In Brief

• Our proprietary analysis of trends among the 100 largest corporate pensions by assets found that the 2019 headline funded status rose only slightly as 30%-plus public equity returns were offset by record declines in pension discount rates.
  
  
• In aggregate, broad asset allocation was largely unchanged but sponsors again decreased expected return assumptions and increased fixed income duration. Meanwhile, employer contributions by the group dropped to a four-year low.
  
  
• While 80% of plan sponsors have outperformed their expected return assumptions over the last decade, only one-third have investment returns exceeding liability growth.

• Pension risk transfer activity increased but is yet to have a meaningful impact on the pension market. Lump sums were attractively priced for plan sponsors during 2019 but costs are slated to increase during 2020.


• For plan sponsors seeking to de-risk against a backdrop of higher volatility and market uncertainty, in addition to traditional long duration bonds, we recommend seeking protection in equity diversifiers such as low-beta hedge fund strategies; high income core alternative assets; option strategies; and hedge portfolio credit diversifiers.
  
  

In 2019, corporate pension plans experienced significant volatility but, as they did in 2018, ended the year essentially where they started. A heroic 31.5% return on the S&P 500 was largely offset by a historic decline in pension discount rates of 100 basis points (bps).

How have plan sponsors adapted their pension strategy to continuing funded status volatility, against a backdrop of late-cycle concerns, and rising asset prices? This report highlights historical trends and expectations for the industry, drawing on data from the largest 100 plans,¹ as well as insights from our engagements with J.P. Morgan’s corporate pension clients. The accompanying slide deck provides further context to elaborate on the ideas we discuss below.

After a decade of contributions and asset returns, dollar deficits are unchanged

The 2019 headline funded status for the top 100 corporate pension plans peer group was slightly improved, up 40bps from the prior year to 87.7%, but dollar deficits grew as both assets and liabilities ballooned. In fact, after 10 years of modest contributions and better-than-expected investment returns, duration mismatches and discount rate driven liability growth left the dollar deficit for our peer set right where it started (Exhibit 1A and 1B).

Although the top 100 plans’ funded status ended 2019 roughly where it began, there was significant volatility intra-year

EXHIBIT 1A: 2019 ESTIMATED INTRAYEAR FUNDED STATUS

Source: Company 10-K filings, J.P. Morgan Asset Management; data as of December 31, 2019. Numbers may not add up due to rounding. GAAP: Generally accepted accounting principles. For illustrative purposes only.

EXHIBIT 1B: ANNUAL GAAP FUNDED STATUS, TOP 100 PLANS

Source: Company 10-K filings, J.P. Morgan Asset Management; data as of December 31, 2019. Numbers may not add up due to rounding. GAAP: Generally accepted accounting principles.

During 2019, funded status peaked at 90.3% in April and bottomed out at 82.5% in August, tracking the horse race between bonds and equities (Exhibit 2). Similarly, year-to-date U.S. large cap stocks’ outperformance relative to U.S. Long Credit topped out in April, hit a trough in August and rebounded during the last third of the year as the bond rally faded against equities’ continued march higher (Exhibit 2). Investment performance, 19.0% on average, hit post-2008 global financial crisis highs but slightly trailed liability growth when service cost accruals are included. This felt a bit like 2012 and 2014, when liabilities outpaced double-digit asset returns, leading funded status lower for the year.

Public equity outperformed long credit in the first half and troughed in August; plans’ funded status followed suit.

EXHIBIT 2: STOCK-BOND RELATIVE PERFORMANCE DROVE FUNDED STATUS

Source: Bloomberg Barclays, Company 10-Ks, J.P. Morgan Asset Management; data as of March 1, 2020. For illustrative purposes only.

Employer contributions among the peer group of the top 100 corporate pension plans dropped to their lowest level since 2015, delivering a modest 2.0% to aggregate funded status—amounting to about 5% of the value of these plan sponsors’ share buybacks and dividend distributions. For the 2018 regulatory plan year, the majority of plan sponsors had no required contributions (33%) or used their credit balance (37%)—a shadow account tracking historical excess contributions—to satisfy required contributions. Absent new legislation or outsized market-driven funded status gains, we expect pension contributions to tick higher as pension relief wears away, pushing down regulatory discount rates during the next couple years.³

Asset allocation by corporate pension plans

As funded status and plan circumstances change, the required return to achieve funding objectives evolves with it. We continue to see plan sponsors re-underwrite their asset allocation positioning when return hurdles shift, for good or ill.

