The 2017 edition of the Milliman Corporate Pension Funding
Study (PFS) marks the 17th annual analysis of the financial
disclosures of the 100 largest corporate defined benefit (DB)
pension plan sponsors. These 100 companies are ranked highest
to lowest by the value of their pension assets reported to the
public, to shareholders, and to the U.S. federal agencies that
have an interest in such disclosure.
These pension plans finished 2016 with pension assets of
$1.395 trillion and projected benefit obligations (PBO) of
$1.718 trillion. The funded ratio at the end of 2016 was 81.2%.
This was comparable to the funded status at the end of 2015
of 81.9% and 2014 of 81.6%.
But a ratio is just one number divided by another, which doesn’t
tell the complete story. The 0.7% decline in the pension funded
ratio and the resulting $22 billion fiscal year (FY) 2016 funded
status deterioration, masks:
- Volatility in the interest rate environment, which caused the
discount rate to decline steeply by 30 basis points in FY2016
- FY2016 pension trusts’ investment return of 8.4% compared to
the average return expectation for the Milliman 100 companies
of 7.0% in 2016.
- Employers’ 2016 plan contributions of $42.8 billion, a 38.0%
increase from $31.1 billion in 2015.
- The second consecutive fiscal year of further decreases in
future life expectancy resulting in an approximate $3.8 billion
reduction in the actuarially determined benefit obligations for
six Milliman 100 companies for which the value of the change
was disclosed (IBM, GM, GE, AT&T, Verizon, and 3M).
- Seven companies exited the Milliman 100. For various
reasons, seven plan sponsors had plan assets decline enough
in 2016 that they are no longer Milliman 100 companies
covered in the 2017 PFS:
- CBS, Computer Sciences, RR Donnelley, J.C. Penney, Macy’s,
PPG, and Sears.
- HP has changed as it divested into two companies: HP and
HP Enterprises, both of which were included in our study.
- Seven companies joined the Milliman 100: HP Enterprises,
Ameren, Kimberly Clark, Molson, Travelers, US Bancorp, and
FY2016 pension expense (the charge to the income statement
under Accounting Standards Codification Subtopic 715)
decreased 7.6% to about $30.8 billion from about $33.3 billion
Forty-six of the Milliman 100 companies in this 2017 PFS
indicated they have adopted or plan to adopt a “spot rate”
approach for calculating pension expense. Thirty-seven
companies included such disclosures for our 2016 PFS.
Pension risk transfer transactions and strategies from the
plan sponsors to insurance companies continued in 2016.
(Perhaps not news to the readers of the Milliman PFS, but
enough news for pension risk transfer transactions to make
page A1 of the March 13, 2017, Wall Street Journal.) FY2016
pension risk settlement payments to former but not yet
retired participants continued as well. We’ve estimated that
the sum of the pension risk transfers to insurance companies
(sometimes referred to as “pension lift-outs”) and the
settlement payments totaled about $13.6 billion compared
with $11.6 billion in FY2015. Examples among Milliman 100
companies are: Westrock ($2.5 billion), United Technologies
($1.6 billion), Hewlett Packard ($1.3 billion), Verizon ($1.3
billion), International Paper ($1.2 billion), and Pepsi ($1.0
billion). These pension risk transfer strategies also relieve the
plan sponsor of the Pension Benefit Guaranty Corporation
(PBGC) premium payments that are required for these former
employees who are part of the participant head count.
FIGURE 1: HIGHLIGHTS (FIGURES IN $ BILLION)
||FISCAL YEAR ENDING
|MARKET VALUE OF ASSETS
|PROJECTED BENEFIT OBLIGATION
|NET PENSION INCOME/(COST)
|ACTUAL RATE OF RETURN
|Note: Numbers may not add up precisely due to rounding.
In addition to defined benefit pension plans, the PFS has been
tracking the actuarial obligations of postretirement health care
benefits. FY2016 marks the first year that the aggregate reporting
of these accumulated postretirement benefit obligations
(APBOs) is under $200 billion, at about $197.9 billion. This is
consistent with the trend of divesting other postemployment
benefits (OPEB) liabilities by plan sponsors over the last decade.
