How does your retirement plan stack up?

How does your retirement program stack up?


		E-mail this page				Print this page

Contact: Charles Salmans
Tel: +1 212 345 4512

The new retirement plan landscape - changes, challenges and opportunities

Last updated: 20 October 2007

Earlier this year, Mercer published highlights of its analysis of retirement plan data disclosed by the Standard & Poor’s 500 companies in their 2006 10-K reports.

The analysis tracks important benchmarks in the investment policies, contributions and design of retirement programs – from traditional defined benefit (DB) pension and defined contribution (DC) plans, to postretirement medical and life insurance plans. We’ve now completed more detailed analysis in our full report, How does your retirement program stack up?

You can download a complete 36-page PDF of the Mercer report or an 18-page PDF of the highlights on the right side of this page.

This full report expands the key findings outlined in the highlights report, providing additional detail on the funded status and funding, investment and benefit policies of the companies in the S&P 500. We also include results specific to the following economic sectors:

Consumer discretionary
Consumer staples
Energy
Financials
Health care
Industrials
Information technology
Materials
Telecommunication services
Utilities

Some of the key findings relating to the comparison of these sectors include:

Relative to revenue, it is industrials, materials and utilities that have the highest cost for ongoing retirement benefits of all types; the energy sector has the lowest cost.
The best-funded pension plans are in the telecommunication services and financials sectors (financials includes banks that have unique incentives to fund at higher levels).
The sectors with the worst-funded status – in terms of pension plan funded status percentage – are information technology and health care. However, under funding as a percentage of market capitalization is highest for the utilities, materials and industrials sectors (industrials includes the airline industry).
The consumer staples group was the most negatively affected by the balance-sheet recognition provisions of FAS 158; utilities were the least affected (largely due to offsetting regulatory assets) and financials were a close second.

Pension plan funded status improves

The financial health of pension plans improved significantly during 2006. Pension plan health, as measured by a plan’s funded status, improved mostly due to strong asset returns and a rise in discount rates. This combination – the opposite of the 2000 – 2002 “perfect storm” – helped to increase median funded status for S&P 500 plan sponsors from 83 percent in 2005 to 89 percent in 2006.

Plan sponsor contributions, which in past years have added significantly to funded status, played a lesser role in 2006. During 2006, median plan sponsor contributions were only slightly larger than the value of benefits earned during the year.

Exhibit 1 shows the improvement in pension funded status. Funded status is defined as the ratio of a plan’s assets to its liabilities. For this exhibit, we have compared the market value of plan assets to the total projected benefit obligation (PBO). The PBO reflects the pension benefits employees have earned based on current service and includes the projected effect of future salary increases.

Asset returns exceed expectations

Better-than-expected asset returns help to close the gap between plan assets and liabilities, thereby improving plan funded status. For the fourth year in a row, plan asset returns met or exceeded expectations. The median actual return for 2006 was 13.4 percent, compared with the median expected return of 8.2 percent. On the whole, 2006 was a very good year for asset performance – at least 90 percent of plans exceeded expectations, with even the 10th percentile actual return at 8.8 percent. Exhibit 2 illustrates the distribution of 2006 plan asset returns and compares the results to those of the four previous years.

Liability growth

Plan asset returns do not entirely account for the change in a plan’s funded status. The change in plan liabilities must also be considered. Just as asset returns are measured as the net percentage increase in assets after contribution and benefit payments have been reflected, we define liability “return” as the net percentage increase or decrease in plan liabilities after adjusting for benefit accruals, benefit payments, and any plan amendments or curtailments. A positive return – an increase in liabilities – implies a decrease in the funded ratio; similarly a negative return – a
decrease in liabilities – improves the funded ratio.

Two factors account for most of the liability return: the interest “earned” on the liability during the year and the change in the value of the liability resulting from any change in the discount rate. Absent changes in the discount rate, the liability return should generally be quite close to the discount rate at the beginning of the measurement period. However, this is rarely the case, because discount rates are adjusted at each measurement date to reflect current market yields.

When the discount rate increases, the liability decreases (a negative liability return); when the discount rate decreases, the liability increases (a positive return). In addition to the two discount rate related factors, the liability return may be affected by other sources of gain or loss, such as the difference between assumed and actual experience or changes in assumptions.