Asset allocations in aggregate were largely unchanged but individual plan sponsors within the peer set made some large shifts, both de-risking and re-risking. Seven plan sponsors boosted fixed income by 10% or more while five plan sponsors increased public equity by 10% or more. Few plans that increased their risk posture disclosed a change in target allocations, so the shifts toward public equity could be more a function of market-driven allocation drifts than a deliberate rebalancing toward risk assets.

Fixed income allocations were unchanged in aggregate but we tracked another year of increases in fixed income duration and interest rate hedge ratios. Our analysis of GAAP allocation data has limitations due to the non-standardized format and increasingly opaque disclosures as more and more sponsors group together assets valued at net asset value (NAV) as a practical expedient. For example, plan sponsors may be employing derivatives—for capital-efficient duration, leverage, or even as a form of dry powder for quickly putting on exposures to take advantage of market dislocations without having to be a forced seller elsewhere in the portfolio.

For a deeper insight, where transparency permits we’ve analyzed the sub allocations within public markets. Unsurprisingly, we found that fixed income portfolios are dominated by traditional government and credit issues, with only a small portion (less than 3.0%) identified as dedicated securitized allocations. As plan sponsors de-risk, their reliance on corporate credit exposure to match the liability increases, but so does the concentration of portfolio risk. We anticipate the portion dedicated to securitized, and other diversified, fixed income sectors will increase as de-risking plan sponsors seek to balance minimizing funded status volatility against liability outperformance. Within public equities, we find a slight home bias towards U.S. equity but geographic diversification in aggregate is roughly in line with market capitalization.

Asset allocations for the peer group are largely a function of plan characteristics. For example, plans with service cost accruals of less than 0.5% and funded status greater than 90% have an average fixed income allocation of 69%. That compares to a 48% fixed income allocation for the Top 100 corporate pension plans overall and 38% for plans below 80% funded with service cost accruals of greater than 1.5%.

While 80% of sponsors have outperformed expected returns, only one-third have outperformed liabilities

Expected return on assets (EROA) assumptions continued their monotonic decline, to a plan-weighted average of 6.8% and an asset-weighted average of 6.9%. The asset-weighted average is dominated by larger plans that tend to run with larger allocations to alternatives and, in some cases, employ higher alpha expectations. In 2010, 72% of plan sponsors had an EROA assumption of 8% or greater. In 2019, fewer than 10% assumed an 8% or higher EROA.

While pension plans of all types have been criticized for clinging to unrealistic return expectations, over the last decade 80% of the top 100 corporate pension plans have actually outpaced their EROA assumptions—the average excess return is almost 100bps (Exhibit 3A). Unfortunately, when we compare asset returns to liability returns, we find that only about one-third of plan sponsors have outperformed (Exhibit 3B).

Plans get criticized for unrealistic return expectations but 80% of the top 100 corporate pension plans have outpaced EROA assumptions over a decade. However, only one-third of plan sponsors have outperformed their liability growth.

EXHIBIT 3A: DISTRIBUTION OF TOP 100 PLANS’ 10-YEAR PERFORMANCE, ACTUAL VS. EXPECTED RETURN (BPS)

Source: Company 10-Ks, J.P. Morgan Asset Management; data as of March 1, 2020. For illustrative purposes only.

EXHIBIT 3B: TOP 100 CORPORATE PENSION PLANS’ 10-YEAR LIABILITY RETURN VS. ASSET RETURNS (%)

Source: Capital IQ, Company 10-Ks, J.P. Morgan Asset Management; data as of March 1, 2020. Liability returns include interest cost and actuarial (gain)/loss and do not include the impact of service cost accruals. For illustrative purposes only.