As we write this report in April 2017, we acknowledge that our
year-end 2016 results do not reflect the gains in the financial
markets since January 1, nor do they reflect any volatility in
interest rates that were caused by the Federal Reserve’s actions
in March to raise short-term interest rates by 25 basis points.
For 2017 results, see our April 2017 Pension Funding Index (PFI),
which will review the funded status changes that occurred
during the first quarter of 2017. The April PFI will be released
within one week of the release date of this report and will reflect
the FY2016 results included in the 2017 Pension Funding Study.
The most significant factors offsetting the adverse effect of
the decrease in discount rates on the funded ratio were the
favorable investment returns and an increase in employer
contributions during 2016:
1. The actual return on the pension trusts was 8.4% when
the expectation was an investment return of 7.0%—an
investment gain of $16 billion.
- For the eight years 2009 to 2016, there has always been a net
investment increase over the asset value at the start of the
year, and except for 2011 and 2015, pension plan investment
gains have exceeded the expected return set at the start of the
fiscal year. We also note that in those eight years, pension
plan assets allocation to equities has decreased to about
36.0%, from about 46.0%, while fixed income allocation has
increased to about 44.0% from about 36.0%.
2. Employers increased 2016 cash contributions by almost $12
billion compared with 2015. About $43 billion was contributed
in 2016, compared with about $31 billion in 2015. Among other
contribution strategies, there may have been a desire to reduce
the PBGC premium applicable to pension plan underfunding.
Pension expense in 2016 declined $2.5 billion to $30.8 billion from
$33.3 billion in 2015. One may reasonably conclude that pension
expense will decrease in 2017 because of the favorable 2016 asset
performance, however, the effect of lower discount rates, plan
settlements, and more plan sponsors adopting the spot interest
rate method will also need to be considered.
Future reductions in pension expense can result from changes
in the assumptions under which pension expense is calculated.
Forty-six of the Milliman 100 companies indicated in Form
10-K that they have adopted or plan to adopt a “spot rate”
approach to calculate the interest cost component of pension
expense, which is a refined use of the individual “spot” interest
rates on the corporate bond yield curve used to develop the
actuarial liabilities or projected benefit obligation (PBO).
For an upwardly sloping yield curve, the use of the spot rate
method is expected to lower interest cost and therefore total
pension expense in comparison with using the former singleweighted
average discount rate methodology. In fact, if all
of the Milliman 100 companies were to adopt the spot rate
accounting method to calculate the interest cost component
of pension expense in 2017, the pension expense savings is
estimated to be $11.1 billion. This calculation assumes a 20.0%
reduction in the interest cost for a typical company in the
Milliman 100 study adopting the spot rate methodology.
De-risking transactions continued in 2016, and the estimated
dollar volume of pension risk transfers collected from the
accounting disclosures was higher in 2016 ($13.6 billion)
compared with 2015 ($11.6 billion). Note that the federal
Department of Labor (DOL) prefers the use of “pension risk
transfer” (PRT) when referring to these transactions in which
the complete divestiture of DB plan obligations to participants
or to insurance companies occurs.
PRT transactions may continue to occur in 2017, spurred by the
significant increases in the premiums payable to the PBGC.
Strong year for investment
returns—especially for plans with
significant allocations to U.S. versus
FIGURE 2: ESTIMATED RATES OF RETURN EARNED IN 2016 FOR
THE 88 PLANS WITH CALENDAR FISCAL YEARS BY THEIR
ALLOCATION TO FIXED INCOME
Rates of return earned in 2016 for the 88 pension plans with
calendar year fiscal years ranged from 0.5% to 18.4%, with an
average of 8.7%. Returns mostly fell in the 6.0% to 12.0% range
(73 plans) with seven plans earning returns below 6.0% and eight
plans earning returns above 12.0%. Generally, plans with greater
allocations to fixed income earned slightly higher returns, but
differences in the basic asset allocation among equities, fixed
income, and other assets did little to explain differences in returns.
