Professional Documents
Culture Documents
Americas/United States
Accounting & Tax
Equity Research
research team
• We would not be surprised to see the FASB add a major project on pension
accounting to its agenda this year, especially now that it’s done with employee
stock options.
• With that said, we have updated our analysis on what would happen to the
earnings and balance sheets of the companies in the S&P 500 if we replace the
magic of pension accounting with what actually happened to the pension plans.
We find over $1.1 trillion in off-balance-sheet pension liabilities and nearly $900
billion in off-balance-sheet pension assets.
ANALYST CERTIFICATIONS ARE IN THE DISCLOSURE APPENDIX. FOR OTHER IMPORTANT DISCLOSURES,
visit www.csfb.com/ researchdisclosures or call +1 (877) 291-2683. U.S. Disclosure: CSFB does and seeks to do
business with companies covered in its research reports. As a result, investors should be aware that the Firm may have a
conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in
making their investment decision. Customers of CSFB in the United States can receive independent, third party research on
the company or companies covered in this report, at no cost to them, where such research is available. Customers can access
this independent research at www.csfb.com/ir or call 1 877 291 2683 or email equity.research@csfb.com to request a copy of
this research.
The Magic of Pension Accounting, Part III 07 February 2005
Table of Contents
Executive Summary ..........................................................................................................3
Conclusion..................................................................................................................... 9
Our Key Findings ............................................................................................................10
Pension Reform ..............................................................................................................16
The Administration’s Proposal.....................................................................................17
Timing..........................................................................................................................20
Pension Plan Asset Allocation ........................................................................................22
Pension Plan Asset Allocation at the Macro Level......................................................23
Pension Plan Asset Allocation—The S&P 500 ...........................................................25
Do the Asset Allocations Make Sense? ......................................................................33
An Asset Allocation Shift .............................................................................................38
Who Has Exposure? .......................................................................................................42
Forecasting..................................................................................................................43
Our S&P 500 Forecast ................................................................................................44
Shareholders’ Exposure ..............................................................................................47
Cash Flow Exposure ...................................................................................................48
Earnings Exposure ......................................................................................................50
Balance Sheet Exposure .............................................................................................52
Sensitivity Analysis ......................................................................................................53
The Impact of Changing the Expected Return ............................................................55
Pension Report Card.......................................................................................................59
The Magic of Pension Accounting ..................................................................................62
Three Steps to Strip out Pension Accounting Magic...................................................63
Step 1—Put the Pension Plan Assets and Obligation on the Balance Sheet .............65
Step 2—Calculate “Real” Pension Expense or Income: Show Me the Volatility.........72
Step 3—Divide Pension Expense/Income into Three Buckets ...................................80
Summary .....................................................................................................................90
Analysis of Assumptions .................................................................................................91
Overview......................................................................................................................91
Expected Rate of Return on Plan Assets ....................................................................91
Discount Rate ..............................................................................................................97
Salary Inflation Rate ..................................................................................................101
Do the Assumptions Make Sense?...............................................................................104
The SEC Inquiry ........................................................................................................104
Expected Return........................................................................................................108
Discount Rates ..........................................................................................................118
Funded Status...............................................................................................................125
Industry and Company Concentrations .....................................................................126
Underfunded..............................................................................................................128
Overfunded................................................................................................................129
Quality of Earnings........................................................................................................131
Appendix A: S&P 500 Companies That Don’t Have a Defined Benefit Pension Plan..139
Appendix B: Historical and Projected Funded Status for the S&P 500 ........................140
2
The Magic of Pension Accounting, Part III 07 February 2005
Executive Summary
With a strong stock market over the past two years, we have noticed a drop off in
interest among equity investors in our favorite topic: defined benefit pension plans.
However, we think that’s about to change, as 2005 is getting set to be a big year for
pension plans for three key reasons:
1. Pension reform. The Pension Benefit Guaranty Corporation (PBGC) taking over US
Airways’ remaining pension plans and possibly United Airlines’ pension plans could
be the spark that lights the pension reform fire in Washington D.C. The president’s
plan for pension reform, released on January 10, laid out a number of potential
paths reforms might take, including reform the funding rules, reform the premiums,
and enhance disclosures. Ever since the president’s plan was released, we’ve
heard lots of speculation about whether or not pension reform could cause a shift in
asset allocation for defined benefit pension plans away from equities and toward
fixed income. That could bode well for the bond market but poorly for the stock
market. Some even think it is happening already and attribute the recent flattening
of the yield curve to pension plans buying long-dated bonds.
2. Pension accounting. The short list for major projects on which the FASB and IASB
would like to work together includes three items: leasing, intangibles, and last but
certainly not least, pensions. In our view, it’s just a matter of time before the FASB
adds a major project on pension accounting to its agenda. In fact, we would not be
surprised to see it on the agenda before the end of this year, especially if real
pension reform catches fire. A project on pension accounting would be beyond
contentious; the uproar from companies would be deafening, and some could even
threaten to shutdown their defined benefit pension plans. That threat would get the
unions up in arms to protect an enormously valuable benefit for their members.
(Interesting that you could have companies and unions arguing the same side of an
issue; not something you see every day.)
Why? What in the world could the little ol’ FASB propose that would get everyone
so worked up? A complete overhaul of pension accounting. We expect the FASB to
propose scrapping the current model with all of its various smoothing mechanisms
and replace it with one where the real economics of the pension plan are reported in
the financial statements instead of being tucked away in a footnote. In other words,
the board could propose to put the pension plan assets and obligation on the
balance sheet at their fair values—with changes in each going through earnings, or
as a compromise through shareholders equity—a type of mark-to-market approach,
similar to the methodology we first described in The Magic of Pension Accounting.
3. The SEC. Then there is the SEC, which has a renewed focus on defined benefit
pension plans. For one, pensions will be part of the SEC’s upcoming report to
Congress on off-balance-sheet activity. Second, the SEC staff has made clear that
companies need to do a better job of supporting their pension assumptions. And
last, pensions are among the risk-based initiatives for the SEC’s Enforcement
Division, which is currently investigating the pension accounting for six companies.
3
The Magic of Pension Accounting, Part III 07 February 2005
With that as the backdrop, we bring you The Magic of Pension Accounting, Part III. The
third in a series of fast-paced, action-packed reports covering the thrills, chills, and spills
of defined benefit pension plans. We have not only updated some of the more popular
analyses from our prior reports, we have also included lots and lots of new stuff. In this
report we focus on five key themes:
1. Pension reform. With the president’s plan for pension reform now out in the open,
we provide some of our initial thoughts on the reform proposals, and their potential
timing.
2. Asset allocations. With companies required to disclose pension asset allocation
information for the first time, we grabbed this data, analyzed it, and attempted to
figure out whether the allocations make sense. In addition, we provide our thoughts
on whether or not we really could see a shift in asset allocations for defined benefit
pension plans.
3. Who has exposure? We updated some of our forecasts and forward-looking
pension analysis to reflect asset returns and interest rates through the end of 2004,
to help us answer the simple question: Which companies and industries have the
most exposure to defined benefit pension plans?
4. The magic of pension accounting. With the FASB potentially overhauling pension
accounting, we thought it was time to update our analysis on what would happen to
the earnings and balance sheets of the companies in the S&P 500 if we eliminate
the magic of pension accounting and instead simply reflect the economics of the
pension plans.
5. The assumptions. With the SEC clearly focused on company pension assumptions,
we dig deep into the various pension assumptions, including an attempt to
determine whether or not they make sense.
Over the next few pages, we guide you through the report, highlighting some of the new
bells and whistles that we have added since Part II.
Pension Reform
The first step down the path to pension reform in 2005 was taken on January 10 by
Secretary of Labor Elaine Chao in a speech, “Protecting the Retirement Security of
America’s Workers: The President’s Plan for Reforming Defined Benefit Pension Plans.”
The president’s plan is based on three main principles: reform the funding rules, reform
the premiums, and enhance disclosures.
The president’s plan is just that, a plan; actually to date, it’s more like a few bullet points.
Pension reform has been discussed for years and plans to reform the defined benefit
pension system have come and gone. So what’s different this time around? A spark: to
move pension reform past the planning stage, Washington needs a spark. That spark
could come from the Pension Benefit Guaranty Corporation (PBGC) taking over US
Airways’ remaining pension plans and possibly United Airlines’ pension plans.
We expect the Administration to come out with much more detail on its proposal for
pension reform shortly. We would recommend taking a look at the Budget proposal on
February 7 and keeping a close watch on the Department of Labor’s Web site,
www.dol.gov, for more details. Also on the DOL’s Web site, you can find what the
Administration has proposed so far.
4
The Magic of Pension Accounting, Part III 07 February 2005
As we get our hands on proposals from Congress and more detail from the
Administration, we plan on following up with an in-depth analysis of pension reform in a
future report. In the meantime, we run you through some of our initial thoughts, including
a quick review of the Administration’s proposal and potential timing in the Pension
Reform section of this report.
Asset Allocations
Information on asset allocations had been sorely missing from the pension footnote. To
get a clue as to pension plan asset allocations, investors were forced to rely on macro
data from the Federal Reserve or various surveys. In the 2003 10-Ks, for the first time,
companies were required to disclose information about their pension plan asset
allocations. We gathered this data for each company in the S&P 500 with a defined
benefit pension plan.
Pension plan assets for the companies in the S&P 500 totaled $1,142 billion at the end
of 2003. In other words, the S&P 500 companies pension plans comprised about 61% of
the total $1,872 billion of assets in U.S. private sector defined benefit pension plans.
Our results show a 62% allocation to equities, 29% to fixed income, 3% to real estate,
and 6% to other asset classes for 2003.
We analyze the asset allocations in defined benefit pension plans in a variety of different
ways, including a look at the macro data from the Federal Reserve, the asset allocations
of the 10 companies with the largest pension plans, companies with the highest
allocations to various asset classes, shifts in asset allocation, the various approaches to
asset allocation, and the amount of plan sponsor’s stock in the pension plan.
Remember, the sole purpose of pension plan assets is to pay the pension benefits
promised to the employees. In theory, asset allocation should be driven by the pension
obligation, choosing the assets that best allow the company to meet the obligation it has
to its employees. However, some companies may have lost sight of that simple point for
a number of reasons, including the magic of pension accounting, where a higher
allocation to historically “higher” returning asset classes results in higher earnings. In
fact, in the 1990s many pension plans were magically transformed into profit centers.
That all changed starting in 2000, and then throughout 2001 and 2002, as the pension
plans were clobbered by the double whammy of a declining stock market and record low
interest rates. As a result, some companies—that didn’t previously—actually started
considering their pension obligations when determining the asset allocation in the
pension plan.
With a renewed focus on pension plans over the past few years, investors are
wondering whether or not the asset allocations make sense: Does the company have
the right mix of assets to meet its pension obligation? Is there an asset-liability
mismatch, and if so, how much risk is the company taking on as a result? The key to
answering these questions involves comparing the asset allocation to each company’s
pension obligation. For example, a longer-term pension obligation—younger employees
and fewer retirees—should in theory call for a different type of asset allocation than a
shorter-term pension obligation with older employees and lots of retirees.