Mortality matters: Impacts of assumption changes

In October 2019, the Society of Actuaries (SOA) published an updated base mortality table, Pri-2012,³ and an updated mortality improvement scale, MP-2019, used to project future longevity enhancements. The Pri-2012 table, the first update since 2014, incorporates about twice as many data points, of which 40% were from Taft-Hartley plan experience. All else equal, most plan sponsors would see a slight decrease in pension liability from switching to these new tables, although increases are possible depending on the demographic and collar breakdown of the participants. Similarly, the updated mortality improvement scale, incorporating more recent data showing higher than projected deaths, will slightly decrease pension liabilities. For U.S. GAAP valuations, plan sponsors have flexibility in choosing the most appropriate mortality assumption and while many switched to these new tables, others made adjustments or used their own plan-specific data. Out of the Top 100 plans who disclosed detailed mortality assumptions, roughly 90% disclosed a switch to Pri-2012 or some customized variant. Exhibit 4 details some examples and the resulting impacts:

What are some of the outcomes of mortality assumption updates?

EXHIBIT 4: SAMPLE OF IMPACTS ON SELECTED COMPANIES

Source: Company 10-Ks, J.P. Morgan Asset Management; data as of December 31, 2019. For illustrative purposes only.

We also were bestowed with new, more conservative mortality tables, introduced for regulatory funding purposes for the first time in a decade, based on Internal Revenue Service guidance from October 2017. While the adoption was mandatory for lump sum calculations, plan sponsors were able to delay usage until their 2018 plan year contribution calculations. About 50% of plan sponsors opted to delay implementation, 45% adopted the new tables for 2018 and 5% used substitute mortality assumptions based on their own plan experience data. Switching from the prior to current legislation mortality assumptions could increase regulatory liabilities as much as 5%, likely a main driver of the low election rate. This delayed implementation is another factor that could have lowered sponsor contributions during the 2019 calendar year—and another reason to expect a future increase, as adoption of more conservative mortality tables becomes mandatory.

Pension risk transfer in 2019: No meaningful impact on pension market

Pension risk transfer (PRT) increased during 2019, with a total of $28 billion in single premium buyout sales, according to data from the Life Insurance Management Research Association (LIMRA). Despite the attention-grabbing headlines, PRT buyouts are not at this time making a meaningful impact on the overall pension market. We estimate that, on average, over the last five years, liabilities transferred through PRT were roughly 10% of total benefits disbursed by single-employer defined benefit plans, equivalent to roughly 1% of pension industry assets per annum. The LIMRA data shows 500 transactions during 2019 equated to an average deal size of $50 million (Exhibit 5A and 5B). At this pace, annuity buyouts would not meaningfully transform the corporate pension industry’s structure.

EXHIBIT 5A: ANNUITY BUYOUT TRANSACTIONS FROM 2010 THROUGH 2019

Source: Life Insurance Management Research Association, Company 10-K filings, J.P. Morgan Asset Management; data as of March 1, 2020.

EXHIBIT 5B: ANNUITY BUYOUTS IN THE CONTEXT OF TOTAL SINGLE EMPLOYER DB DISBURSMENTS

Life Insurance Management Research Association, 5500 EBSA Abstract, Company 10-K filings, J.P. Morgan Asset Management; data as of March 1, 2020. Total disbursements and assets from 2010 through 2017 are sourced from DOL 5500 filings for all single-employer DB pension plans. Total disbursements for 2018 and 2019 are estimated using the 10-year CAGR of 3.9% per annum. Total assets for 2018 are estimated from ICI Retirement Statistics and 2019 total assets are estimated by using the year-over-year change observed in our Top 100 plans data.

From a plan sponsor cost perspective, 2019 was a great year for lump sum windows. Exhibit 6 shows the discount rates used to calculate lump sum minimum present values, which are commonly locked for the year based on the level in November of the prior year.⁴ The peak in corporate bond yields coincided precisely with the lock period. This ensured that 2019 lump sum costs would be cheaper than 2018, all else equal. Lump sums became even more attractive as the year progressed and spot yields continued to rally: the last four months of 2019, lump sums could be paid at less than 90 cents on the dollar, relative to GAAP balance sheet liability carrying values.

Lump-sum discount rates locking in November 2018 made lump-sum offers an attractive option for plan sponsors in 2019

EXHIBIT 6: LUMP SUM DISCOUNT RATES

Source: Internal Revenue Service, J.P. Morgan; data as of March 1, 2020. Lump sum discount rates are generally locked for the year based on the setting of November of the prior year. The lookback and stability period can and do vary by plan. For illustrative purposes only.