Instead, we believe the return differences were most likely
explained by the plans’ allocations to U.S. equities and to U.S.
corporate bonds, especially long-duration bonds. U.S. equities
significantly outperformed non-U.S. equities in 2016, and U.S.
corporate bonds significantly outperformed U.S. government
bonds and interest rate swaps in 2016. Plans with heavy allocations
to fixed income as part of a liability-driven investment (LDI)
strategy typically have significant allocations to long-duration
corporate bonds. Returns on long-duration corporate bonds
almost matched the returns on U.S. equities, with both of these
assets contributing double-digit returns in 2016.
FIGURE 3: FIXED-INCOME ALLOCATION 50% OR HIGHER
(CALENDAR YEAR FISCAL YEARS ONLY)
50% OR HIGHER
Equity allocations in the pension portfolios dropped to 36.1%
by the end of 2016. The companies comprising the Milliman
PFS have generally shifted toward higher allocations to fixed
income investments. This trend has surfaced as plan sponsors
reconfigured allocations to de-risk their pension plans over the
past several years.
The actual asset return for the plan sponsor with the highest
allocation to equities (81.1%) was 12.0%, which was a little better
than the return of 10.4% for the plan sponsor with the lowest
allocation to equities (7.0%) in 2016. The highest asset return
among all companies with calendar year fiscal years was 18.4%,
while the lowest was 0.5%.
In prior years, investment allocations made by plan sponsors
had showed a trend toward implementing LDI strategies.
Generally, this involves shifting more assets into fixed income
positions. This trend continued in 2016. The percentage of
pension fund assets allocated to equities, fixed income, and
other investments was 36.1%, 44.1%, and 19.8%, respectively,
at the end of 2016, compared with 37.4%, 42.6%, and 20.0%,
respectively, at the end of 2015.
Unlike in 2015, when plans with high allocations to fixed
income (over 50.0%) performed comparably to the other plans
(0.06% average return compared with -0.07%), in 2016 the plans
with high allocations to fixed income outperformed the other
plans (9.84% compared with 8.38%).
Over the last five years, the plans with consistently high
allocations to fixed income have slightly underperformed the
other plans while experiencing lower funded ratio volatility.
Among the 88 companies in the Milliman PFS with calendar
year fiscal years, 23 pension plans had fixed income allocations
greater than 40.0% at the end of 2011 and maintained an
allocation of at least 40.0% to fixed income through 2016. Over
this five-year period, these 23 plans experienced lower funded
ratio volatility than the other 65 plans (an average funded ratio
volatility of 4.4% versus 6.2% for the other 65 plans) while
earning a slightly lower five-year annualized rate of return (an
average of 8.0% versus 8.4%). Similar to 2015, when these 23
plans outperformed relative to the other 65 plans (0.3% average
return versus -0.2%), they also outperformed the other plans in
2016 (10.0% average return versus 8.3%).
FIGURE 4: ESTIMATED AVERAGE RATE OF RETURN BY ALLOCATION
TO FIXED INCOME – 2014-2016
Overall allocations to equities decreased during 2016, resulting in
an average allocation of 36.1%—the lowest equity allocation in the
17-year history of the Milliman PFS. None of the 100 companies
had increases to its equity allocations of more than 10.0% in 2016.
Only three companies decreased their equity allocations by more
than 10.0% in 2016, compared with four companies in 2015, 11 in
2014, five in 2013, three in 2012, and 12 in 2011.
FIGURE 5: ASSET ALLOCATION – EQUITIES
FIGURE 6: ASSET ALLOCATION – FIXED INCOME
Overall allocations to fixed income increased in 2016, resulting
in an average allocation of 44.1%. Only two companies had a
decrease of more than 10.0% in their fixed income allocations.
Five companies, however, increased their fixed income
allocations by more than 10.0% in 2016, compared with three in
2015, seven in 2014, four in 2013, two in 2012, and seven in 2011.
FIGURE 7: ASSET ALLOCATION – OTHER
Other asset classes include real estate, private equity, hedge
funds, commodities, and cash equivalents. More specific details
on how investments are allocated to the other categories are
generally not available in the companies’ U.S. Securities and
Exchange Commission (SEC) filings. Overall allocations to
other asset classes remained stable in 2016. Eight companies
increased their allocations by 5.0% or more to other asset
classes during 2016.