5
The Magic of Pension Accounting, Part III 07 February 2005
The only problem is that no information is provided about the duration of the pension
obligation. What’s an investor to do? We now have all this asset allocation information
and nothing to compare it to. What we need is a method for estimating whether a
company’s pension obligation appears to have a longer duration or a shorter duration.
After much soul searching, head scratching, and break dancing, we came up with one:
the pension benefits paid as a percentage of the pension obligation. Our thought
process here is quite simple: the higher the benefits paid as a percentage of the pension
obligation, the more retirees; the more retirees, the shorter the duration of the obligation,
and vice versa. Clearly not perfect; however, it may get us in the ballpark.
So has the president’s plan for pension reform sparked a big shift recently in asset
allocation within defined benefit pension plans out of equities and into fixed income?
Anything is possible, however, there is still a lot of uncertainty as to how the reforms will
turn out; remember, the president’s plan is still just a plan. Let’s not forget that a big shift
in asset allocation out of equities and into fixed income by companies with defined
benefit pension plans would also put a dent in earnings.
However, defined benefit pension plans face a number of potential events and key
trends that could, from a macro perspective, spark a shift in asset allocation by pension
plans. The potential events include a change in pension accounting, pension reform,
pension plans getting healthier, and more pension plans getting dropped on the PBGC.
The key trends are demographics and companies continuing to move away from
defined benefit plans. The question is whether or not we are approaching some sort of
tipping point where the stars have aligned so that these events and trends actually
converge at or around the same time.
We explore asset allocations in more detail, including the potential for a shift in asset
allocation in the Pension Plan Asset Allocation section of this report.
Remember, if the company is in the S&P 500, all you have to do is type in the ticker
symbol and the historical pension data will update automatically. There is no data entry
required unless you want to input your own assumptions about the future to generate a
forecast. For those interested in analyzing companies outside the S&P 500, give us a
call; we have another version of the model that we can make available to you. We
describe the model and its limitations, and demonstrate its flexibility, in the Magic of
Pension Accounting, Part II.
6
The Magic of Pension Accounting, Part III 07 February 2005
Now that we find ourselves in a new year, we thought it would make sense to update
our pension forecasts. Taking into account the phenomenal performance of the stock
market in the last quarter of 2004, we now assume a 9.38% return on plan assets in
2004. In addition, as interest rates continued to trail down last year, we take the
discount rate down 25 basis points for 2004. As a result, we now estimate that the
pension plans of the companies in the S&P 500 got a bit less healthy in 2004, going
from $172 billion underfunded at the end of 2003 to our revised estimate of $185 billion
underfunded at the end of 2004. Exhibit 28 summarizes the key assumptions and
aggregate findings for our base-case forecast. This is just one of countless scenarios
that can be run through our model.
To determine which companies have the most exposure to their defined benefit pension
plans, we first divide the companies in the S&P 500 into two camps: those that have
defined benefit plans and those that don’t. We estimate that approximately 370
companies in the S&P 500 (74%) have defined benefit pension plans. In other words,
the 130 companies (26%) that don’t have such plans face no pension issues at all; they
are listed in Appendix A.
We focus on four types of exposure: shareholder, cash flow, earnings, and balance
sheet. In our new and improved pension report card, we bring together our analysis and
grade each industry group in the S&P 500 based on how much exposure they appear to
have to defined benefit pension plans. No surprises at the top of the list—Auto
Components, Autos, and Airlines appear to have the most exposure.
We also take you through a sensitivity analysis that has been talked about for quite
some time and is now getting more attention with the SEC focused on pension
assumptions: What would happen to earnings if the expected rate of return assumption
were to drop? Specifically: What would happen to pension cost in 2005 if all the
companies in the S&P 500 were to ratchet down their expected return assumptions? For
example, if we drop the expected rate of return by 100 basis points, we estimate
aggregate pension cost for the S&P 500 would increase by $12 billion, from $31 billion
to $43 billion, reducing aggregate S&P 500 earnings by nearly 2%.
At the company level, we found 52 companies where we estimate pension cost would
increase by at least $0.10 per share, including the 17 companies in Exhibit 45 with an
increase of at least $0.20 from a 100-basis-point decline in the expected rate of return.
We also found 29 companies in Exhibit 46 where a 100-basis-point drop in the expected
rate of return could reduce 2005 earnings by more than 5%.
We explore the exposure to defined benefit pension plans in more detail in the Who Has
Exposure? and Report Card sections of this report.
In December 2003, the FASB took a small step toward fixing pension accounting by
providing investors with some new disclosures. In our view, it’s just a matter of time
before the FASB takes a big step, adding a major project on pension accounting to its
agenda. In fact, we would not be surprised to see it on the agenda before the end of this
year, especially now that employee stock options have been taken care of.
As the board may soon start work on pension accounting, we thought it was time to
update our analysis on what would happen to the earnings and balance sheets of the
companies in the S&P 500 if we eliminate the magic of pension accounting and instead
simply reflect the economics of the pension plans in the financial statements.
7
The Magic of Pension Accounting, Part III 07 February 2005
From an analytical perspective, and if we could rewrite the accounting rules, we would
treat the pension plan as an investment subsidiary by first reporting the pension plan
assets as an asset on the balance sheet and the pension obligation as a liability. We
would then report the changes in plan assets (actual returns) and the pension obligation
(service cost, interest cost, and actuarial gains and losses) in the appropriate place on
the income statement.
We estimate the off-balance-sheet pension liability for the S&P 500 was $1,137 billion
(yup, that’s over $1 trillion) at the end of 2003 and $964 billion for 2002. The off-
balance-sheet pension asset for the S&P 500 was $887 billion and $768 billion for 2003
and 2002, respectively. In the aggregate, placing the pension plan assets on the
balance sheet would increase total assets for the S&P 500 by 5% and 4%, respectively,
for 2003 and 2002; treating the PBO as debt would increase aggregate debt for the S&P
500 by 18% and 17%. The difference between the increase in assets and the increase
in liabilities would reduce total equity for the S&P 500 by $163 billion, or 5%, at the end
of 2003. We expect this type of analysis to get some more attention when the SEC
releases its report on off-balance-sheet activity, which will feature pensions among other
topics.
If we were to reflect what really happened to the pension plans through earnings,
aggregate earnings for the companies in the S&P 500 would have increased by
approximately 3% in 2003. Contrast that with the prior four years, when earnings would
have dropped 67% in 2002, 67% in 2001, and 9% in 2000, while increasing 25% in
1999, as displayed in Exhibit 60.
One of the main reasons pension cost reported in the income statement can vary so
dramatically from what actually happened to the pension plan is the difference between
the actual return on plan assets and the expected return. Cumulatively from 1999
through 2003, the actual return on pension plan assets for the companies in the S&P
500 was $223 billion versus an expected return of $485 billion; that’s an unexpected
loss of $262 billion. The pension plans of only 15 companies outperformed their
expected returns from 1999 through 2003; on the other hand, the actual return on
pension plan assets fell short of the expected return for 349 companies in the S&P 500.
Focusing on the expected return on plan assets, we find that it has been right around
10% of aggregate operating income for the S&P 500 over the past five years, not the
highest quality of earnings. For specific companies, the expected return appears even
more significant. It accounted for over 50% of operating income in 2003 for the 30
companies in Exhibit 76.
We eliminate the pension cost reported in operating income (as defined by Compustat)
for each of the companies in the S&P 500 and replace it with service cost. We then
compare the reported operating income and margins with our adjusted operating
income. Operating income for the S&P 500 would remain unchanged in 2003 and
decline by 3% in 2002 and 4% in 2001 after making our adjustments. For 37 companies,
operating income would decline by 5% or more in 2003. The 20 companies in Exhibit 79
would experience a decline of over 10%; 14 of these companies would experience a
decline of over 10% in each of the past five years. At the other end of the spectrum,
operating income would increase by 5% or more for 48 companies in 2003. The 19
companies in Exhibit 80 would experience an increase of over 10%.
8
The Magic of Pension Accounting, Part III 07 February 2005
The Assumptions
With a heightened level of investor interest in this topic as a result of the SEC inquiry
into pension accounting, we decided to take another close look at the pension
assumptions for the companies in the S&P 500.
In 2003, we saw the median pension assumptions of companies in the S&P 500 move
to levels not seen in the last 10 years. For instance, the median expected rate of return
dropped 81 basis points from its peak of 9.31% in 2000 to 8.5% in 2003, its lowest level
of the decade. Meanwhile, the median discount rate hit 6.1%, also a 10-year low.
Finally, the median salary inflation rate, which hovered around 5% in the early 1990s,
fell to 4% in 2003.
We dig deeper to try and answer a simple question: Do the assumptions make sense
(specifically the expected return and discount rate)? In other words, we are looking for
support of the pension assumptions, which appears to be the very same thing that the
SEC is looking for (and apparently in some cases is having difficulty finding). The
inability to support the assumptions should cause you to pause; if a company can’t
support its pension assumptions, then you start to wonder which of the myriad other
accounting assumptions it can’t support.
We analyze the pension assumptions in more detail in the Analysis of Assumptions and
Do The Assumptions Make Sense? sections of this report.
Funded Status
In this section of the report, we take a quick look at the historical data on pension plan
funded status.
Quality of Earnings
Last, we analyze the historical data to focus on quality of earnings issues. For example:
What percentage of net income is from the pension plan, and how much of the growth in
earnings or decline in earnings is due to the pension plan?
Conclusion
It’s no surprise that once again we find the Auto Component, Auto, and Airline industry
groups have the most exposure to their defined benefit pension plans.
We may be approaching a tipping point for defined benefit pension plans. Potential
changes in pension accounting, pension reforms, and SEC scrutiny could eventually
lead to changes in behavior among companies and investors. Without all the
distractions created by misleading accounting and convoluted funding and premium
rules, companies should pay closer attention to the economics of their pension plans.
That could result in shifts in asset allocation, more transparent negotiations with the
workers, a restructuring of the plans (including hybrid plans), or companies dropping the
plans altogether. For investors, if they are no longer wasting their time on the complex
smoothing mechanisms and how they affect earnings, they can instead pay attention to
what’s really important, the funded status of the pension plan, future funding needs (the
claim on cash flows), the assumptions, and a comparison between the duration of the
pension obligation and the asset allocation to measure the amount of risk in the plan.
9
The Magic of Pension Accounting, Part III 07 February 2005
• Pension plan assets for the companies in the S&P 500 totaled $1,142 billion at the
end of 2003. In other words, the S&P 500 companies pension plans comprised about
61% of the total $1,872 billion of assets in U.S. private sector defined benefit pension
plans. Our results show a 62% allocation to equities, 29% to fixed income, 3% to real
estate, and 6% to other asset classes for 2003.
• In Exhibit 13, we provide the asset allocations for the 10 companies with the largest
pension plans in the S&P 500; they account for 38% of the total S&P 500 pension
plan assets.
• Focusing on the outliers among the companies in the S&P 500 in Exhibit 14, we
highlight the five companies that have the highest allocations to each of the four major
asset classes.