Many plan sponsors offered lump sum windows during 2019 (Exhibit 7)—to take advantage of attractive pricing or for other strategic reasons, including PBGC premium optimization.

Many plan sponsors offered lump-sum windows in 2019, to take advantage of attractive pricing or for other strategic reasons

EXHIBIT 7: EXAMPLES COMPANIES THAT MADE LUMP SUM OFFERINGS

Source: Company 10-Ks, J.P. Morgan Asset Management; data as of December 31, 2019. Where possible, calculations are based on U.S. qualified pension liabilities only. PBO: Projected benefit obligation.

In fact, 2019 was a milestone year: it had the highest level since 2012 of settlements, liability distributions including buyouts, and lump sums that exceed a certain threshold. If we exclude General Motor’s 2012 risk transfer activities, 2019 was the largest settlement year in our data set. We note that the use of PRT—both lump sum and buyouts—has evolved. It has become more of a surgical instrument than a sledgehammer in the de-risking toolbox, along with asset allocation and liability design changes.

An example of a pension strategy that encapsulates this ethos comes from Ford Motor Company, one of the Top 100 plans that can reasonably be considered to be at, or at least near, its end game. As Ford stated in its 2019 10-K filing:

Our strategy is to reduce the risk of our funded defined benefit pension plans, including minimizing the volatility of the value of our pension assets relative to pension liabilities and the need for unplanned use of capital resources to fund the plans. The strategy reduces balance sheet, cash flow, and income exposures and, in turn, reduces our risk profile. Going forward, we expect to:

• Limit our pension contributions to offset ongoing service cost or meet regulatory requirements, if any;
  
  
• Maintain target asset allocation of about 80% fixed income investments and 20% growth assets, which better matches plan assets to the characteristics of the liabilities, thereby reducing our net exposure; and
  
  
• Evaluate strategic actions to reduce pension liabilities, such as plan design changes, curtailments, or settlements
  
  

While smaller PRT transactions are the norm, Bristol-Myers Squibb (BMY) bucked the trend by terminating its $4.0 billion principal U.S. plan.⁵ The plan paid out $1.3 billion in lump sums to those who elected them and the remaining $2.6 billion was transferred to Athene through the purchase of a group annuity. The termination also resulted in a $1.6 billion pension settlement charge, immediately absorbing all previously unrecognized losses associated with the plan through the income statement. Perhaps a cautionary tale, the plan was overfunded by roughly $425 million at termination, a common concern among plan sponsors contemplating contributions to already well-funded pensions. It appears that Bristol has chosen to use these excess assets to fund contributions to a qualified replacement plan (QRP).⁶ If at least 25% of the surplus is transferred to a QRP, the excise tax on any reversion of remaining surplus to the sponsor would decrease from 50% to 20%.

PBGC cost increases—and the darker side of smoothing

PBGC premiums paid by the top 100 corporate pension plans rebounded from the declines of 2018, increasing from 16bps to 23bps of plan assets. The size of the increase was a bit counterintuitive and requires a dive into some of the mechanics of how premiums are derived. Recall that premiums are comprised of a flat-rate premium (FRP) charged per participant, and a variable-rate premium (VRP) charged on the unfunded vested deficit but capped at a ceiling of $541 per participant for 2019. When valuing the liability, plan sponsors can choose between the Standard or Alternative method discount rate. The Standard is based on spot corporate bond yields, similar to U.S. GAAP.⁷ The Alternative is based on 24-month smoothed corporate bond rates, without regard to pension relief.

Plan sponsors that used the Standard method experienced large drops in both assets and liabilities, leaving funded status slightly improved on average (Exhibit 8). However, plan sponsors using smoothed rates experienced the same asset declines while the benefit of rising liability discount rates was delayed by the smoothing effects, leading to larger unfunded deficits. The intermingling of smoothed values and market values in an asset-liability management (ALM) framework is a recurring battle for corporate pensions across accounting, regulatory and PBGC frameworks. We generally find that when the focus of an investment program shifts to minimizing funded status volatility, the costs of smoothing outweigh the benefits. The smoothing effect also occurred against the backdrop of higher VRP and FRP rate levels. Plans with a surplus would still have seen their premiums increase about 8%, all else equal, as the FRP rose from $74 to $80 per participant.