For comparison purposes, we have looked at historical
changes since 2005, the first year when the Milliman 100
companies consistently made allocation information available.
The allocation to equities was down from 61.7% at the end of
2005 and the allocation to fixed income instruments was up
from 28.8% at the end of 2005. The percentage of investments
in other asset classes was also up from the 9.6% allocation at
the end of 2005.
Pension Risk Transfer activities continue
Similar to the past few years, plan sponsors continued to execute
pension risk transfer (PRT) activities in 2016 as a way of divesting
pension obligations from their DB plans and corporate balance
sheets. Large-scale pension buyout programs were transacted for
six of the Milliman 100 companies, as pension assets and liabilities
were transferred to an insurance company. Westrock, United
Technologies, Hewlett Packard, Verizon, International Paper, and
Pepsi reported transactions of $2.5 billion, $1.6 billion, $1.3 billion,
$1.3 billion, $1.2 billion, and $1.0 billion, respectively. PPG and J.C.
Penney, former Milliman 100 companies not included in the 2016
Study, had PRT transfers during 2016 of $2.3 billion and $1.6 billion,
respectively. These settlements were significant enough to drop
these companies from the largest 100 plan sponsor companies
listing of the Milliman PFS.
The 2016 PRT market was slightly more active when compared
with the 2015 market. Extracting the dollar volume of PRT
activities from Form 10-K is an estimate and it appears that
the dollar volume in 2016 was $13.6 billion, an increase of $2.0
billion compared with the 2015 dollar volume of $11.6 billion.
PRTs are deemed an effective way to reduce a pension plan’s
balance sheet footprint by plan sponsors, but generally they have
an adverse effect on the plan’s funded status, as assets paid to
divest accrued pension benefits are higher than the corresponding
actuarial liabilities that are extinguished from plans. Much of this
incongruity stems from Internal Revenue Service (IRS) pension
plan valuation rules differing from an insurance company’s
underwriting assessment of its new future risks.
Last year, we reported that a more prevalent de-risking measure
came in the form of a “lump-sum window” program, in which
some plan sponsors settled the pension obligation by distributing
a payment to a specific group of former participants. However,
the IRS issued Notice 2015-49 that effectively and permanently
ended the ability of a plan sponsor to offer a lump-sum settlement
to retirees or their surviving beneficiaries who were collecting
annuities. On the other hand, lump-sum offerings via windows to
terminated vested plan participants continued in 2016 and more
are expected in 2017 as well as plan sponsors’ rush to divest some
additional liability before the new IRS-required mortality tables
applicable for determining lump-sum distribution amounts from
qualified retirement plans possibly become effective.
Last year, we also reported on an analysis of mortality
experience of participants in all U.S. DB plans. While we don’t
plan to delve into the development of life expectancy factors
in this study, we reference below a couple of noteworthy items
that will affect funded status.
1. A further refinement of the mortality study by the Society
of Actuaries in October 2016, reduced expected rates
of mortality improvements. The revisions shorten life
expectancy by a few years and reduce the fiscal year-end
PBO. While we are unable to collect specific details of the
reduction for all companies included in our study, there were
several noteworthy disclosures regarding the impact on this
refinement on the pension liability for several companies.
General Motors, AT&T, General Electric, IBM, Verizon, and
3M each noted that the adoption of the refined mortality
improvement scale reduced their pension and OPEB
liabilities by $888 million, $793 million, $600 million, $600
million, $500 million, and $440 million respectively.
2. We’ve been reporting for a few years that the IRS has
planned to update the federal pension regulations that
dictate the rates of mortality used for the valuation of DB
pension plan actuarial obligations. In December 2016, the
IRS finally proposed such regulations, based on the abovementioned
Society of Actuaries report, with an expected
effective date of plan years starting after December 31, 2017.
We remain somewhat pessimistic about these regulations
affecting the 2018 contributions and, in particular, lump-sum
payments for plans that offer these one-time settlements.
The executive orders issued by President Trump in January
2017, freeze or delay the effective date of all federal agencies’
proposed regulations, and the president has not yet set up
review teams he has proposed for the IRS (or DOL), who will
serve as the authorities to review and set regulation release
dates. The “mortality table regulations” are one of those
The Bipartisan Budget Act of 2015 included increases in the
PBGC’s flat rate and variable rate premiums. These premiums
are paid to insure certain accrued pension benefits to
participants if an employer sponsoring a single-employer DB
pension plan becomes insolvent.