The median rate of return on liabilities during 2006 was 6.0 percent, slightly more than the 2005 median discount rate of 5.5 percent. Experience losses, primarily due to the effects of currency conversion for non-US plans, were sufficient to offset negative returns produced by a slight increase in discount rates. Without the effect of currency conversion losses, the median return on plan liabilities was 4.2 percent, well below the 5.5 percent return expected. Exhibit 3 illustrates the distribution of liability returns for the past three years.

The improvement in funded status in 2006 can thus be summarized as the net effect of the difference between plan asset and liability returns. Using median asset and liability returns as an example (and assuming a hypothetical plan with a 2005 funded status of 83 percent), the net improvement in 2006 funded status can be illustrated as follows:

Discount rates

As Exhibit 4 illustrates, weighted average discount rates increased in the last year, from 5.50 percent in 2005 to 5.75 percent in 2006. Rates increased consistently across all percentiles, with 2006 rates approximately 25 basis points higher than 2005 rates. As in previous years, there continues to be only moderate variation in discount rates across companies. The spread from the 10th to the 90th percentile ranged from 5.16 percent to 6.00 percent at the end of 2006.

This variation is not unusual and reflects normal differences among plans, such as:

Differences in measurement date
Differences in plan population – plans with a higher percentage of retirees may have a different discount rate than plans with a higher percentage of younger active employees
Differences in plan design – cash balance plans that pay lump sums may have a different discount rate than annuity based plans
Differences in methodologies used by plan sponsors to estimate the appropriate discount rate

Gradual changes in asset allocation

While the target allocations to equities among plans in the S&P 500 declined by three percentage points over the past four years, actual allocations to equity and to real estate were little changed, at 63 percent and 2 percent of plan assets, respectively, as illustrated in Table 1 below. The target allocation to assets in the “other” category experienced a two-percentage point increase since 2003, which reflected a modest expansion in aggregate allocations to “alternative assets” – private equity, timberland, infrastructure, hedge funds, etc. – to provide diversification benefits as well as enhanced performance opportunities. Individual plan allocations to alternatives varied widely, from zero in many cases to 15 percent or more in the top 10 percent of cases.

Risk and reward

Although plan sponsors are concerned about the investment return volatility associated with equity exposure, the majority have not chosen to reduce their equity allocations. Some plan sponsors have lengthened the duration of their fixed income portfolios to reduce funded status volatility by increasing the correlation between assets and liabilities. But most have chosen to retain the higher expected return to be gained from high levels of equity exposure and have kept fixed income investments benchmarked to the Lehman Aggregate Bond Index. The strong equity returns of the past four years may have erased the painful memories of the “perfect storm” years of extreme funded status volatility and may continue to make equity-heavy asset allocations attractive.

It remains to be seen what impact the 2006 Pension Protection Act (PPA) and mark-to-market accounting rules will have on plan sponsors’ asset allocation decisions. Outside the US, where mark-to-market accounting and funding rules have been in effect for some time, plan sponsors have reduced equity exposure and lengthened fixed income portfolio durations. Perhaps we will see the same trend take shape in the US.

Managing pension plan risk

Pension plans help companies manage many different types of risk. Benefit design policies are typically targeted toward human resource objectives and risks – attracting, retaining and retiring the right employees with an appropriate level of retirement income and an appropriate allocation of risk and cost between plan sponsor and participant. Traditional DB pension plans enable participants to mitigate retirement income risk; they also allow employers to pool longevity risk, take advantage of economies of scale and offer retirement benefits at favorable “group” rates. However, pension plans also create risks for companies. In the typical pension arrangement, the plan sponsor bears the financial risk of providing the promised pension benefits, regardless of asset performance; in a DC plan, this financial risk is borne by individual participants.

The financial risk of the pension plan is typically mitigated by increasing the correlation between the performance of assets and liabilities. As we discussed in the section on liability returns, the current discount rate – and subsequent changes in the discount rate – are the main drivers of liability return. The main tool for improving the correlation between asset and liability performance is thus to increase the plan assets’ exposure to fixed income instruments that are similarly responsive to the discount rate.

Our recent Perspective, "The new world of pension plan financial management – Optimally balancing risk and return," discusses the concept of pension financial risk in greater detail – and how investment, contribution and benefit policies can be coordinated to help manage it.

Funding policy keeps pace with benefits earned

Plan sponsor contributions are another important component of the increase in funded status, particularly contributions over and above the value of benefits earned during the year. For the past several years, sponsors have been making such contributions to help improve plan funded status. However, in 2006, it appears that many companies have funded these legacy costs, as cash contributions were much closer to the cost of new benefit accruals.