• We found 94 companies at the end of 2003 that talked about holding their own stock
in the pension plans they sponsor, including the 8 companies in Exhibit 20 where their
own stock represents more than 10% of the plan assets.
• We came up with a simple method for estimating the relative duration of the pension
obligation; the pension benefits paid as a percentage of the pension obligation. For
the companies in the S&P 500, the median ratio of benefits paid to the pension
obligation was 6%, with 212 companies clustered between 4.0% and 7.9%.
• We compared our estimate of the relative duration of the pension obligations to the
pension plan asset allocations. In Exhibit 24, we do not find any relationship between
the allocation to fixed-income assets and the duration of the pension obligation. That
leaves us with one of three takeaways: either our method for ranking the relative
duration of the pension obligations does not work, the asset allocations don’t make
any sense, or we are lacking the right information to run this analysis. We think it may
be a little bit of each.
• Defined benefit pension plans face a number of potential events and key trends that
could, from a macro perspective, spark a shift in asset allocation by pension plans.
The potential events include a change in pension accounting, pension reform, pension
plans getting healthier, and more pension plans getting dropped on the PBGC. The
key trends are demographics and companies continuing to move away from defined
benefit plans. The question is whether or not we are approaching some sort of tipping
point where the stars have aligned so that these events and trends actually converge
at or around the same time.
• To begin answering the question of which companies have the most exposure to their
defined benefit pension plans, we first divide the companies in the S&P 500 into two
camps: those that have defined benefit plans and those that don’t. We estimate that
approximately 370 companies (74%) have defined benefit pension plans in the S&P
10
The Magic of Pension Accounting, Part III 07 February 2005
500. In other words, the 130 companies (26%) that don’t have such plans face no
pension issues at all; they are listed in Appendix A.
• We estimate that the pension plans of the companies in the S&P 500 got a bit less
healthy in 2004, going from $172 billion underfunded at the end of 2003 to our revised
estimate of $185 billion underfunded at the end of 2004. Exhibit 28 summarizes the
key assumptions and aggregate findings for our base-case forecast. In Exhibit 30 and
Exhibit 31, we show you how we arrive at a weaker pension plan in 2004.
• We compare the funded status of the pension plan with each company’s equity
market capitalization to get an idea as to which pension plans have the largest claim
on the shareholders’ stake in the company. The distribution of this relationship for the
companies in the S&P 500 is included in Exhibit 32. We estimate that the eight
companies in Exhibit 33 had defined benefit pension plans that were underfunded by
more than 25% of their equity market capitalization as of December 31, 2004.
• Next, we evaluate which companies have the largest potential exposure, the highest
leverage to defined benefit pension plans, by comparing the pension obligation to the
equity market capitalization. We estimate that the 20 companies in Exhibit 34 had
projected benefit obligations that exceeded their equity market capitalization as of
December 31, 2004.
• S&P 500 companies contributed $74 billion into their pension plans in 2003, on top of
the $46 billion they contributed in 2002 and up from only $14 billion in 2001. In their
2003 10-Ks, the companies in the S&P 500 disclosed they would contribute $24 billion
to their pension plans during 2004. We estimate the companies in the S&P 500 will
contribute an aggregate $14 billion to their pension plans in 2005.
• We estimate that 232 companies will experience an increase in pension cost between
2004 and 2005, causing an earnings headwind that further reduces earnings in 2005.
We estimate that the 18 companies in Exhibit 38 will experience an increase in
pension cost between 2004 and 2005 of at least $0.10 per share.
• We provide you with a high-level sensitivity analysis, which we generated using our
model, to give an indication of how sensitive the aggregate funded status, pension
cost, and pension contributions are to changes in the actual return on plan assets and
the discount rate. Exhibit 41 provides the sensitivity of the aggregate pension plan’s
funded status, Exhibit 42 provides the sensitivity of total estimated pension cost, and
Exhibit 43 provides the sensitivity of total estimated pension contributions in 2005.
• In Exhibit 44, we display what would happen to aggregate pension cost for the
companies in the S&P 500 if we were to ratchet down the 2005 expected return
11
The Magic of Pension Accounting, Part III 07 February 2005
• For 52 companies we estimate pension cost would increase by at least $0.10 per
share, including the 17 companies in Exhibit 45 with an increase of at least $0.20 per
share from a 100-basis-point decline in the expected rate of return. In other words, a
100-basis-point drop in the expected rate of return would reduce 2005 earnings by
more than 5% for each of the 29 companies in Exhibit 46.
• As the board may soon start work on pension accounting, we thought it was time to
update our analysis on what would happen to the earnings and balance sheets of the
companies in the S&P 500 if we eliminated the magic of pension accounting and
instead simply reflected the economics of the pension plans in the financial
statements. We estimate the off-balance-sheet pension liability for the S&P 500 was
$1,137 billion at the end of 2003 and $964 billion for 2002. The off-balance-sheet
pension asset for the S&P 500 was $887 billion and $768 billion for 2003 and 2002,
respectively.
• The 10 companies with the largest off-balance-sheet pension liabilities are included in
Exhibit 51, accounting for over 36% of the total for the S&P 500. It should be no
surprise that these same 10 companies also have the largest off-balance-sheet
pension assets, accounting for 34% of the total for the S&P 500.
• In the aggregate, placing the pension plan assets on the balance sheet would
increase total assets for the S&P 500 by 5% and 4%, respectively, for 2003 and 2002;
treating the PBO as debt would increase aggregate debt for the S&P 500 by 18% and
17%. The difference between the increase in assets and the increase in liabilities
would reduce total equity for the S&P 500 by $163 billion, or 5%, at the end of 2003.
Some specific impacts include:
• If the PBO is considered the equivalent of debt, debt would more than triple for 43
companies, including the 24 companies in Exhibit 52 that had over $1 billion in debt
outstanding. The pension obligation is very similar to debt, except that the creditors
are the employees. In effect, a company is borrowing from its employees, which
may not be the most efficient way to raise capital.
• If the pension plan’s funded status were reported on the balance sheet,
shareholders’ equity would decline by 25% or more for the 19 companies in Exhibit
54. In other words, the shareholders’ equity for these companies may be overstated
by more than 25% due to pension accounting.
• To calculate “real” pension cost, we replace the reported pension cost from the
income statement with what actually happened: the actual changes in the pension
plan assets and PBO. Applying this methodology to the companies in the S&P 500
would have increased aggregate earnings by approximately 3% in 2003. Contrast that
with the prior four years, when earnings would have dropped 67% in 2002, 67% in
2001, and 9% in 2000, while increasing 25% in 1999, as displayed in Exhibit 60.
• Cumulatively from 1999 through 2003, the actual return on pension plan assets for the
companies in the S&P 500 was $223 billion versus an expected return of $485 billion,
12
The Magic of Pension Accounting, Part III 07 February 2005
• In Exhibit 74, we break down aggregate net pension cost for the companies in the
S&P 500 into its various components for each of the past five years. We then take
each component and compare it to aggregate operating income, as that is where net
pension cost is currently reported. For example, the expected return on plan assets in
2003 was $97 billion, or 10% of total operating income for the companies in the S&P
500, clearly not the highest-quality earnings.
• For specific companies, the expected return appears even more significant. It
accounted for over 25% of operating income for 65 companies, including the 30
companies in Exhibit 76 where the expected return on plan assets was more than
50% of the companies’ operating income in 2003.
• If we strip out all the components of pension cost and leave behind service cost,
operating income for the S&P 500 would remain flat in 2003, while declining by 3% in
2002, and 4% in 2001. For the 20 companies in Exhibit 79 operating income would
decline by more than 10%; 14 of these companies would experience a decline of over
10% in each of the past five years. At the other end of the spectrum, operating income
would increase by 10% or more for the 19 companies in Exhibit 80.
• Exhibit 91 shows the distribution of expected rates of return across the companies in
the S&P 500 during 2003. Even as companies have ratcheted down their expected
return assumptions over the past few years, 273, or almost 75% of the companies
with defined benefit plans in the S&P 500, still use expected rates of return of 8% or
higher. The median expected rate of return was 8.50% in 2003, down from 8.80% in
2002 and 9.24% in 2001.
• Early in the fourth quarter of 2004, pension plan assumptions were once again on the
hot seat—vaulting defined benefit pension plans back into the spotlight—as six
companies—Boeing, Delphi, Ford, General Motors, Navistar, and Northwest
Airlines—announced that the SEC was taking a look at the accounting for their
pension and OPEB plans. The key to whether or not any of these companies will
have a problem with the SEC will depend upon how they support their pension
assumptions.
• We found that 270 of the 370 companies with pension plans actually employed an
expected rate of return that falls within 100 basis points of our estimate based on
historical returns, while 170 companies employed an expected rate of return that falls
within 50 basis points of our estimates. Not a surprise, it appears as if historical
returns are driving the expected return assumption.
• In Exhibit 127, we plot out the Moody’s Aa rate at the end of each month during 2003.
In addition, we gathered the pension plan year-ends (measurement dates) for each
company in the S&P 500 with a defined benefit pension plan. We then grouped the
companies by the month in which their pension plan years end. For each month, we
noted the number of companies with a pension plan year-end that falls during that
month and plotted the median discount rate for that group of companies.
• In general, the monthly median discount rate appeared to track the Moody’s Aa
corporate bond index spot rate at the end of each month reasonably closely
13
The Magic of Pension Accounting, Part III 07 February 2005
throughout 2003. However, it is interesting to note that the median discount rate was
higher than the Moody’s rate for eight out of twelve months during the year.
Remember, a higher discount rate results in a smaller pension obligation and lower
pension cost.
• In Exhibit 129, we plot out the ratio of benefits paid to the pension obligation versus
the discount rates used by each company in the S&P 500. In theory, as we move from
a longer-duration pension obligation toward a shorter-duration pension obligation, we
would expect to see the discount rates drop. Instead, we find that no matter the
estimated duration of the pension obligation, companies appear to be using around
the same discount rate.
• Defined benefit pension plans for the companies in the S&P 500 were the picture of
good health in 1999, when they were $248 billion overfunded (128% funded) in total.
At the end of 2003, we see a dramatically different picture, as the plans ended the
year $172 billion underfunded (87% funded). In other words, the health of the pension
plans had deteriorated by $420 billion over that five-year period. What caused the
deterioration? Simple, the pension plan assets could not keep pace with the pension
obligations. Note that in Exhibit 133, from 1999 to 2003 the pension plan assets grew
by $10 billion, a compound annual growth rate of less than 1%, while the pension
obligations grew by $430 billion, a compound annual growth rate of roughly 10%.
• In Exhibit 134 and Exhibit 135, we find that the pension plan assets and obligations
are pretty well concentrated, with 13 industry groups accounting for over 70% of the
total pension plan assets and obligations for the companies in the S&P 500. Note that
three industry groups—Automobiles, Aerospace and Defense, and Diversified
Telecom—make up over 30% of the total.
• Exhibit 138 contains a distribution of the funded status of the pension plans of the
companies in the S&P 500. For example, 164, or 44% of the companies that offer
defined benefit plans, have pension plans that are 70-89% funded.