The flat dollar amount increases to $69 in 2017 from $64 in 2016.
The variable rate premium has increased to 3.4% of the pension
plan’s PBGC-funded status deficit in 2017, from 3.0% of the 2016
deficit. (PBGC’s funded status deficit uses interest rates and
mortality assumptions that are different from the funded status
of the Milliman 100 companies.)
The 2016 funded ratio of 81.2% was slightly higher than we
reported in the January 2017 Milliman 100 Pension Funding
Index (PFI). The January 2017 PFI funded ratio of 81.0% was
based on data collected for the 2016 Milliman Pension Funding
Study. This revised funded ratio of 81.2% from our current
study reflects the collection and collation of publicly available
information. Investment returns and contributions during 2016
were higher than we had forecasted, both key factors for the
higher funded ratio.
The higher-than-expected cash contributions during 2016 are
likely in response to rising PBGC premiums. As background,
PBGC variable rate premiums are calculated based on a plan’s
funded status determination, without respect to interest rate
funding relief as afforded to plan sponsors under the Employee
Retirement Income Security Act (ERISA), for purposes of
minimum funding requirements. Faced with the prospect of
escalating PBGC premiums in future years, many plan sponsors
have developed strategies to narrow their funded status gaps
sooner than what may be required based on minimum funding.
This often results in higher-than-required (under IRS rules)
cash contributions being made.
Impact of decreasing discount rates
evident in 2016 financial statements
of the Milliman 100 companies
Discount rates used to measure plan obligations, determined
by reference to high-quality corporate bonds, decreased during
2016, thereby increasing liabilities and reversing the trend from
the prior year. The median discount rate decreased to 3.99%
at the end of 2016 from 4.29% in 2015. The 3.99% discount rate
at the end of 2016 was the lowest in the 17-year history of the
Milliman PFS. Discount rates had been generally declining from
7.63% at the end of 1999. Discount rates were 237 basis points
higher at the end of 2008.
The impact of the decreasing discount rates in 2016 and increased
PBO was partially offset with a favorable investment gain of 8.4%.
This resulted in a modest decrease in the funded status. The 2016
funding deficit of $323.4 billion is a $21.7 billion increase over the
year-end 2015 funding deficit of $301.7 billion. It is the fourthlargest
deficit in the 17-year history of the Milliman PFS.
FIGURE 8: PENSION FUNDING SURPLUS/(DEFICIT)
Pension expense—the charge to company earnings—decreased
to $30.8 billion in 2016 as compared with $33.3 billion during
fiscal year 2015, a $2.5 billion decrease. The peak level of
pension expense occurred in 2012, when it was $55.9 billion.
In addition, 46 of the Milliman 100 companies indicated that
they have adopted or plan to adopt a “spot rate” approach
for estimating the service and interest cost components of
net periodic benefit costs. Thirty-seven companies included
such disclosure for our 2016 PFS. This approach is likely to
produce a pension expense savings in the near term. In spite
of the decrease in discount rates during 2016, the 2017 pension
expense is not likely to increase significantly, which is due to
the investment gains experienced during 2016 and the change
in accounting method to a spot rate approach.
The “spot rate” approach, which is a refined use of the individual
“spot” interest rates on the corporate bond yield curve used to
develop the actuarial liabilities or PBO. This contrasts with the
measurement of PBO utilizing a customized bond matching
model. The plan sponsor can choose to use only one of the
valuation methodologies, and cannot change it each year unless
there is agreement with the auditors to do so.
For an upwardly sloping yield curve, the use of the spot rate
method is expected to lower the “interest cost” component
of pension expense, thus lowering the total pension expense
in comparison with using the former single-weighted
average discount rate methodology. This method leads to an
expectation of PBO losses when the PBO is remeasured at the
end of fiscal year 2017 for pension disclosure.
We’ve estimated an $11.1 billion decrease in FY2017 pension
expense if all Milliman 100 companies adopted the spot rate
accounting method to calculate the interest cost component.