For the median plan sponsor, cash contributions in excess of the cost of new benefit accruals represented only 0.42 percent of the plan’s PBO. Thus, only 0.42 percent of the improvement in the median plan’s funded status can be attributed to sponsor contributions. (See Exhibit 5) In contrast, cash contributions to fund legacy costs of poorly funded plans represented 7.79 percent of PBO. However, for both the median and 90th percentile plans, contributions in excess of benefit accruals were down significantly from 2005.

An additional reason for the decline in cash contributions may be the introduction of FAS 158. FAS 158 eliminated the minimum liability rules that could have resulted in large balance sheet charges if assets were less than the accumulated benefit obligation (ABO). Because there was no longer any accounting benefit to be gained from fully funding the ABO, plan sponsors may have had less incentive to make 2006 contributions in excess of the value of benefits earned during the year.

Pension Protection Act

Under the 2006 PPA, new funding rules will require US plan sponsors to fund each year’s benefits as they are earned and any unfunded accrued benefits over a seven-year period. PPA’s funding target is similar to the ABO; we thus use the ABO as a proxy for the funding target and show in Exhibit 6 that the median plan sponsor was 98 percent funded on an ABO basis. Note that the actual funded ratio for the subset of US qualified plans – those to which the PPA applies – may be slightly larger, as the reported ABO includes plans that are typically not advance-funded, such as nonqualified plans or certain non-US plans. Once their legacy costs are fully funded under PPA, plan sponsors will have the opportunity to focus on investment and design policies that maintain funded status and mitigate risk.

FAS 158

In September 2006, the Financial Accounting Standards Board (FASB) issued Statement No. 158, which requires plan sponsors to recognize in their balance sheets the net difference between plan assets and benefit obligations – for both pension and other postretirement benefit plans. Benefit obligations for this purpose are the PBO for pension plans and the accumulated postretirement benefit obligation (APBO) for other postretirement benefit plans.

Prior to the adoption of FAS 158, corporations were required to record a balance sheet adjustment only if pension assets were less than the ABO. After FAS 158, the balance sheet adjustment applies to both pension and other postretirement benefit plans, regardless of funded status. As a result of this new statement, many plan sponsors recorded an after-tax charge to shareholder equity at the end of 2006 to adjust for the difference between the new and old accounting standards. Most corporations with non-calendar fiscal years will record this adjustment in 2007.

Last year, we estimated that the median after-tax reduction in shareholder equity would have been about 3.2 percent in 2004 and 2.8 percent in 2005, had the new standard always been in effect. Largely due to 2006 asset returns that exceeded liability growth, the actual median effect was a reduction of only 1.7 percent for 304 companies that reported adopting FAS 158 in 2006. Reductions in equity for the middle 50 percent of companies ranged from 0.4 percent to 5.2 percent. These results are shown in more detail in Exhibit 7.

Despite dire predictions of the consequence of putting these liabilities on the balance sheet, analysts and investors may have already factored unfunded postretirement benefit obligations into their thinking; we note that the S&P 500 has recently reached record highs. And, we think that most companies have successfully managed the effect of the accounting change on loan covenants, reimbursement rates, compensation programs or other areas that depend on equity metrics.

Future accounting changes

Discussions continue at both the FASB and the International Accounting Standards Board (IASB) on the proper accounting for pension and other postretirement benefits. We expect these discussions to result in progress toward convergence of the two standards and, possibly, immediate recognition in income of all changes in funded status, including gains and losses and prior service costs.

On a parallel track, the accounting standard-setters are also working on financial statement presentation – considering how companies report "income” and how income is divided into categories such as “operating,” “financing,” “investment,” or “gains and losses.” A change in pension accounting rules might thus be more palatable to companies if it is coupled with broader accounting rule changes that would permit volatile gains and losses to be recognized outside of “operating” income. At present, neither the IASB nor the FASB have issued any definitive guidance; although the FASB is expected to publish a Preliminary Views document on financial statement presentation in the fourth quarter of 2007.

Separately, the SEC has recently proposed allowing non-US companies to issue securities in the US markets without reconciling their financial reporting under International Financial Reporting Standards (IFRS), promulgated by the IASB, to US accounting standards. That proposal is currently open for comment. Of course, if non-US companies are not required to follow US standards, there will inevitably be pressure to permit US companies to do the same, and the SEC may invite comment on these issues later this summer.