• Even as pension plans got healthier in 2003, the powerful smoothing mechanisms
built into pension accounting worked their magic, driving pension costs up and
earnings down. Between 2002 and 2003, net income from continuing operations for
the S&P 500 increased by $202 billion. In the aggregate, pretax pension costs
increased by $25 billion ($17 billion after tax), from $3 billion in 2002 to $28 billion in
2003, slowing the growth in earnings for most companies with defined benefit pension
plans. Another example of the disconnect between pension accounting and the
economics of the pension plan: as the plans grew stronger, pension costs increased.
• To measure whether or not the change in pension cost was meaningful to a particular
company’s earnings, we first compare the earnings headwind from the pension plan in
2003 to 2002 reported earnings. We found that the earnings headwind created a drag
of at least 5% for 89 companies and at least 10% for 50 companies, including the 20
companies in Exhibit 152 that suffered an earnings drag of at least 20%.
• We base our analysis on the companies in the S&P 500 index, reflecting changes to
the index through January 1, 2004.
14
The Magic of Pension Accounting, Part III 07 February 2005
Pension Plan Stats 1999 2000 2001 2002 2003 2004E 2005E 2006E
Number of Companies Overfunded 254 238 122 35 51 40 52 66
Number of Companies Underfunded 88 110 237 335 319 326 314 300
Plan Assets (US$ in billions) $1,132 $1,177 $1,050 $955 $1,142 $1,192 $1,222 $1,252
Projected Benefit Obligation (PBO) (US$ in billions) $884 $954 $1,040 $1,171 $1,314 $1,377 $1,362 $1,347
Funded Status $ (US$ in billions) (Under) Over $248 $222 $10 ($216) ($172) ($185) ($141) ($96)
Funded Status % 128% 123% 101% 82% 87% 87% 90% 93%
Contributions (US $ in billions) NA $16 $14 $46 $74 $27 $14 $15
Benefits Paid (US$ in billions) NA $68 $75 $82 $84 $87 $89 $92
Number of Companies—Pretax Pension Income 100 148 149 94 45 38 30 27
Number of Companies—Pretax Pension Expense 235 193 203 266 320 330 338 341
Pre-Tax Pension Cost—(Income)/Expense (US$ in billions) $(2) $(12) $(7) $3 $28 $26 $31 $31
After-Tax Pension (Income)/Expense—% of Net Income 0% (2%) (2%) 1% 4% 3% 3% 3%
Median Expected Rate of Return on Plan Assets 9.25% 9.31% 9.24% 8.80% 8.50%
Actual Return 16.94% 5.13% (7.38%) (8.85%) 18.80%
Median Discount Rate 7.50% 7.50% 7.25% 6.75% 6.10%
Median Salary Inflation Rate 4.53% 4.50% 4.50% 4.15% 4.00%
Source: Company data, CSFB estimates.
15
The Magic of Pension Accounting, Part III 07 February 2005
Pension Reform
The first step down the path to pension reform in 2005 was taken on January 10 by
Secretary of Labor Elaine Chao in a speech, “Protecting the Retirement Security of
America’s Workers: The President’s Plan for Reforming Defined Benefit Pension Plans.”
The idea behind pension reform is quite simple: if the system is broke then fix it. The
president’s plan is based on three main principles: reform the funding rules, reform the
premiums, and enhance disclosures for defined benefit pension plans offered by U.S.
corporations (the proposed reforms would not affect public (state, local, etc.) pension
plans).
The president’s plan is just that, a plan; actually to date, it’s more like a few bullet points.
Pension reform has been discussed for years and plans to reform the defined benefit
pension system have come and gone. Remember, the Administration came out with a
similar proposal to improve the accuracy and transparency of pension information back
in July 2003. Various bills were proposed, and in the end, the U.S. Congress passed the
Pension Funding Equity Act in April 2004. Instead of reforming the defined benefit
pension system, however, the Act provided companies with pension-funding relief,
allowing them to contribute less to their pension plans in 2004 and 2005. In addition, the
U.S. Congress actually provided “extra special” relief to the Airlines and Steel
companies, allowing the companies with the weakest pension plans to contribute even
less. (For more on pension funding relief, please see our April 8, 2004, report, Pension
Funding Relief Is Here.)
16
The Magic of Pension Accounting, Part III 07 February 2005
lobbying effort, pension reform is likely to result in less tax dollars for the U.S.
government (as companies make additional tax-deductible contributions to their pension
plans), not ideal when running large budget deficits, and Washington seems more
focused on another retirement issue: Social Security reform.
Some have even speculated that pension reform could cause a shift in asset allocation
within corporate defined benefit pension plans in the U.S.—from equities toward fixed
income. We explore this thought process a bit more, in the Pension Plan Asset
Allocation section of this report.
We expect the Administration to come out with much more detail on its proposal for
pension reform shortly. We would recommend taking a look at the Budget proposal on
February 7 and keeping a close watch on the Department of Labor’s Web site,
www.dol.gov, for more details. Also on the DOL’s Web site, you can find what the
Administration has proposed so far.
As we get our hands on proposals from Congress and more detail from the
Administration, we plan on following up with an in-depth analysis of pension reform I
n a future report. In the meantime, we wanted to run you through some of our initial
thoughts, starting out with a quick review of the Administration’s proposal and potential
timing.
The basic idea behind reforming the funding rules is to get companies to fully fund their
pension plans without all the twists and turns of the current rules. The Administration’s
plan includes the following ideas:
• Replace multiple measures of the pension obligation with one measure, adjusted to
reflect the risk that the pension plan is terminated. We would also expect the reforms
to replace the multiple measures of the pension plan assets with the fair value of the
plan assets.
17
The Magic of Pension Accounting, Part III 07 February 2005
• Base the pension plan funding requirements on the company’s financial health. A
weaker company with a weaker pension plan will be required to make larger
contributions to its pension plan.
• Stop financially strapped companies from making pension promises they can’t afford.
If the yield-curve approach were used, how would a company know which point to pick
on the yield curve when setting its discount rate? That would depend upon when the
future pension benefits are to be paid. Using a very simple example, if 50% of the
pension obligation is due in 5 years and 50% is due in 15 years, the discount rate would
be a weighted average based half on the current yield on 5-year corporate bonds and
half on the current yield on 15-year corporate bonds. The July 2003 proposal from the
Administration would have transitioned from the 4-year, weighted-average yield on long-
term corporate bonds to the yield-curve approach by 2008.
According to the 2003 proposal, “Using the yield curve is essential to match the timing
of future benefit payments with the resources necessary to make the payments.” We
agree it is the economically correct way to price the obligation. The only problem with
this approach is that it could hurt the companies that the funding relief helped the
most—those with the largest pension problems, the old-line companies with lots of
retirees. A large portion of retirees and older workers in the pension plan results in a
shorter-term obligation. Therefore, under a yield-curve approach, the shorter end of the
yield curve would be used to set the discount rate. A normal, upward-sloping yield curve
would result in a lower discount rate, a larger pension obligation, and increased funding
requirements. That’s the reality for these companies: They have an obligation that is
18
The Magic of Pension Accounting, Part III 07 February 2005
coming due in the near future. If they don’t step up and fund it, you and I will be bearing
that cost when the government has to step in and bail out the PBGC.
The PBGC is also an insurance company that has no control over the premiums that it
charges, a recipe for failure. The premiums are set by the U.S. Congress and have not
changed since 1991. Companies currently pay a flat-rate premium of $19 per employee
or retiree plus a variable-rate premium of $9 for each $1,000 the pension plan is
underfunded (using a vested benefit obligation). The PBGC’s premium revenue in 2004
was $1.5 billion; meanwhile, it paid out $3 billion in pension benefits last year.
Reforming the premiums that companies pay to the PBGC revolves around moving to
more of a risk-based premium structure. To reform the pension insurance premiums, the
Administration has proposed the following:
Flat-Rate Premiums
• Increase the flat-rate premium to $30 per employee to reflect the growth in worker
wages since 1991.
• Index the flat-rate premium in the future to the growth in worker wages.
• Risk-Based Premiums
• Allow the PBGC’s board (Secretaries of Labor, Commerce, and Treasury) to adjust
the risk-based premium so that premium revenue is sufficient to cover expected
losses and to improve the PBGC’s financial condition.
Notice there is no mention of tying the risk-based premiums to the pension plan’s asset
allocation, where a pension plan with what appeared to be a riskier asset allocation (i.e.,
heavy equity allocation) would pay a higher premium than a pension plan with a more
conservative allocation (i.e., high proportion of fixed-income assets). Instead, the
Administration appears content with measuring risk based on the funded status of the
pension plan and the financial health of the company. Apparently, the Administration
does not want to be seen as directly influencing the asset allocation decisions made by
pension plans.
Enhance Disclosure
The idea here is simple, provide “. . . workers, investors, regulators and the public with
more and better information about the financial health of defined benefit pension plans,”
including:
19
The Magic of Pension Accounting, Part III 07 February 2005
• Disclose the actual financial health of the pension plan to workers, including recent
trends in funded status.
• Make available to the public more information that is filed with the PBGC on
underfunded plans. This refers to the information from the 4010 filings that certain
companies, including those with a pension plan that is more than $50 million
underfunded, must file with the PBGC. Such a filing contains all types of information
about the company and the pension plan, including the market value of pension plan
assets and the pension obligation calculated on a termination basis. About 400
companies file 4010 reports with the PBGC; these companies account for nearly 80%
of the total estimated pension underfunding in the U.S. Currently, the PBGC is
prohibited from disclosing the information in the 4010 filings to the public.
Other Reforms
A variety of other pension reforms will likely be discussed over the course of this year,
including enhancing the PBGC’s status in bankruptcy and allowing for transition to
hybrid plans (cash balance).
Timing
Now that we have taken you through the Administration’s proposed reforms, we wanted
to give you an idea as to potential timing. For one, we expect the Administration to come
out with much more detail on its proposal for pension reform shortly, maybe even within
the next week or so. We recommend taking a look at the Budget proposal on February 7
and keeping a close watch on the Department of Labor’s Web site, www.dol.gov, for
more details. Also on the DOL’s Web site, you can find what the administration has
proposed so far.
Once we see more detail from the Administration, bills from Congress and
Congressional hearings should eventually follow (we would expect things to heat up in
the Spring). Four committees in Congress would get involved with pension reform: two
in the Senate—the Senate Finance Committee and the Senate Committee on Health,
Education, Labor & Pension—and two in the House—the House Ways and Means
Committee and the House Education and the Workforce Committee. The chairmen for
each committee are listed in Exhibit 2.
20
The Magic of Pension Accounting, Part III 07 February 2005
Source: CSFB.
For an idea as to what direction the congressional proposals might take, you may want
to take a look at what the chairmen of these committees have proposed in the past. For
example, Congressman John Boehner has been quite vocal on pension reform. That’s
not a surprise when you think about the state he represents; probably lots of people in
Ohio with defined benefit pension plans. Back in September 2004, he laid out the
following “six principles to reform and strengthen the defined benefit pension system.”
Many of which are in harmony with the Administration’s proposal.
• Reduce funding volatility in pension plans to ensure that companies make adequate
and consistent contributions to their plans.
• Companies and unions should not make promises to workers they can’t keep.
