that perfectly matches the movements of plan liabilities. The only contributions are related to plan expenses (they are paid from the plan, causing a deficit, and then contributed by the company to restore full funding). The NPV of plan hibernation for the stable pension plan is $100 million, which is the sum of annual expenses ($5 million each year) plus the termination premium at Year 10 ($100 million). In this case the hibernation cost equals the termination cost (by design), and the company would be indifferent as to the two choices. There is not much to comment on in this situation, as it was created to act wholly as a marker for the more realistic type of plan we are looking at next.

Case Study 2: The mature pension plan The next comparison is for a mature pension plan. A mature pension plan is one that ceased new benefit accruals many years ago and is now likely to have a small active employee population that is not accruing benefits and a large retiree population that is receiving benefit payments. As seen from Exhibit 2, this frozen plan will start with high amounts of benefits payments (shown in light blue) each year; there will be a decline through time as retirees die and the plan population decreases. The values of both the termination liability (shown in dark blue) and the stated liability (show in gray) will decrease year over year as benefits are paid out at a very rapid rate. Also, note that the termination premium – the difference between the value of the termination liability and the value of the stated liability – decreases each year. This happens for two reasons: the size of the liability decreases, and the uncertainty of the liability valuation decreases as the plan population ages.

Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

4

$80

$1,200

$60

$900

$40

$600

$20

$300

$0

Liability ($M)

Benefit payments ($M)

Exhibit 2

$0 0

5

10

15

20 Years

25

Annual benefit payments

30

35

Stated liability

Termination liability

Based on the demographics of the mature pension plan (which differ from those of the stable pension plan), we have calculated the difference between the termination liability and the stated liability, or the immediate termination premium, to be $110 million.

Exhibit 3 YEAR

0

1

2

3

4

5

6

7

8

9

10

1,000

978

955

931

904

876

847

815

782

747

710

Benefit payments

70

69

70

70

70

70

70

70

69

69

Interest cost

48

46

46

43

42

41

38

37

34

32

5

5

5

5

5

4

4

4

4

Liability (and asset value)

Plan expenses (also contributions)

4

Termination premium NPV cost of plan hibernation

54 67

In million $.

Exhibit 3 shows the NPV analysis for the mature pension plan. The cost of plan hibernation is $67 million, which is the sum of annual expenses plus the termination premium at Year 10. This is substantially less than the $110 million cost of the immediate termination strategy, and therefore the company would decide for plan hibernation over immediate termination. The key driver in this case study is that the plan is paying off its own liability at a very rapid pace (see the liability’s decline from $1 billion to $710 million over the 10-year period). Therefore, at the end of the hibernation period, a much smaller termination premium is paid, due to the plan’s decreased liability size and its aging population. In this example, the termination premium is $110 million (11% of $1 billion) at the beginning of the period as compared to $54 million (7.5% of $710 million) at the end of 10 years. Over this period, the annual dollar amount of plan expenses also decreases (along with decreasing asset size), which helps hibernation. The case studies above, though simplified, do show a basic structure for a sponsor’s evaluation of the costs of plan hibernation relative to costs of immediate termination and should help guide analysis and decision-making processes. In the following section we will begin to add complexities to the case studies above, through either a full stochastic analysis or a sensitivity analysis wherein we will discuss the individual variables we held constant during the simplified and stochastic analyses and look at how the evaluation might change with a change in those variables.

Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

5

The stochastic example – an introduction of investment risk and interest rate volatility This section is dedicated to an example that demonstrates the uncertainty of hibernation cost by use of a stochastic modeling process. During our stochastic process, we will work with 1,000 scenarios (or paths) that last for 10 years, which will be the hibernation period. The 1,000 scenarios represent possible future economic outcomes for the plan sponsor, and will work by projecting assets and valuing plan liabilities using our capital markets assumptions, and comparing those values to determine any required plan contributions and the eventual termination premium at the end of the 10-year hibernation period.8 Along each of these 1,000 paths we will value the cumulative cost of plan hibernation, on a present value basis, and compare that to the cost of an immediate plan termination. The pension plan we will use for this analysis is the mature pension plan used in Case Study 2. This plan is fully funded at the time of evaluation and pays annual benefits that will cause liability to decrease in size, as represented in Exhibit 2. (Further information for this analysis, including data, methods and assumptions, can be seen in the Appendix.) To analyze our preference for various investment strategies, our key metrics will be the following: 1.