Postretirement benefit expense as a percentage of revenue

In Exhibit 8 and Exhibit 9, we examine how much money the S&P 500 companies spent on total retirement benefits – pensions, savings plans and PRM benefits. These exhibits divide the expense between “operational” (service cost for pension and other postretirement benefits plus contributions to DC plans) and “legacy” (interest and amortization costs) components. To facilitate comparisons between companies, we have scaled retirement benefit costs as a percentage of reported revenue; the costs can thus be evaluated in context of their effect on pretax profit margins.

Since 2003, total median operational expense as a percentage of revenue declined about 8 percent, from 0.90 percent to 0.83 percent. The range of costs for the second and third quartiles showed a similar change: 0.40 percent to 1.23 percent in 2006, down from 0.51 percent to 1.35 percent in 2003. The change in this ratio is primarily driven by revenues growing faster than postretirement benefit expense. Over the four year period from 2003 through 2006, total revenue grew at an average rate of 8.8 percent per year, compared to an average growth rate of 5.9 percent per year for postretirement benefit operational expense.

Of the 484 companies in the survey, 99 reported no DB or postretirement obligations and thus have no “legacy” costs. Of the remaining 385 companies, median legacy costs were about 0.13 percent of revenue, or about 13 percent of total postretirement cost. However, for 10 percent of the companies, legacy costs consumed at least 0.86 percent of revenue, nearly a third of total retirement benefit expense. As shown in Exhibit 9, legacy costs have come down somewhat since 2003, as some of the “perfect storm” losses have been amortized. Given the improvement in funded status at year-end 2006, we expect further declines in legacy costs in 2007.

The potential size of the legacy costs is also a reminder that pension risks can have a very long tail – for nearly half the companies in our survey, the bad news of 2000 – 2002 still had a significant effect on pension cost in 2006; this effect will probably continue for at least several more years.

When we divide the total operational expense into its DB, DC and PRM components (see Exhibit 10), we see that the cost for DC plans has remained relatively constant over the 2003 – 2006 period and that much of the decrease in total retirement spending is attributable to DB plans.

Conventional wisdom holds that many plan sponsors are freezing their DB plans and replacing them with DC benefits. To examine this hypothesis, we computed operational expense as a percentage of the benefit obligation. To the extent that sponsors are freezing their plans, we expect to see an increasing proportion of plans with low (or zero) operational cost to benefit obligation ratios. (Because of the way that companies aggregate plans and sometimes report the annual administrative cost of the plan as “service cost,” it is possible that a company with a substantially frozen plan might have a nonzero service cost to benefit obligation ratio.)

Exhibit 11 shows the distribution of these ratios in 2003 and 2006. In 2003, 43 plans out of 371, or 12 percent, had ratios of less than 1 percent. By 2006, this had grown to 48 out of 366 plans, or 13 percent, indicating that while there has been some movement away from DB plans, there has not, so far, been a stampede to the exits. We will be examining these issues in more detail in our full report.

Summary and outlook

2006 was a year of significant change for DB plan sponsors – the Pension Protection Act and FAS 158 combined to cause plan sponsors to rethink many aspects of their pension programs.

Financially, the PPA and FAS 158 drive home several key points:

Pension plans are an important – and integral – part of a company’s financials.
The appropriate measure of a pension plan’s impact on the company’s financials is the net surplus (deficit) in the plan, on a mark-to-market basis.
The liabilities in pension plans are properly viewed as debt-like promises, separate from the underlying assets, and should be valued accordingly. Plan sponsors that invest in equities, betting that outperformance of these investments will offset the growth in plan liabilities, may find that the risks of doing so outweigh its rewards.

Although the details are yet to be worked out in regulations, the PPA also offers some interesting new opportunities for pension plan design, such as leveling the playing field between DC and cash balance plans or offering the promise of phased retirement to help manage the human capital aspects of the baby boomer population, the leading edge of which is fast approaching retirement age.

We expect plan sponsors to continue re-evaluating their pension programs in light of the new parameters of FAS 158 and PPA. With mark-to-market valuation now required for both funding and balance sheet recognition, the calculus for deciding on the mix and level of DB and DC benefits is changing. We will be keeping a close watch on these changes, to spot new trends and patterns that might emerge.