• Provide more accurate and meaningful disclosure to workers about the status of their
pension plan.
• Ensure that hybrid plans (e.g., cash balance plans) are a viable part of the defined
benefit pension system.
Or take a look at Senator Chuck Grassley’s 2003 proposal, the National Employee
Savings and Trust Equity Guarantee (NESTEG) Act, which included the yield-curve
approach.
Whether we can arrive at pension reform by the end of 2005 before pension funding
relief expires will partially depend upon the status of Social Security reform, as Social
Security appears to be the higher priority for the White House. It will be difficult to get
both done this year. Therefore, if Social Security reform really gets traction in
Washington, pension reform could get placed on the back burner; on the other hand, if
Social Security reform gets bogged down, pension reform could move to the front of the
line.
Even if pension reform is at the front of the line in Washington, whether it gets done and
in what form is dependent upon the strength of the lobbying effort against it. If
companies and unions team up against the reform, pension reform could get squashed
or if it gets through drastically watered down. On the other hand companies may not put
up as much a fight against the reforms as expected since they may want to use pension
reform as an excuse to stop offering a defined benefit pension plan.
One other thing to keep in mind, there is likely to be a multiple year transition period to
phase in the new reforms.
21
The Magic of Pension Accounting, Part III 07 February 2005
The FASB requires that for each major asset class in the pension portfolio—which most
companies have interpreted as equities, fixed income, real estate, and other—the
company provide the percentage of pension plan assets invested in each asset class
based on the fair value of the plan assets at the pension plan year-end. The FASB did
not define the major asset classes as equity, fixed income, real estate, and other;
however, the board did provide an example of the disclosure that included these four
asset classes. As is often the case, the FASB examples quickly find their way into actual
practice. The thought process being that if companies follow an FASB example, how
can they go wrong.
22
The Magic of Pension Accounting, Part III 07 February 2005
2003 2002
December 31 Target Actual Actual
allocation allocation allocation
Equity securities 51-63 % 60 % 56 %
Plan fiduciaries set investment policies and strategies for the GE Pension Trust. Long-term strategic investment objectives include
preserving the funded status of the trust and balancing risk and return. The plan fiduciaries oversee the investment allocation process,
which includes selecting investment managers, commissioning periodic asset-liability studies, setting long-term strategic targets and
monitoring asset allocations. Target allocation ranges are guidelines, not limitations, and occasionally plan fiduciaries will approve
allocations above or below a target range.
Trust assets are invested subject to the following policy restrictions: short-term securities must be rated A1/P1 or better; investments in
real estate–7% of trust assets at year end–may not exceed 25%; other investments in securities that are not freely tradable–11% of trust
assets at year end–may not exceed 20%. GE common stock represented 6.3% and 6.0% of trust assets at year-end 2003 and 2002,
respectively, and is subject to a statutory limit when it reaches 10% of total trust assets.
Source: GE’s 10-K filing for the year ended December 31, 2003.
23
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 4: U.S. Private Sector—Total Defined Benefit Pension Exhibit 5: U.S. Private Sector—Defined Benefit Pension Plan
1
Plan Assets, 1985–2003 Asset Allocation, 2003
US$ in billions US$ in billions, unless otherwise stated
$1,500
$1,000
$795
Equities Other
$500 $1,033 $353
55% 19%
$0
5
3
8
0
19
19
19
19
19
19
19
19
20
20
1
Source: Federal Reserve Statistical Release from the Board of Governors Equities include mutual fund shares and corporate equities; fixed
of the Federal Reserve System; Flow of Funds Accounts of the United income includes all credit market instruments; and other includes cash,
States—Flows and Outstandings, Fourth Quarter 2003, Table L.119.b cash equivalents, and miscellaneous investments.
Private Pension Funds: Defined Benefit Plans. Source: Federal Reserve Statistical Release from the Board of
Governors of the Federal Reserve System; Flow of Funds Accounts of
the United States—Flows and Outstandings, Fourth Quarter 2003,
Table L.119.b Private Pension Funds: Defined Benefit Plans.
In Exhibit 6, we take a look at the pension plan asset allocations over time and find that
the allocation to fixed income and equities converged at 39% back in 1990. In contrast,
asset allocations to fixed income and equities reached their greatest point of divergence
in 1999—61% of defined benefit pension plan assets in the private sector were
allocated to equities and 23% were allocated to fixed income. With the roaring bull
market in equities through the 1990s, it appears as if many pension plans simply let it
ride.
24
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 6: U.S. Private Sector—Defined Benefit Pension Plan Asset Allocations over Time,
1
1985–2003
US$ in billions, unless otherwise stated
$1,200 60%
$1,000 50%
$800 40%
$600 30%
$400 20%
$200 10%
$0 0%
1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
1
Equities include mutual fund shares and corporate equities, fixed income includes all credit market
instruments, and other includes cash, cash equivalents, and miscellaneous investments.
Source: Federal Reserve Statistical Release from the Board of Governors of the Federal Reserve System;
Flow of Funds Accounts of the United States—Flows and Outstandings, Fourth Quarter 2003, Table L.119.b
Private Pension Funds: Defined Benefit Plans.
We gathered the pension plan asset allocation data from the 10-Ks for each company in
the S&P 500 with a defined benefit pension plan. The aggregate pension plan asset
allocation for the companies in the S&P 500 is not much of a surprise. It mirrors, for the
most part, the fed data that we just ran through and many of the asset allocation
surveys. For example, according to the Pensions & Investments survey of the top 1,000
defined benefit pension plans in the U.S., published on January 26, 2004, U.S. pension
plan portfolios held 61% in equities (45.2% domestic stocks, 15.8% international
stocks), 28% in fixed income, 3.1% in equity real estate, 3.9% in private equity, and
4.0% in cash and other asset classes. Our results show a 62% allocation to equities,
29% to fixed income, 3% to real estate, and 6% to other asset classes for 2003. For
those of you that like pie charts, we display the 2003 and 2002 pension plan asset
allocations for the S&P 500 companies side by side in Exhibit 7 and Exhibit 8.
25
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 7: S&P 500 Companies—Defined Benefit Pension Exhibit 8: S&P 500 Companies—Defined Benefit Pension
Plan Asset Allocation, 2002 Plan Asset Allocation, 2003
US$ in billions, unless otherwise stated US$ in billions, unless otherwise stated
Other
Other $72
$53 6%
6%
Equities Equities
$563 $708
59% 62%
Source: Company data, CSFB estimates. Source: Company data, CSFB estimates.
Exhibit 9: S&P 500 Pension Plan Asset Allocations— Exhibit 10: S&P 500 Pension Plan Asset Allocations—Fixed
Equities, 2003 Income, 2003
240 200
195
200 155
Number of Companies
Number of Companies
160
138
160
120
120
85 80
80
56
40 31
40 19
7 10 9 9 10
- -
Below 30% 30%-49.9% 50%-59.9% 60%-69.9% 70%-84.9% 85% or Above Below 10% 10%-19.9% 20%-29.9% 30%-39.9% 40%-49.9% 50% or Above
Source: Company data, CSFB estimates. Source: Company data, CSFB estimates.
26
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 11: S&P 500 Pension Plan Asset Allocations—Real Exhibit 12: S&P 500 Pension Plan Asset Allocations—Other
Estate, 2003 Investments, 2003
280 160
251 141
240
Number of Companies
Number of Companies
120
200
90 90
160 81
80
120
80
40
32 30 25
40 15 8 12
9
- -
0% 0.1%-2.49% 2.5%-4.9% 5%-7.49% 7.5%-9.9% 10% or Above 0% 0.1%-2.49% 2.5%-9.9% 10%-14.9% 15%-19.9% 20% or Above
Source: Company data, CSFB estimates. Source: Company data, CSFB estimates.
In Exhibit 13, we provide the asset allocations for the 10 companies with the largest
pension plans in the S&P 500; they account for 38% of the total S&P 500 pension plan
assets. Note how the allocations vary among this group of companies, from Ford with a
70% allocation to equities, to Verizon with a 24% allocation to Other.
1
Exhibit 13: Allocations of Top 10 Largest Pension Plans Based on Assets
US$ in millions, unless otherwise stated
2003 Reported Actual Asset Allocations
Total
Pension
Plan Assets Fixed Real
Company Ticker 2003 Equity Income Estate Other
General Motors Corp GM $ 93,729 49% 31% 8% 12%
Intl Business Machines Corp IBM 68,225 58% 33% 3% 7%
Ford Motor Co F 53,564 70% 29% 0% 1%
General Electric Co GE 43,879 60% 20% 7% 13%
Verizon Communications VZ 42,841 56% 17% 3% 24%
Boeing Co BA 33,209 55% 38% 3% 4%
SBC Communications Inc SBC 28,154 66% 27% 3% 4%
Lockheed Martin Corp LMT 20,913 63% 33% 0% 4%
AT&T Corp T 17,997 68% 23% 9% 0%
Du Pont (E I) De Nemours DD 17,967 63% 24% 4% 9%
1
Amounts represent combined asset allocations for U.S. and international plans.
Source: Company data, CSFB estimates.
Focusing on the outliers among the companies in the S&P 500 in Exhibit 14, we
highlight the five companies that have the highest allocations to each of the four major
asset classes, including Intel with a 100% allocation to equities, Merrill Lynch with an
84% allocation to fixed income, Weyerhaeuser with a 13% allocation to real estate, and
Biogen with an 88% allocation to other assets. In Biogen’s case, the other category
includes cash and cash equivalents.
27
The Magic of Pension Accounting, Part III 07 February 2005
1
Exhibit 14: Highest Asset Allocations among Different Asset Classes, 2003
Asset Allocations (%)
Company Ticker 2003A 2002A
Equity Securities:
Intel Corp INTC 100% 100%
Cincinnati Financial Corp CINF 97% 98%
U S Bancorp USB 97% 96%
General Dynamics Corp GD 96% 66%
Electronic Data Systems Corp EDS 87% 85%
Fixed Income Securities:
Merrill Lynch & Co MER 84% 89%
Starwood Hotels & Resorts Wrld HOT 79% 45%
Viacom Inc-Cl B VIA.B 65% 74%
AES Corp. (The) AES 64% 62%
Loews Corp LTR 63% 53%
Real Estate:
Weyerhaeuser Co WY 13% 17%
United Parcel Service Inc UPS 11% 15%
Southern Co SO 11% 12%
Delphi Corp DPH 11% 9%
Delta Air Lines Inc DAL 11% 12%
Other2:
Biogen Idec Inc BIIB 88% NA
Weyerhaeuser Co WY 82% 93%
Freeport McMoran Cop & Gld FCX 51% 44%
Centurytel Inc CTL 35% 28%
Amsouth Bancorporation ASO 25% 26%
1
Amounts represent combined asset allocations for U.S. and international plans.
2
Includes private equity, hedge funds, short-term investments, cash, and cash equivalents.
Source: Company data, CSFB estimates.
When looking at individual companies, we found that the majority also had relatively
minor changes in their asset allocations, with 201 companies’ asset allocations
changing by less than five percentage points. However, for some companies, the
swings in asset allocation were significant, including the 20 companies in Exhibit 15 that
had the biggest increase in asset allocation by asset class.