Winning and losing – In what percentage of our 1,000 scenarios does hibernation have the lowest cost, and in what percentage does immediate termination have the lowest cost.

2.

Variability of hibernation cost – What investment strategy can minimize the worstcase scenario of plan hibernation (defined as the 95th-percentile cost in the plan hibernation distribution).

The portfolios we will analyze will range from 70% return-seeking and 30% liability-driven fixed income (labeled “70/30”) down to 0% return-seeking and 100% liability-driven fixed income (labeled “0/100”), and the allocations will be shown in 10% increments. These will be assumed to be static investment policies that are frequently rebalanced to their target allocations.

Winning and losing – Based purely on a look at Exhibit 4, it would seem that any of the strategies could be used to justify a plan sponsor’s choice to hibernate rather than terminate the plan. For each portfolio, hibernating is less costly in more of the scenarios than is the choice of immediate plan termination.

Exhibit 4: Probability that hibernation cost would be less than plan termination cost, and therefore that plan hibernation would be the preferred strategy 70/30

60/40

50/50

40/60

30/70

20/80

10/90

0/100

72.2%

72.3%

71.5%

69.3%

68.5%

67.1%

62.9%

54.4%

Portfolios labels read “% return-seeking / % liability-driven fixed income”

Variability of hibernation cost – “Winning” and “losing” are not always the best measures when looking at strategies, especially when the losses can be quite large. Exhibit 5 shows the distribution of the cost of plan hibernation. The graph is read in the following way,  For the “70/30” portfolio, we can see that the median (50th percentile) cost to hibernation would be $8 million, meaning that half of our 1,000 scenarios produce a hibernation cost above $8 million and the other half $8 million below. This amount generally increases with lower return-seeking allocation and higher liability-driven fixed income allocation.  The black line on the chart shows the immediate termination premium of $110 million.  Sections of the bar chart below this line would favor a hibernation strategy; sections above this line, an immediate plan termination.  The other key measurement is the 95th percentile (or maybe the 99th percentile); this is the outcome where only 5% (or 1%, in the case of the 99th percentile) of our stochastic scenarios have produced an outcome of this or a higher amount. This is certainly a scenario that a plan would like to avoid, if the goal is to keep hibernation cost below immediate termination cost. For instance, the 95th percentile of cumulative contributions is $427 million for the “70/30” portfolio and $247 million for the “10/90” portfolio. Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

6

From this chart we can see that we can minimize the 95th percentile with the “10/90” allocation, showing that a hibernation strategy would favor an asset allocation with a high level of liability-driven fixed income, but with some recognition that surplus generation could offset the eventual termination premium at the end of the hibernation period. Sponsors that find the portfolios with high allocation attractive to the hibernation process are simply taking a chance that the markets will generate a return to cover the termination premium in the short term.

Exhibit 5: Hibernation cost at year 10 Hibernation cost

Mature pension plan $600 $110 $450 $300 $150

$0

Portfolio 70/30

Portfolio 60/40

Portfolio 50/50

Portfolio 40/60

Portfolio 30/70

Portfolio 20/80

Portfolio 10/90

Portfolio 0/100

99%

542

493

455

397

367

359

341

345

95%

427

396

364

332

287

254

247

263

75%

137

137

140

138

140

139

142

154

50%

8

9

18

25

41

60

80

103

25%

0

0

0

0

0

12

31

52

The above analysis shows that the key to successful plan hibernation is an investment policy that can maximize the probability of reduced cost during the hibernation period and minimize the uncertainty surrounding future costs. This would cause the investment decision makers to lean toward an asset allocation with a low allocation to high-volatility return-seeking investments and a high allocation to liability-driven fixed income investments. The portfolios on the right side of Exhibit 5 balance the goals of investing to match plan liabilities (and therefore to minimize costs, due to contributions during the hibernation period) and investing to grow surplus (and therefore to have the ability to reduce the termination premium at the end of the hibernation period).

Sensitivity analysis Each of the following items was represented in the stochastic analysis as a known variable; below, we have included a summary of some additional complexities that should be part of the evaluation process of plan termination versus plan hibernation.