We’ve made the assumption of a 20.0% reduction in the
interest cost for each of the other 54 Milliman 100 companies
for this calculation.
FIGURE 9: PLAN ASSETS AND OBLIGATIONS
The effect of a decrease of 30 basis points in discount rates was
partially offset by the favorable investment returns during 2016,
revisions to life expectancy assumptions, and the impact of
The net 3.2% increase in pension obligations generated by
the decrease in discount rates, revisions to life expectancy
assumptions used to measure pension plan obligations (at a
median rate of 3.99% at year-end 2016, down 30 basis points
from 4.29% at year-end 2015), and PRT activity resulted in a
liability increase of $54.0 billion. Pension liabilities for IBM
and General Motors remained below the $100 billion pension
obligation mark in 2016, which helped their plans to improve
their funded status, assisted by investment gains of 8.3% and
The 8.4% investment gain (actual weighted average return
on assets during 2016) resulted in an increase of $32.3 billion
in the market value of plan assets when combined with the
higher contributions, and approximately $13 billion paid out in
annuity purchases or lump-sum settlements. The Milliman 100
companies had set an expectation that 2016 investment returns
would be, on average, 7.0%.
Funded ratios barely decrease
The funded ratio of the Milliman 100 pension plans decreased
during 2016 to 81.2% from 81.9% at the end of 2015 (81.4% for
plans with calendar year fiscal years in 2016, down from 81.7%
for 2015). The aggregate pension deficit increased by $12.8
billion during these calendar year companies’ 2016 fiscal years
to $292.8 billion, from an aggregate deficit of $280.0 billion at
the end of 2015. For fiscal year 2016, funded ratios ranged from a
low of 49.0% for Delta Airlines to a high of 148.0% for NextEra
FIGURE 10: FUNDED RATIO – ASSETS/PROJECTED BENEFIT
The 0.7% decrease in the 2016 funded ratio reversed the 2015
increase. The funded ratio had been 81.6% at the end of 2014.
Note that there has not been a funding surplus since the 105.8%
funded ratio in 2007.
Only eight of the 88 Milliman 100 companies with calendaryear
fiscal years reported surplus funded status at year-end
2016, compared with nine companies in 2015, eight in 2014, 19 in
2013, and six in 2012. These numbers pale in comparison with
the 50 companies with reported surplus funded status at yearend
2007. Because of the offsetting impact of higher investment
returns and the increase in liabilities caused by lower discount
rates, only 38 of the Milliman 100 companies with calendar
fiscal years reported an increase in funded ratio for 2016
compared with 51 for 2015.
FIGURE 11: DISTRIBUTION BY FUNDED STATUS – 2011-2016
(CALENDAR YEAR FISCAL YEARS ONLY)
2016 pension expense decreases
There was a net decrease in 2016 pension expense: a $30.8
billion charge to earnings ($2.5 billion lower than in 2015). This
is well below the $55.9 billion peak level in 2012. Twenty-two
companies recorded 2016 pension income (i.e., a credit to
earnings). Sixteen companies also recorded income in 2015 and
17 companies in 2014 and 2013, up from 10 in 2012.
The discount rate for 2016 pension expense is based on the
year-end 2016 SEC disclosures. We estimate that 2017 pension
expense will decrease to $23.5 billion, a $7.3 billion decrease
compared with 2016, under the assumption of a continued
3.99% discount rate. This reduction includes an estimated $5.1
billion decrease in expense that is due to 46 of the Milliman
100 plan sponsor companies adopting the spot rate method for
calculation of the interest cost component of pension expense.
FIGURE 12: PENSION EXPENSE (INCOME) AND CONTRIBUTIONS
The aggregate 2016 cash contributions of the Milliman 100
companies were $42.8 billion, an increase of $11.7 billion from
the $31.1 billion contributed in 2015, and an $18.7 billion decrease
from the 2012 record high level of $61.5 billion. Contributions
had been decreasing in 2015 and 2014 ($31.1 billion and $40.0
billion, respectively) from the 2013 level of $43.5 billion.
However, the contributions of $42 .8 billion in 2016 reverses this
trend. We speculate that this is due to increased contributions
by many plan sponsors to minimize their PBGC premium
increases, as discussed earlier.