28
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 15: Biggest Increase in Asset Allocations among Different Asset Classes, 2003
Asset Allocations (%)
Company Ticker 2003A 2002A Increase
Equity:
General Dynamics Corp GD 96% 66% 30%
TJX Companies Inc TJX 62% 40% 22%
Hilton Hotels Corp HLT 50% 32% 18%
Gannett Co GCI 63% 45% 18%
Moodys Corp MCO 71% 53% 18%
Fixed Income:
Starwood Hotels & Resorts Wrld HOT 79% 45% 34%
Dover Corp DOV 37% 10% 27%
Hershey Foods Corp HSY 37% 15% 22%
CMS Energy Corp CMS 52% 32% 20%
AON Corp AOC 54% 38% 16%
Real Estate:
Cigna Corp CI 8% 0% 8%
Coors (Adolph) - Cl B RKY 7% 0% 7%
Air Products & Chemicals Inc APD 5% 0% 5%
Sprint Fon Group FON 5% 0% 5%
Nisource Inc NI 5% 0% 5%
Other1:
Biogen Idec Inc BIIB 88% 0% 88%
Deere & Co DE 20% 0% 20%
Federal Home Loan Mortg Corp FRE 19% 6% 13%
General Motors Corp GM 12% 1% 11%
Lilly (Eli) & Co LLY 11% 1% 10%
1
Includes private equity, hedge funds, short-term investments, cash, and cash equivalents.
Source: Company data, CSFB estimates.
General Motors provided the following rationale for shifting toward other types of assets,
including alternatives in its pension plan, in its 2003 10-K:
The current strategic mix for U.S. pension plans has reduced exposure to equity
market risks and increased allocation to asset classes which are not highly
correlated as well as asset classes where active management has historically
generated excess returns and places greater emphasis on manager skills to produce
excess return while employing various risk mitigation strategies to reduce volatility.
When fully implemented, GM pension assets will have the following allocations:
global equity: 41%-49%, global bonds: 32%-36%, real estate: 8%-12%, and
alternatives: 9%-13%. Overall, the current strategic policy mix is expected to result in
comparable but less volatile returns than GM's prior asset mix.
29
The Magic of Pension Accounting, Part III 07 February 2005
Also note how companies’ views on asset allocation can appear to vary dramatically,
even for those in the same sector. For example, in 2003, Hilton Hotels shifted its
pension portfolio more toward equities, upping the allocation from 32% in 2002 to 50%
while trimming fixed income from 68% to 50%. On the other hand, Starwood made the
exact opposite decision, completely eliminating the allocation to equities and upping the
allocation to fixed income.
We find it interesting that Hilton is taking on more equity exposure in a pension plan that
has been frozen since 1996 (i.e., employees have not accrued new benefits since
1996).
For other companies, the allocation to certain asset classes decreased significantly,
including the 20 companies in Exhibit 16. Note that 11 companies appear in both Exhibit
15 and Exhibit 16, which makes a lot of sense, as a large increase in the allocation to
one asset class is likely to result in a large decrease in the allocation to another asset
class.
Exhibit 16: Biggest Decrease in Asset Allocations among Different Asset Classes, 2003
Asset Allocations (%)
Company Ticker 2003A 2002A Decrease
Equity:
Biogen Idec Inc BIIB 0% 65% (65%)
Dover Corp DOV 53% 85% (32%)
Starwood Hotels & Resorts Wrld HOT 0% 26% (26%)
CMS Energy Corp CMS 48% 68% (20%)
AON Corp AOC 43% 62% (19%)
Fixed Income:
General Dynamics Corp GD 4% 34% (30%)
Biogen Idec Inc BIIB 12% 35% (23%)
Gannett Co GCI 29% 47% (18%)
Hilton Hotels Corp HLT 50% 68% (18%)
Moodys Corp MCO 21% 38% (17%)
Real Estate:
MBNA Corp KRB 0% 25% (25%)
Citizens Communications Co CZN 0% 10% (10%)
Principal Financial Grp Inc PFG 0% 5% (5%)
Medco Health Solutions Inc MHS 0% 4% (4%)
Weyerhaeuser Co WY 13% 17% (4%)
Other1:
TJX Companies Inc TJX 6% 29% (23%)
Anthem Inc ATH 2% 21% (19%)
Morgan Stanley MWD 14% 32% (18%)
Hershey Foods Corp HSY 2% 17% (15%)
Wachovia Corp WB 5% 20% (15%)
1
Includes private equity, hedge funds, short-term investments, cash, and cash equivalents.
Source: Company data, CSFB estimates.
30
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 17: Viewing the World Differently: Interesting Asset Allocation Disclosures Highlight Various Approaches
Company Ticker Investing Emphasis Disclosure
US Bancorp USB Equity Based on an analysis of historical performance by asset class, over any 20-year period since the mid-
1940s, investments in equities have outperformed other investment classes but are subject to higher
volatility. While an asset allocation including bonds and other assets generally has lower volatility and
may provide protection in a declining interest rate environment, it limits the pension plan's long-term
up-side potential. Given the pension plan's investment horizon and the financial viability of the
Company to meet its funding objectives, the Committee has determined that an asset allocation
strategy investing in 100% equities diversified among various domestic equity categories and
international equities is appropriate.
National City Corp NCC Equity The investment objective for the defined benefit pension plan is to maximize total return with tolerance
for slightly above average risk. Asset allocation strongly favors equities, with a target allocation of
approximately 80% equity securities, 15% fixed income securities, and 5% cash. Due to volatility in the
market, the target allocation is not always desirable and asset allocations will fluctuate between the
acceptable ranges. A core equity position of large cap stocks will be maintained. However, more
aggressive or volatile sectors will be meaningfully represented in the asset mix in pursuit of higher
returns. Higher volatility investment strategies such as credit risk, structured finance, and international
bonds will be appropriate strategies in conjunction with the core position. It is management's intent to
give the investment managers flexibility within the overall guidelines with respect to investment
decisions and their timing. However, certain investments require specific review and approval by
management. Management is also informed of anticipated changes in nonproprietary investment
managers, significant modifications of any previously approved investment, or anticipated use of
derivatives to execute investment strategies.
Merrill Lynch & Co MER Fixed Income The U.S. tax-qualified plan, which represents approximately 78% of Merrill Lynch's total plan assets as
of September 26, 2003, is 100% invested in a group annuity contract which is currently 100% invested
in fixed income securities.
Assets and liabilities of the U.S. tax-qualified plan are dollar duration matched, such that the assets are
expected to mature within one half year of the time when liabilities come due. The asset portfolio's
investment objective calls for a concentration in fixed income securities, the majority of which have an
investment grade rating.
AES Corp AES Various The U.S. Plans seek to achieve the following long-term investment objectives: (1) Maintenance of
sufficient income and liquidity to pay retirement benefits and other lump sum payments; (2) Long-term
rate of return in excess of the annualized inflation rate; (3) Long-term rate of return (net of relevant
fees) that meets or exceeds the assumed actuarial rate; (4) Long term competitive rate of return on
investments, net of expenses, that is equal to or exceeds various benchmark rates based on a full
investment cycle of 3 to 5 years, including a "policy index" consisting of 35% S&P 500 Index, 10%
Russell 2500 Index, 10% MSCI EAFE Index, 5% NAREIT, 35% Lehman Brothers Aggregate Bond
Index, and 5% Lehman Brothers High Yield Index. Consistent with the above, the allocation is
reviewed intermittently to determine a suitable asset allocation which seeks to control risk through
portfolio diversification and takes into account, among possible other factors, the above-stated
objectives, in conjunction with current funding levels, cash flow conditions and economic and industry
trends.
Equifax Inc EFX Various The USRIP [the U.S. Retirement Income Plan], in an effort to meet its asset allocation objectives,
utilizes a variety of asset classes which have historically produced returns which are relatively
uncorrelated to those of the S&P 500. Asset classes included in this category are Alternative Assets
(hedge funds-of-funds), venture capital (including secondary private equity), and real estate. The
primary benefits to the Plan of using these types of asset classes are: (1) their non-correlated returns
reduce the over-all volatility of the Plan's portfolio of assets, thereby moving the Plan closer to the
efficient investment frontier, and (2) they produce superior risk-adjusted returns, as measured by
standard metrics such as Jensen's Alpha and the Information Ratio. Additionally, the USRIP allows
certain of its managers, subject to specific risk constraints, to utilize derivative instruments, in order to
enhance asset return, reduce volatility, or both. Derivatives are primarily employed by the Plan in its
fixed income portfolio and in the hedge fund-of-funds area.
Weyerhaeuser Co WY Various The strategy of the U.S. pension trust is to invest directly and via total return partnership swaps in a
diversified mix of nontraditional strategies, including hedge funds, private equity, opportunistic real
estate and other externally managed alternative investment funds. Various financial instruments, such
as S&P 500 swaps and fixed income futures, are used to supplement the market exposures embedded
in such investments so the total effective exposures are maintained close to a target benchmark of 60
percent equity, 35 percent bonds and 5 percent cash on the trust's net asset value. As of December 1,
2003, the equity target was changed to 55 percent and the bond target was changed to 40 percent.
Source: Company 10-Ks.
31
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 18: Weyerhaeuser’s Asset Allocation Disclosure Exhibit 19: Bellsouth Corp’s Asset Allocation Disclosure
Allocation of Assets by Category Other
Pension Benefits
Qualified and registered pension plans DECEMBER 28, 2003 DECEMBER 29, 2002
-------------------------------------------------------
--------------------------------------------------- ----------------- ----------------- December 31 December 31
Private equity and related funds 29.2% 32.2%
Real estate and related funds 13.0 16.9
------------------------------
Common stock and S&P 500 total return index exposure 5.1 (10.1) Asset Category Target 2002 2003 Target 2002 2003
Bonds, Lehman government/credit and T-bills 0.3 0.4 -------------------------------------------------------
Short-term investments, cash and cash equivalents 21.2 18.0
Hedge funds 31.2 43.3
Equity securities 45-75% 56% 57% 50-85% 75% 78%
Net receivables 0.5 0.6 Debt securities 10-25 20 19 0-10 7 5
Accrued liabilities (0.5) (1.3) Real estate 5-15 10 10 5-15 5 4
----------------- ----------------- Other 5-15 14 14 15-25 13 13
Total 100.0% 100.0% -------------------------------------------------------
Total 100% 100% 100% 100%
----------------- -----------------
-------------------------------------------------------
Source: Weyerhaeuser’s 12/31/03 10-K filing. Source: Bellsouth’s 12/31/03 10-K filing.