Plan expenses – Plan expenses were estimated by using 50 basis points compared to plan assets. Higher plan expenses will make plan termination more attractive; lower expenses will make hibernation more attractive. Knowing this, if a sponsor chooses hibernation, the plan must commit to tightly controlling pension-related expenses if the hibernation is to be successful. It is also the case that not all plan expenses are based on asset size; a substantial portion of pension plan expenses are fixed. Fixed expenses are those that stay the same, regardless of plan size, and for which there are few or no opportunities for economies of scale. Therefore, at some future point, plan size will be small enough that fixed expenses will have grown prohibitive and plan termination will have become more favorable.

Liability valuation – During the hibernation period, the liability valuation process becomes even more important. This calls for using the most appropriate demographic and economic assumptions in the valuing of plan liabilities. These assumptions would make use of the most up-to-date mortality tables (maybe even a mortality table consistent with that used by the insurance company when valuing the plan’s liability) and a market-based interest rate, rather than a rate that may be highly customized from a bond model. Unrealistic actuarial assumptions would lead to liability losses that must eventually be funded by the plan and will show up in the termination premium as well, given that the plan will have been funding to the value of an understated liability. Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

7

Plan demographics – Plan demographics will change the way the plan liability evolves through time. It is possible for a plan with a young demographic profile to see its liability increase for several years, maybe even for as long as a decade. This would mean that the plan would need to have a longer hibernation period if it seeks to terminate when its liabilities have decreased. This longer hibernation period incurs risk, given that there will be more time to incur plan expenses, and more exposure to investment risk (specifically asset-liability mismatch risk). This may make hibernation of a younger-demographic plan more expensive. Number of plan participants is also important, as higher numbers of participants lead to increased PBGC flat rate premiums.

Termination premium – The plan termination premium is based on many factors, including current market conditions, plan demographics, liability size, the assumptions the plan sponsor and the insurance company use to value plan liability, supply and demand for annuity buyouts, even the plan sponsor’s negotiation skills. The methodology for the exact development of the premium is beyond the scope of this paper. It is safe to say that the termination premium will not be constant from the beginning of the period to the end of the period. By choosing to hibernate, the plan sponsor takes on the risk that the termination premium will change. Thus, we advise sponsors to repeat this analysis in response to changes in the cost of annuitization. A plan that chooses to hibernate today does preserve the option to terminate in the future in response to changes in the above pricing factors. We believe there to be a structural effect that would cause the termination premium to decrease during the hibernation period, to the benefit of plan sponsors. First, the liability value will typically be lower at the end of the hibernation period, especially for a mature pension plan, as the plan will be paying benefits that would cause liabilities to decrease year after year (an example of this was shown in Exhibit 2 and 3). Second, the aging of a participant demographic reduces uncertainty about the assumptions the sponsor used to value plan liabilities, and therefore there are fewer differences between the sponsor’s valuation assumptions and the insurance company’s underwriting assumptions. This will lead to a decrease in the difference between the termination liability and the stated liability, the longer a plan stays in hibernation (again, this is illustrated in Exhibit 2).

Funded status – The above analysis was for a fully funded plan at the time of evaluation. If a plan is overfunded, immediate termination may be more advantageous, because part of the cost of termination would be paid out of the plan surplus and not by the plan sponsor. The analysis also supports that if a hibernating plan grows its funded status, either due to asset returns or interest rate movements, to the point where it could terminate, it should do so, because termination will be “costless” to the company (i.e., no contribution needs to be made to fund the termination; funding will be from the surplus generated in the plan). The decisions for an underfunded plan may be less clear. The cost of immediate plan termination would include both the termination premium and any deficit at the time, and thus could be very expensive. However, additional expenses would accrue during the hibernation period, including payment of a PBGC variable-rate premium based on any deficit. Further, an underfunded pension plan may adopt an investment policy that adjusts with funded status and is – unlike those shown in our paper’s stochastic analysis section – not static. An underfunded pension plan may adopt a liability-responsive asset allocation.9 In this dynamic allocation strategy the plan would reallocate from return seeking to liability-responsive fixed income strategies as funded status improves.