Many plan sponsors may continue to contribute at these higher
levels for 2017 if they can’t find better uses for their corporate
cash. We expect that some plan sponsors undertaking PRT
activities (e.g., lump-sum payouts, annuity purchases, etc.) may
have to increase contributions to maintain funded status. Also
some plan sponsors that want to minimize PBGC premium
costs may choose to accelerate plan funding to close pension
Pension deficit decreases slightly as a
percentage of market capitalization
The total market capitalization for the Milliman 100 companies
increased by 10.3%. The increase in pension obligations (which
is due to lower discount rates) resulted in a small decrease
in the unfunded pension liability as a percentage of market
capitalization of 4.3% at the end of 2016 from 4.4% at the end
of 2015. Pension deficits represented less than 10.0% of market
capitalization for 74 of the Milliman 100 companies in 2016 and
75 of the Milliman 100 companies in 2015 and 2014 (down from
81 companies in 2013). However, this is still an increase from
2012, when only 60 companies had deficits that were less than
10.0% of their market capitalizations.
Since 2011, we have had investment returns exceeding
expectations in four out of five years, and this has resulted in
elevated levels of market capitalization. There is one company
whose deficit exceeds 50.0% of market capitalization in 2016,
down from two companies in 2015. This is down from nine in
2012, the year we first started tracking this figure.
FIGURE 13: UNDERFUNDED PENSION LIABILITY AS A PERCENTAGE
OF MARKET CAPITALIZATION 2013-2016
The weighted average investment return on pension assets
for the Milliman 100 companies’ 2016 fiscal years was 8.4%,
which was above their average expected rates of return of 7.0%.
Seventy-three of the Milliman 100 companies exceeded their
expected returns in 2016, including all six that had off-calendar
fiscal years. Only three companies exceeded their expected
returns in 2015 and all three had off-calendar-year fiscal years.
But 80 companies in 2014 exceeded their expected returns
compared with 77 in 2013, 93 in 2012, 20 in 2011, and 98 in 2010.
The 2016 investment return was favorable, and now investment
returns above expectations during six out of the last eight years
have been earned by the plan sponsors of the Milliman 100
companies. At December 31, 2016, total asset levels were
$1.395 trillion. This is $109.7 billion above the value of
$1.285 trillion at the end of 2007, prior to the collapse of the
worldwide financial markets.
During 2016, investment gains offset by the combination
of annuity purchases and lump-sum settlements increased
the market value of assets by $32.3 billion. The Milliman
100 companies’ actual investment return for 2016 was $110.4
billion compared with the expected return of $94.4 billion, a
difference of $16.0 billion. For the five-year period ending in
2016, investment performance has averaged 8.28% compounded
annually. There were three years of investment losses over
the past 17 years (2001, 2002, and 2008), contributing to an
annualized investment return of only 6.0% over that period.
FIGURE 14: INVESTMENT RETURN ON PLAN ASSETS
Expected rates of return
Companies continued to lower their expected rates of return
on plan assets to an average of 7.0% for 2016, as compared with
7.1% for 2015, 7.3% for 2014, 7.4% for 2013, 7.6% for 2012, 7.8% for
2011, and 8.0% for 2010. This represents a significant drop from
the average expected rate of return of 9.4% back in 2000.
Only one of the Milliman 100 companies utilized an expected
rate of return for 2016, 2015, 2014, and 2013 of at least 9.0%,
whereas three companies also assumed an expected rate of
return of at least 9.0% in 2012, 2011, and 2010, but this was down
from five in 2009 and a high of 84 in 2000.
What to expect in 2017 and beyond
Our expectations in the coming year include:
- Contributions are expected to stay at their current levels,
given the anticipated desire of plan sponsors to fund in
excess of IRS minimum requirements in an effort to stave off
PBGC premium increases.
- Pension expense is expected to decrease compared with the
2016 level, which is due to reductions in interest cost as a
result of the interest rate spot method and the investment
gains experienced during 2016. The expense reduction will
be tempered by the drop in discount rates and the resulting
decline in pension funded status during 2016.