Using a smoothed discount for PBGC liability valuation can drive ALM mismatches

EXHIBIT 8: YEAR-OVER-YEAR CHANGE IN PBGC LIABILITY VALUATION BY DISCOUNT RATE ELECTION

Source: Internal Revenue Service, Society of Actuaries, J.P. Morgan; data as of March 1, 2020. For illustrative purposes only. Calculations based on hypothetical 12-year duration liability.

There are several proposals that seek to address the burden of PBGC premiums on single-employer plans. A bill, The Pension Budget Integrity Act seeks to de-link premiums from the federal budget. The President’s Budget proposed for fiscal year 2021 calls for a freeze on inflation indexing but an increase in the VRP cap to $900 per participant.⁸ PBGC premiums have been a big motivator of pension risk transfer, especially small balance annuity buyouts and lump sum offers. However, we think that most of the low-hanging fruit has been picked in terms of premium optimization transactions. Future risk transfer activity will more likely be driven by relative value and idiosyncratic corporate actions related to the pension.

Conclusion: Our recommended approaches for 2020

Plans are navigating a more volatile market environment in 2020 with continued uncertainty about interest rate levels, equity valuations and questions about the probability, and potential severity, of a possible recession. For those plan sponsors seeking to reduce risk but wary of adding duration at current rate levels, we would recommend considering the following approaches during 2020:

• Equity diversifiers, such as quantitative hedge fund strategies, that offer novel equity diversification: low beta and low correlation.
  
  
• High income alternative assets, including global infrastructure equity, which can dampen funded status volatility without increasing the hedge ratio. To be risk-reducing, these strategies should generally be on the core end of the spectrum, meaning that a large portion of the total return comes from accurately forecastable cash income.
  
  
• Option strategies that can provide explicitly defined protection against market events and can be implemented in both the rates and equities spaces.
  
  
• For those plan sponsors adding to hedge portfolios, be wary of the risks lurking within concentrated corporate credit portfolios and look for diversification across a wider opportunity set, balancing liability tracking error against liability outperformance.

Uncertainty creates risks, but also, for plan sponsors positioned to act—seeking to reduce risk but wary of buying duration at these rate levels—we hope our 2019 findings and recommendations can help uncover “Once in a lifetime” opportunities.

¹ Largest plans by assets as of December 31, 2019. Analysis excludes international plans and/or non-qualified plans where data transparency makes this separation possible. Where transparency is not available, figures may include these elements. Plan sponsors with fiscal-year ends between May 26, 2019 and January 4, 2020 are included. Analysis includes only public companies. Because of rounding, some numbers such as percentage change from start-to end-of-year headline funded status) may appear not to add up.
² For a discussion of the drivers and implications of pension relief wear-away, see “Michael Buchenholz, “Pension indigestion: Considerations for the end of regulatory relief,” J.P. Morgan Asset Management, December 2019,
https://am.jpmorgan.com/us/en/asset-management/institutional/investment-strategies/pension-strategy/pension-indigestion-considerations-for-the-end-of-regulatory-relief/
³ “Pri-2012 Private Retirement Plans Mortality Tables,” Society of Actuaries, 2012.
⁴ Actual lump-sums costs will depend on the plan’s specific stability and lookback periods, which may differ from this analysis.
⁵ Due to the termination of its primary U.S. pension plan during the year, BMY is not included this year in the Top 100.
⁶ A QRP must cover at least 95% of the active participants in the terminated plan and if it is a defined contribution plan, the amounts transferred must be allocated to participants over a period of not more than seven years.
⁷ Note that an election to use the Alternative Premium Funding Target is irrevocable for a period of five years.
⁸ For more on the budget proposal’s implications, see Michael Buchenholz, “The 2021 budget proposal’s impact on PBGC premiums: Potential implications for Pension plans,” J.P. Morgan Asset Management, February 2020,
https://am.jpmorgan.com/us/en/asset-management/institutional/investment-strategies/pension-strategy: The 2021 budget proposal’s impact on PBGC premiums.

0903c02a8283d7c6