Corporate situation – The decision processes discussed above don’t take place in a vacuum, and although an NPV analysis is important, it is not the only important factor. Certain companies will opt for plan termination even when analysis points toward hibernation. For example, a pension plan may be large relative to the sponsoring organization, which will therefore be very sensitive to the financial risk the plan represents. The sponsor may choose upfront termination to avert exposure to future risks during a hibernation period. A similar decision might be made when a pension plan is small relative to the sponsoring organization. In such case a sponsor may be willing to take on the risk that hibernation may ultimately be more expensive than immediate termination, if such outcome is unlikely to adversely impact the organization. Conversely, a sponsor in this situation may be willing to terminate the plan up front, even if analysis favors hibernation, if doing so would enables the company to better concentrate on its core strategy.

Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

8

Conclusion An open and ongoing pension plan has many goals that must be evaluated and traded off over a long period of time. These plans could reasonably seek returns to lower future contributions, or could adopt liability-driven investment strategies designed to maintain funded status and minimize surplus volatility and thereby stabilize contributions. The chosen strategy may even change through time, based on funded status or the financial health of the sponsoring organization. Implicitly, all frozen pension plans are choosing between the strategies of upfront plan termination and plan hibernation. They are opting either to run the plan into the future or to terminate today because they believe, typically, that their choice will reduce costs. Once a frozen plan has opted to hibernate, the pension plan goal is to minimize costs – both present, ongoing and future (uncertain) costs – related to contributions and plan expenses, rather than to maximize portfolio returns. Thus, plans must be judicious about taking risk during the hibernation process. Excess return derived from the taking of investment risk could reduce the cost of hibernation relative to immediate termination, but could also increase the (longer-term) cost of hibernation over that of upfront plan termination. It is important to achieve consensus among those who operate the pension plan and to completely document decisions on the investment strategies to be used during the hibernation period. Those who have influence over the pension plan (committee members, board members, treasury staff, etc.), must not let themselves be sidetracked by attractive investment opportunities that may increase the costs and uncertainty associated with plan hibernation. Plans that don’t have confidence in their ability to maintain expenses and to invest significantly in liability-driven strategies may be better off terminating the plan, rather than adopting a hibernation strategy as hibernation may be more costly over time, as the cost of funding periodic deficits begins to eat into the favorable economics shown above.

Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

9

Appendix Data, methods and assumptions for stochastic study Evaluation date – January 1, 2014

Plan data Pension plan is frozen to new benefit accruals Initial liability value using corporate bond curve – $1 billion Initial liability value using Treasury bond curve – $1.11 billion Initial termination premium for comparison to the hibernation cost – $110 million PBGC flat rate premium based on participant count of 16,000 as of January 1, 2014 Investment portfolio: Return-seeking asset portfolio represented by diversified global equity portfolio index and liability-driven fixed income portfolio represented by a long credit fixed income index

Methods Hibernation period lasts until the plan can fully pay the termination premium with surplus (i.e., without an additional contribution) Plans that don’t have the ability to terminate during the study period (10 years) are assumed to fund the cost of plan termination after the 10th year Hibernation cost is calculated along each scenario as the sum of costs during the hibernation period, including: plan expenses, contributions and termination premium adjusted for any plan surplus or deficit Termination premium as of the time of hibernation is the difference between the liability calculated using a corporate bond yield curve and the liability using the Treasury yield curve Full 24-month asset smoothing chosen for funding purposes Full recognition of the Bipartisan Budget Act of 2013 for PBGC premiums Funding policy – Minimum required contribution under Pension Protection Act (2006) with recognition of all relevant amendments related to discount rate methodologies.

Assumptions Liabilities are valued in line with that of an insurance company looking to bid on the annuity purchase; therefore, no part of the insurance premium is based on differing demographic assumptions Recognizes 0.5% mortality loss annually as the only actuarial demographic loss; all other demographic assumptions are fully realized Participant count assumed to decrease each year with mortality (approximately 1% per year) Asset policies in the graphs are assumed to be static policies that are frequently rebalanced to targets Annual expenses (non-PBGC) – $5 million adjusted annually for inflation and asset size Russell capital market assumptions used as of 12/31/2013

Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

10

1

Some may think of hibernation to be a low-risk investment strategy adopted by a well-funded pension plan. For the subject of this paper, hibernation is framed in relation to costs, of which investment risk is only one aspect. I am using a more expansive definition of hibernation. I am using the term to mean the phase you enter after you decide not to terminate. This is a phase where the plan sponsor attempts to lower the costs (and increase the certainty of those costs) below the cost of immediate termination. My definition of hibernation allows for both overfunded and underfunded plans, as well as a variety of investment strategies. The important point is that the plan sponsor enters this phase with confidence that it will have less cost. I choose to call this “hibernation,” but it may rightfully go by another name. 2

I realize that termination is not an all-or-nothing decision in the sense that a plan sponsor may choose to purchase annuities for a subset of plan participants (retirees, for example) or choose to buy out other participants (terminated vested participants, for example) through a lump sum cashout program. Those are valid choices that fall between termination and hibernation and can be evaluated using a similar process as presented in this paper. 3

I refer specifically to an “insurance company” because at the current time only an insurance company can be the chosen firm for an annuity purchase after plan termination. Department of Labor Internal Bulletin 95-1 refers to an “annuity purchase issued by an insurance company.” 4

In this paper we proxy the plan termination premium as the difference between the present value of future benefit payments using a high-quality corporate bond yield curve and the present value of those same benefit payments using a US Treasury bond yield curve. 5

The development of this premium is beyond the scope of this paper. However we will discuss the calculation’s various inputs, what may influence those inputs, and how they are likely to change through the hibernation period. 6

Some may say a more accurate valuation of plan liability as the valuation of plan liability will undergo a more thorough underwriting process by the insurance company and adjustments may be made to decrement tables (mortality and retirement) that are more reflective of the plan population and its anticipated behavior. 7

The main reason why the reward is limited is that generally there is nearly no value to the corporation for any surplus beyond the point where it could terminate the pension plan. Any remaining assets after a plan termination can only revert to the company after it has been taxed at both the corporate tax rate and an additional excess tax, unless the surplus can be used in a qualified replacement plan. 8

In reality a company would not commit to a fixed hibernation period but would continually monitor the annuity purchase market for a price that they find appealing, the study chooses 10 years for illustration. In reality the hibernation period may be shorter or longer based on an ongoing cost benefit analysis. 9

See Gannon and Collie (2009): “Liability-responsive asset allocation: A dynamic approach to pension plan management,” Russell Investments.

ABOUT RUSSELL INVESTMENTS

Russell Investments is a global asset manager and one of only a few firms that offers actively managed multi-asset portfolios and services, which include advice, investments and implementation. Russell Investments stands with institutional investors, financial advisors and individuals working with their advisors—using our core capabilities that extend across capital market insights, manager research, asset allocation, portfolio implementation and factor exposures to help investors achieve their desired investment outcomes. FOR MORE INFORMATION:

Call Russell Investments at 800-426-8506 or visit russellinvestments.com/institutional Important information Nothing contained in this material is intended to constitute legal, tax, securities, or investment advice, nor an opinion regarding the appropriateness of any investment, nor a solicitation of any type. The general information contained in this publication should not be acted upon without obtaining specific legal, tax, and investment advice from a licensed professional. These views are subject to change at any time based upon market or other conditions and are current as of the date at the beginning of the document. The opinions expressed in this material are not necessarily those held by Russell Investments, its affiliates or subsidiaries. While all material is deemed to be reliable, accuracy and completeness cannot be guaranteed. The information, analysis and opinions expressed herein are for general information only and are not intended to provide specific advice or recommendations for any individual or entity. Russell Investments’ ownership is composed of a majority stake held by funds managed by TA Associates with minority stakes held by funds managed by Reverence Capital Partners and Russell Investments’ management. Frank Russell Company is the owner of the Russell trademarks contained in this material and all trademark rights related to the Russell trademarks, which the members of the Russell Investments group of companies are permitted to use under license from Frank Russell Company. The members of the Russell Investments group of companies are not affiliated in any manner with Frank Russell Company or any entity operating under the “FTSE RUSSELL” brand. Copyright © 2014-2016. Russell Investments Group, LLC. All rights reserved. This material is proprietary and may not be reproduced, transferred, or distributed in any form without prior written permission from Russell Investments. It is delivered on an "as is" basis without warranty. First used: March 2014 (Disclosure revision: July 2016) USI-19147-03-17

Russell Investments // Hibernation versus termination: Evaluating the choice for a frozen pension plan

11