- PBO losses are expected at year-end 2017 due to the use of
the aforementioned spot rate methods for determining the
interest cost component of pension expense.
- PBO gains are expected due to further refinements in
- Further pension risk transfer activities should occur depending
on the movement of discount rates and asset returns in 2017.
HISTORICAL VALUES (All dollar amounts in millions. Numbers may not add up correctly due to rounding.)
The table below shows the trend of the divestiture of OPEB liabilities from $317 billion in 2003 to $198 billion in 2016.
Who are the Milliman 100 companies?
The Milliman 100 companies are the 100 U.S. public companies
with the largest defined benefit pension plan assets for which a
2016 annual report was released by March 5, 2017.
This 2017 report is Milliman’s 17th annual study. The total value
of the pension plan assets of the Milliman 100 companies was
more than $1.39 trillion at the end of 2016.
About the study
The results of the Milliman 2017 Pension Funding Study are
based on the pension plan accounting information disclosed in
the footnotes to the companies’ Form 10-K annual reports for
the 2016 fiscal year and for previous fiscal years. These figures
represent the Generally Accepted Accounting Principles
(GAAP) accounting information that public companies are
required to report under Financial Accounting Standards
Board (FASB) Accounting Standards Codification Subtopics
715-20, 715-30, and 715-60. In addition to providing the financial
information on the funded status of their U.S. qualified pension
plans, the footnotes may also include figures for the companies’
nonqualified and foreign plans, both of which are often
unfunded or subject to different funding standards from those
for U.S. qualified pension plans. The information, data, and
footnotes do not represent the funded status of the companies’
U.S. qualified pension plans under ERISA.
Twelve of the companies in the 2017 Milliman Pension
Funding Study had fiscal years other than the calendar year.
The 2017 study includes seven new companies to reflect
mergers, acquisitions, and other corporate transactions during
2016. Figures quoted from 2015 reflect the 2017 composition
of Milliman 100 companies and may not necessarily match
results published in the 2016 Milliman PFS. Generally, the
group of Milliman 100 companies selected remains consistent
from year to year. Privately held companies, mutual insurance
companies, and U.S. subsidiaries of foreign parents were
excluded from the study.
The results of the 2017 study will be used to update the Milliman
100 PFI as of December 31, 2016, the basis of which will be used
for projections in 2017 and beyond. The Milliman 100 PFI is
published on a monthly basis and reflects the effect of market
returns and interest rate changes on pension funded status.
About the authors
Zorast Wadia, FSA, EA, MAAA, is a principal and consulting
actuary in the New York office of Milliman. He has more
than 17 years of experience in advising plan sponsors on their
retirement programs. Zorast has expertise in the valuation of
qualified and nonqualified plans. He also has expertise in the
areas of pension plan compliance, design, and risk management.
Alan H. Perry, FSA, CFA, MAAA, is a principal and consulting
actuary in the Philadelphia office of Milliman. He has more
than 20 years of experience in advising plan sponsors on asset
allocation and financial risk management. Alan specializes
in the development of investment policies by performing
asset-liability studies that focus on asset mix, liability-driven
investing, and risk hedging.
Charles J. Clark, ASA, EA, MAAA, is a principal and director
of the employee benefits research group in the Washington,
D.C., and New York offices of Milliman. He has over 36 years of
experience as a consulting actuary. Charles provides analysis
of employee benefit legislation, regulations, and accounting
standards to Milliman consultants. He has worked extensively
with plan sponsors, Washington, D.C., employee benefits trade
groups, and lawmakers on employee benefit program strategy,
design, pricing, and interpretation.
The authors thank the following Milliman colleagues for their
assistance in compiling the figures and editing the report
for the Milliman 2017 Pension Funding Study: Keila Cohen,
Mary Der, Karen Drake, Rebecca Driskill, John Ehrhardt,
Jeremy Engdahl-Johnson, Jennifer Faber, Kevin Ferris, Chris
Goodman, Heng Lim, Elizabeth Mattoon, Janie Pascual, Jamie
Phillips, Lesley Pink, Rebecca Ross, Javier Sanabría, Mike
Wilson, Susan Yearick, Lynn Yu, and Delbert Zamora.