32
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 20: Pension Plan Ownership of Company Stock > 10% of Plan Assets
% of Company Stock % of Company Stock
Company Ticker in Pension Plan 2003 in Pension Plan 2002
Rowan Cos Inc RDC 19% NA
Cincinnati Financial Corp CINF 18% 18%
Fifth Third Bancorp FITB 18% 23%
Southtrust Corp SOTR 14% 20%
McCormick & Co MKC 13% NA
Kinder Morgan Inc KMI 11% 10%
Masco Corp MAS 11% 5%
Sherwin-Williams Co SHW 10% NA
33
The Magic of Pension Accounting, Part III 07 February 2005
Remember, the pension obligation represents the promise to pay retirement benefits to
employees in the future. It is the present value of all the future retirement benefits
earned and not yet paid to date. Benefits paid (as its name would imply) are simply the
pension benefits currently paid to retirees. Our thought process here is quite simple: the
higher the benefits paid as a percentage of the pension obligation, the more retirees, the
more retirees the shorter the duration of the obligation. On the other hand, the lower the
benefits paid as a percentage of the pension obligation, the less retirees in the pension
plan, the less retirees the longer the duration of the obligation. Not perfect, but it should
get us in the ballpark, especially when we are looking at 370 companies. The matrix in
Exhibit 21 should hammer home this point.
We put this thought process to work by calculating the ratio of benefits paid to the
beginning-of-year projected benefit obligation (PBO) for each company in the S&P 500
over the past four years. Then, we calculated the four-year median ratio for each
company. We plot out the distribution of the ratios in Exhibit 22.
120 115
Number of Companies
100 97
80
60 52
44
40 29 33
20
-
Below 2.0% 2.0%-3.9% 4.0%-5.9% 6.0%-7.9% 8.0%-9.9% 10.0% or Above
As we move from left to right in Exhibit 22, the ratio of benefits paid to the pension
obligation increases, and we estimate the duration of the pension obligation shrinks.
Now there are a variety of reasons this relationship may not hold true; for example, a
lump-sum distribution could throw our calculation completely out of whack, making it
appear as if the company has a shorter duration pension obligation when its just a
onetime anomaly. We try to factor that into our analysis by using the median ratio for
each company over the past four years.
34
The Magic of Pension Accounting, Part III 07 February 2005
For the companies in the S&P 500, the median ratio of benefits paid to the pension
obligation was 6%, with 212 companies clustered between 4.0% and 7.9%. At one
extreme, we find 29 companies with a ratio of benefits paid to the pension obligation of
less than 2%, including the 10 companies on the left hand side of Exhibit 23 with
pension obligations greater than $100 million. For these companies, it appears as if they
have relatively longer duration pension obligations. At the other end of the spectrum, we
find 33 companies where the ratio is greater than 10%, including the 10 companies on
the right hand side of Exhibit 23 with pension obligations greater than $100 million and a
ratio of 13% or higher. For these companies, the pension obligation appears to have a
relatively shorter duration.
Exhibit 23: Trailing Four-Year Median Ratio of Benefits Paid to Beginning of Year PBO
(PBO > $100 Million)
Longer Duration Obligation Shorter Duration Obligation
Low Ratio of High Ratio of
Benefits Paid Benefits Paid
Company Ticker to PBO Company Ticker to PBO
MBNA Corp KRB 1% SBC Communications Inc SBC 18%
Sabre Hldgs Corp -Cl A TSG 1% Peoples Energy Corp PGL 15%
FedEx Corp FDX 1% Fifth Third Bancorp FITB 15%
National Semiconductor Corp NSM 2% Bank One Corp ONE 14%
MGIC Investment Corp/Wi MTG 2% Qwest Communication Intl Inc Q 14%
Federal Home Loan Mortg Corp FRE 2% Bellsouth Corp BLS 13%
Autozone Inc AZO 2% Stanley Works SWK 13%
Hilton Hotels Corp HLT 2% U S Bancorp USB 13%
AFLAC Inc AFL 2% AES Corp. (The) AES 13%
McCormick & Co MKC 2% Sunoco Inc SUN 13%
Source: Company data, CSFB estimates.
35
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 24: Ratio of Benefits Paid to PBO versus Fixed-Income Asset Allocations
Shorter
Term
18%
SBC
16% PGL
8% WY
LTR
GD
6% AYE MER
4% AOC
INTC HDI
2% EDS HLT
MBNA
FDX TSG
0%
Longer
Term 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
2003 Actual Fixed Income Asset Allocation (%)
• Why do companies that appear to have pension obligations with drastically different
durations have nearly the same allocation to fixed-income assets? For example, SBC
and FDX.
• How can companies like USB and AES that both appear to have pension obligations
with roughly the same duration have asset allocations that are at opposite ends of the
spectrum? USB has allocated 0% of its assets to fixed income, while AES has a 64%
allocation to fixed income. USB answers that question in its 10-K: “ . . . given the
pension plan's investment horizon and the financial viability of the Company to meet
its funding objectives, the Committee has determined that an asset allocation strategy
investing in 100% equities diversified among various domestic equity categories and
international equities is appropriate.”
• MER has an 84% fixed-income allocation. Why do many other companies with
apparently shorter duration obligations have substantially lower fixed-income
allocations? MER’s 10-K states: “ . . . assets and liabilities of the U.S. tax-qualified
plan are dollar duration matched, such that the assets are expected to mature within
one half year of the time when liabilities come due. The asset portfolio's investment
objective calls for a concentration in fixed income securities, the majority of which
have an investment grade rating.”
36
The Magic of Pension Accounting, Part III 07 February 2005
1
Exhibit 25: Companies with Shorter Duration Pension Obligations —Highest Equity and Fixed-Income Asset Allocations
Highest 2003 Equity Allocation Highest 2003 Fixed Income Allocation
Ratio of Equity Fixed Income Ratio of Equity Fixed Income
Benefits Paid Allocation Allocation Benefits Paid Allocation Allocation
Company Ticker to PBO 2003 2003 Company Ticker to PBO 2003 2003
U S Bancorp USB 13% 97% 0% Viacom Inc VIA 10% 30% 65%
Williams Cos Inc WMB 10% 82% 13% AES Corp. (The) AES 13% 30% 64%
Eaton Corp ETN 10% 81% 18% CMS Energy Corp CMS 10% 48% 52%
Eastman Chemical Co EMN 10% 80% 11% Kerr-McGee Corp KMG 10% 55% 41%
Keycorp KEY 10% 73% 25% ConocoPhillips COP 10% 55% 40%
Tribune Co TRB 11% 72% 23% Stanley Works SWK 13% 59% 39%
Fifth Third Bancorp FITB 15% 72% 26% Bank One Corp ONE 14% 63% 37%
Fleetboston Financial Corp FBF 10% 72% 21% Monsanto Co MON 11% 60% 35%
Bank Of America Corp BAC 10% 71% 28% Sunoco Inc SUN 13% 62% 33%
El Paso Corp EP 13% 70% 29% Safeco Corp SAFC 12% 67% 33%
1
Based on trailing 4-year median ratio of benefits paid to PBO as of the beginning of the year (PBO > $100 Million).
Note: Asset allocations are based on combined U.S. and International plan disclosures.
Source: Company data, CSFB estimates.
37
The Magic of Pension Accounting, Part III 07 February 2005
1
Exhibit 26: 2003 Asset Allocations versus Longer Duration Obligations —High Equity versus High Fixed-Income Asset
Allocations
Highest 2003 Equity Allocation Highest 2003 Fixed Income Allocation
Ratio of Equity Fixed Income Ratio of Equity Fixed Income
Benefits Paid Allocation Allocation Benefits Paid Allocation Allocation
Company Ticker to PBO 2003 2003 Company Ticker to PBO 2003 2003
Electronic Data Systems Corp EDS 3% 87% 8% Starwood Hotels & Resorts HOT 3% 0% 79%
MGIC Investment Corp/Wi MTG 2% 80% 16% Teradyne Inc TER 3% 48% 53%
EMC Corp/Ma EMC 3% 75% 25% Hilton Hotels Corp HLT 2% 50% 50%
Albertsons Inc ABS 3% 71% 27% Sabre Hldgs Corp -Cl A TSG 1% 56% 44%
MBNA Corp KRB 1% 71% 19% Allergan Inc AGN 3% 58% 41%
McCormick & Co MKC 2% 68% 32% AFLAC Inc AFL 2% 48% 40%
American Express AXP 3% 66% 26% Thermo Electron Corp TMO 3% 60% 39%
Harley-Davidson Inc HDI 3% 66% 13% Genuine Parts Co GPC 3% 62% 35%
1
Based on trailing 4-year median ratio of benefits paid to PBO as of the beginning of the year (PBO > $100 Million).
Note: Asset allocations are based on U.S. and International plan disclosures.
Source: Company data, CSFB estimates.
So has the president’s plan for pension reform sparked a big shift recently in asset
allocation within defined benefit pension plans? Anything is possible. (Remember, we
are not asset allocation experts, only accountants.) However, there is still a lot of
uncertainty as to how the reforms will turn out, the president’s plan is still just a plan. We
have a long way to go to get to pension reform. Pension reform has been discussed for
years without any real reforms taking place, and if pension reform is enacted, there will
likely be a transition period to phase in the new reforms. Let’s not forget that a big shift
in asset allocation out of equities and into fixed income by companies with defined
benefit pension plans would likely put a dent in earnings by reducing the expected rate
of return on plan assets, which increases pension cost reported on the income
statement. For some, the dent could be quite large. (See the Who Has Exposure?
section of the report for an idea as to what could happen to earnings if the expected
return assumption were to drop.)
However, defined benefit pension plans face a number of potential events and key
trends that could, from a macro perspective, spark a shift in asset allocation by pension
plans. The potential events include a change in pension accounting, pension reform,
pension plans getting healthier, and more pension plans getting dropped on the PBGC,
the key trends are demographics and companies continuing to move away from defined
benefit plans. The question is whether or not we are approaching some sort of tipping
point where the stars have aligned so that these events and trends actually converge at
or around the same time.
38
The Magic of Pension Accounting, Part III 07 February 2005
Another question that comes to mind: If there are shifts in asset allocation do they occur
gradually or suddenly? Remember that defined benefit pension plans still have a large
presence in the United States, comprising about $4.2 trillion in total assets and split
between corporate plans $1.9 trillion and public (state, local, etc.) plans $2.3 trillion. The
asset allocation in the corporate plans is 55% equities, 26% fixed income, and 19% in
various other asset classes. As a point of reference, the total market cap of the U.S.
stock market was about $13.3 trillion at the end of 2004. Clearly sudden shifts in asset
allocation by defined benefit pension plans could cause market dislocations. In addition
as new contributions are made to the pension plans a shift in asset allocation could
affect the demand for equities and fixed income in the future.
We also wonder whether or not the shifts in asset allocation by defined benefit plans—
assuming they take place—can be offset by the asset allocation decisions in other
retirement vehicles, for example, 401k plans and the creation of social security private
accounts.
At this point we still have more questions than answers, however, we plan on doing
more work on this topic, in the meantime we wanted to run you through some of our
thoughts on the potential events and trends facing the defined benefit system that could
spark a shift in asset allocation.
Potential Events
FASB Changes the Accounting
Pension accounting today, with all of its smoothing mechanisms, protects companies
from volatile markets, hiding the volatility in the pension plan and making it appear as if
the plan is virtually risk free and everything’s just fine and dandy. In our view, it is just a
matter of time before the FASB fixes pension accounting—finally requiring that the real
economics of a pension plan be reported in the financial statements, not tucked away in
a footnote—and more of the pension plans’ volatility could find its way onto financial
statements. One potential way to minimize that volatility is to match pension plan assets
more closely with pension obligations.