Strong equity returns and slightly higher market discount rates generally helped to improve plans’ funded status in 2006; that strong performance has continued so far in 2007. Coupled with the PPA requirements to fund pension deficits relatively quickly, plan sponsors may soon reach the point where legacy costs are fully funded, allowing them to focus on design policies that provide sustainable long-term operational cost and investment policies that mitigate the risk associated with legacy costs.

In short, the landscape is changing, and plan sponsors should be evaluating all aspects of their retirement program management in the context of the marketplace and the new operational rules. This review should reflect an integrated framework, such as Mercer’s Integrated Retirement Financial Management (iRFM), so that all policy levers – design, funding and investing – are aligned and optimized with a sponsor’s objectives.

If you would like more information on any of these topics or on how your particular retirement program compares to its peers in the survey, please contact your local Mercer consultant.

About the survey

The survey includes information gathered from the 10-K filings for the S&P 500 companies with fiscal years ending February 2006 through January 2007. Sixteen of the S&P 500 constituent companies as of January 1, 2007 either filed their 10-Ks too late to be included in this report or filed no 10-K due to recent privatizations. Of the 484 companies included in our survey, 367 sponsor a defined benefit (DB) pension plan, many in combination with a defined contribution (DC) or retiree medical plan.

An additional 18 companies have a retiree medical plan but no DB pension plan, and 99 companies sponsor only a DC plan. Total reported pension obligations for the 367 companies that sponsor a pension plan are $1.5 trillion, compared with total reported assets of $1.45 trillion.

About Mercer

Mercer is the global leader for trusted HR and related financial advice, products, and services. In our work with clients, we make a positive impact on the world every day. We do this by enhancing the financial and retirement security, health, productivity, and employment relationships of the global workforce.

Mercer has more than 15,000 employees serving clients from more than 180 cities and 42 countries and territories worldwide. As a wholly owned subsidiary of Marsh & McLennan Companies, Inc., we can also provide access to the complementary consulting services of our sibling companies, Oliver Wyman, Lippincott and NERA Economic Consulting.

About Mercer's investment consulting

Mercer is a leading global provider of investment consulting services, and offers customized guidance at every stage of the investment decision, risk management, and investment monitoring process. We have been dedicated to meeting the needs of clients for more than 30 years, and work with the fiduciaries of pension funds, foundations, endowments, and other investors in 35 countries.

Mercer Inc., is a wholly owned subsidiary of Marsh & McLennan Companies, Inc. (MMC). MMC lists its stock (ticker symbol: MMC) on the New York, Chicago, Pacific, and London stock exchanges. In the US, the investment consulting practice is operated through Mercer Investment Consulting, Inc., a wholly owned subsidiary of Mercer Human Resource Consulting, Inc., the US operating unit of Mercer Human Resource Consulting LLC.

Disclaimer

This report contains confidential and proprietary information of Mercer and is intended for your sole use. The report, and any opinions on or ratings of investment products it contains, may not be modified, sold, or otherwise provided, in whole or in part, to any other person or entity without Mercer's written permission.

Mercer research documents and opinions on investment products (including product ratings) are based on information that has been obtained from the investment management firms and other sources. The views expressed here are those of the author and are based on information obtained by Mercer from sources that are believed to be reliable. Mercer gives no representations or warranties as to the accuracy of such information, and accepts no responsibility or liability (including for indirect, consequential or incidental damages) for any error, omission or inaccuracy in such information other than in relation to information which Mercer would be expected to have verified based on generally accepted industry practices.

Any opinions on or ratings of investment products contained herein are not intended to convey any guarantees as to the future investment performance of these products. In addition:

Past performance cannot be relied upon as a guide to future performance.
The value of stocks and shares, including unit trusts, can go down as well as up, and you may not get back the amount you have invested.
The value of bonds and other fixed income investments, including unit trusts, can go down as well as up, and you may not get back the amount you have invested.
Investments denominated in a foreign currency will fluctuate with the value of the currency.
The value of investments in real property can go down as well as up, and you may not get back the amount you have invested. Valuation is generally a matter of a valuer’s opinion, rather than fact. It may be difficult or impossible to realize an investment because the property concerned may not be readily saleable.
The performance of with-profit policies depends on the profits declared by the insurance company and how these are distributed. Deductions for charges and expenses incurred by the insurance company are greater in the early years, and this affects the amount payable on early surrender.