Pension Reform
• Yield curve approach. If the government institutes the yield-curve approach for
determining the pension obligation for funding purposes, pension funding would
become dependent upon the current yield on high-grade corporate bonds. To
minimize the volatility in required contributions to pension plans, companies with
defined benefit pension plans could shift their pension plan asset allocations toward
high-grade corporate bonds, so that the plan assets move in step with the pension
obligations.
• Funding and premiums. The proposed pension reforms would draw a more direct line
between the health of the pension plan, its funded status, and company contributions
to the pension plan, along with PBGC premiums. In other words, if the funded status
of the pension plan were to deteriorate rapidly, it would hit the company much harder
and more quickly than it does today in the form of higher contributions and PBGC
premiums. To minimize the likelihood that a large contribution is required or PBGC
premium due in any given year, companies may try to match their pension plan assets
more closely with their pension obligations.
39
The Magic of Pension Accounting, Part III 07 February 2005
• Cash balance plans. If the pension reforms ease the transition to cash balance plans,
so that a conversion to a cash balance plan is not age discriminatory, we could see
more companies choose this type of pension plan. A cash balance plan is a cross
between a defined contribution plan and a defined benefit plan. It is a defined benefit
pension plan in which the benefit is defined in terms of a cash balance, so that it looks
like a defined contribution plan. That cash balance typically grows each year with a
pay credit (a percentage of a worker’s salary) and an interest credit (generally tied to
Treasury yields—for example, the U.S. government long bond). If these plans become
popular, we could see a shift toward fixed-income assets in pension plans that would
better match the returns promised to employees.
• Required asset allocations. Some have speculated that pension reform would have an
impact on asset allocations by requiring that companies follow a certain mandated
asset allocation in their pension plans. We are currently aware of no plans to take
reforms down this path. In addition, there had been some discussion about companies
being penalized through higher PBGC premiums for pursuing a “riskier” asset
allocation strategy (i.e., heavy equity allocations) in the pension plan. Reading
between the lines of the president’s proposal, it appears as if the Administration is
content with measuring risk and charging PBGC premiums based on the funded
status of the pension plan and the financial health of the company.
Key Trends
Companies Continuing to Move Away from Defined Benefit Pension Plans
Some of the events that we describe above, a change in pension accounting, pension
reforms, and healthier pension plans could accelerate what has been a trend for quite
some time: the demise of defined benefit pension plans. For example, companies could
complain that if the FASB adopts a mark-to-market approach for pension accounting,
the balance sheet and potentially earnings would become too volatile. Claiming that
investors can’t handle the volatility, they may threaten to shut down their defined benefit
40
The Magic of Pension Accounting, Part III 07 February 2005
plan. Remember, the pension plans are already volatile; it is just that the accounting
today hides it very well. Companies may also threaten to drop the plans if PBGC
premiums increase too high or funding requirements rise or become too volatile.
If companies continue to move away from defined benefit pension plans, and that trend
accelerates, then asset allocation shifts could take place. For example:
• As some companies have transitioned away from defined benefit pension plans for
younger workers, and more continue to do so (e.g., IBM), their pension plans become
a legacy issue. With no new employees coming into the plan, as each year passes,
the pension population gets older, with more and more of the pension plan associated
with older workers and retirees. Therefore, pension obligations eventually become
shorter term, potentially causing a gradual shift in asset allocations toward fixed
income.
• Assuming pension plans get healthier, some companies may even choose to
terminate their plans, as companies are allowed to terminate their pension plan if the
plan is fully funded or overfunded. A company can terminate its pension plan by
paying a lump sum to the pension plan participants, which would involve liquidating
the pension plan assets. Or the company could choose to purchase an annuity
contract from an insurance company, transferring the risk of the pension plan to the
insurance company. (For more on pension plan terminations, see our January 19,
2005 report, The Next Savings and Loan Crisis?)
41
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 27: Percentage of Companies with Defined Benefit Pension Plans by Industry—S&P 500
Industry % Industry % Industry %
Aerospace & Defense 100% Machinery 100% Personal Products 67%
Air Freight & Logistics 100% Office Electronics 100% IT Services 64%
Auto Components 100% Paper & Forest Products 100% Consumer Finance 60%
Automobiles 100% Road & Rail 100% Food & Staples Retailing 60%
Beverages 100% Tobacco 100% Health Care Providers & Services 59%
Building Products 100% Oil & Gas 93% Commercial Services & Supplies 54%
Chemicals 100% Insurance 88% Airlines 50%
Construction & Engineering 100% Metals & Mining 88% Computers & Peripherals 50%
Construction Materials 100% Multi-Utilities & Unregulated Power 88% Hotels Restaurants & Leisure 45%
Containers & Packaging 100% Commercial Banks 83% Communications Equipment 43%
Distributors 100% Electrical Equipment 83% Specialty Retail 41%
Diversified Financial Services 100% Health Care Equipment & Supplies 83% Textiles Apparel & Luxury Goods 40%
Diversified Telecom Services 100% Thrifts & Mortgage Finance 83% Wireless Telecom Services 33%
Electric Utilities 100% Electronic Equipment & Instruments 80% Semiconductors & Semiconductor Equip. 32%
Energy Equipment & Services 100% Media 80% Real Estate 17%
Food Products 100% Household Durables 77% Software 6%
Gas Utilities 100% Pharmaceuticals 75% Internet & Catalog Retail 0%
Household Products 100% Capital Markets 73% Internet Software & Services 0%
Industrial Conglomerates 100% Multiline Retail 73% Trading Companies & Distributors 0%
Leisure Equipment & Products 100% Biotechnology 67%
Source: Company data, CSFB estimates.
Once we have isolated the companies with defined benefit pension plans, we analyze
the data from a bunch of different angles. To answer the simple question—Who has the
most exposure to pension issues?—we focus on four types of exposure: shareholder,
cash flow, earnings, and balance sheet. We start off this analysis by forecasting future
results using our pension forecast model.
42
The Magic of Pension Accounting, Part III 07 February 2005
Forecasting
In our September 8, 2004, report, The Magical World of Pensions: An Update, we
updated some of our forecasts and forward-looking pension analysis for the companies
in the S&P 500. At that point, we were estimating that the defined benefit pension plans
were getting weaker in 2004, potentially ending 2004 in worse shape than in 2002.
Remember, way back then the stock market was virtually flat for the year. Of course,
everything changed in the fourth quarter and the market had a powerful post-election
rally to finish the year in positive territory.
Now that we find ourselves in a new year, we thought it would make sense to update
our pension forecasts. We are changing two key assumptions: the actual return on plan
assets and the discount rate. Taking into account the phenomenal performance in the
stock market in the last quarter of 2004, we now assume a 9.38% return on plan assets
in 2004, up from our original assumption of 1.05%. In addition, as interest rates
continued to trail down last year, we take the discount rate down 25 basis points for
2004; we had originally kept discount rates flat. As a result, we now estimate that the
pension plans of the companies in the S&P 500 got a bit less healthy in 2004, going
from $172 billion underfunded at the end of 2003 to our revised estimate of $185 billion
underfunded at the end of 2004. We kept our many other simplifying assumptions the
same.
Once you’ve accessed CSFB’s Research & Analytics Web site, you’ll find two white
boxes, one labeled “Search” and one labeled “For” in the top left hand corner of the
screen. The “Search” box contains a drop-down menu of different items you might want
to search, but to access our model select “Analyst,” enter “Zion” in the “For” box, and
just hit “GO.” Click on the Excel file “Pension Forecast Model,” and you’re off to the
races.
Remember that if the company is in the S&P 500, all you have to do is type in the ticker
symbol and the historical pension data will update automatically. There is no data entry
required unless you want to input your own assumptions about the future to generate a
forecast (the model already includes our base case assumptions). For those interested
in analyzing companies outside the S&P 500, give us a call; we have another version of
the model that we can make available to you. We describe the model and its limitations,
and demonstrate its flexibility, in the Magic of Pension Accounting, Part II.
43
The Magic of Pension Accounting, Part III 07 February 2005
We dropped the discount rates used for accounting purposes by 25 basis points for
2004, as the yield on the Moody’s Aa corporate bond index dropped 35 basis points in
2004, from 6.01% on December 31, 2003, to 5.66% on December 31, 2004. We use the
Moody’s Aa as a proxy, as the discount rates used to calculate the projected benefit
obligation (PBO) are based on high-grade corporate bonds. After discussions with our
interest-rate research team, we assume that discount rates will rise 25 basis points in
2005 and then go up another 25 basis points in 2006.
44
The Magic of Pension Accounting, Part III 07 February 2005
45
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 29: Funded Status of Defined Benefit Pension Plans for the S&P 500
130%
120%
Percent Funded
110%
100%
90%
80%
70%
'93 '94 '95 '96 '97 '98 '99 '00 '01 '02 '03 '04E '05E '06E
The Key Drivers behind Estimated Pension Plan Assets and Estimated PBO
With strong market returns in 2004, you may be wondering why we are estimating that
the pension plans got weaker last year. In Exhibit 30 and Exhibit 31, we show you how
we got there by presenting the key drivers behind the change in aggregate plan assets
and the aggregate pension obligation for the companies in the S&P 500 using our base-
case forecast. We estimate that the plan assets grew by 4.4%; however, the pension
obligation grew a bit faster at 4.7%. If we started the year in a hole and then we
estimate the pension obligation grows faster than the plan assets, the hole is going to
grow larger.
Exhibit 30: Reconciliation of Plan Assets for S&P 500, 2003–04E Exhibit 31: Reconciliation of PBO for S&P 500, 2003–04E
US$ in millions US$ in millions
Source: Company data, CSFB estimates. Source: Company data, CSFB estimates.
46
The Magic of Pension Accounting, Part III 07 February 2005
Shareholders’ Exposure
We start by determining the defined benefit pension plans that could have the largest
claim on the shareholders’ stake in a company. To do this, we simply compare the
economic value of the pension plan, its funded status, with the value of the
shareholders’ stake in the company, the equity market capitalization. The distribution of
this relationship for the companies in the S&P 500 is included in Exhibit 32.
1
Exhibit 32: Pension Plan Funded Status/Market Capitalization
200
Higher Shareholder Exposure Lower Shareholder Exposure
180 173
160
140
Underfunded Overfunded
# of Companies
120 110
100
80
60
40 31 29
20 8
8
2
0
Below 25% (24.9%) to (10.1%) (10%) to (2.5%) (2.4%) to (0.1%) 0% to 2.4% 2.5% to 9.9% 10% or Above
1
Represents ratio of 12/31/04E funded status to 12/31/04 market capitalization.
Source: Company data, CSFB estimates.
We estimate that the eight companies in Exhibit 33 had defined benefit pension plans at
the end of 2004 that were underfunded by more than 25% of their equity market
capitalization as of December 31, 2004. In other words, the pension plan may have a
claim on over one-quarter of the shareholders’ stake in these companies.
When we ran a similar analysis in The Magic of Pension Accounting, Part II, we
estimated that 30 companies had defined benefit pension plans at the end of 2003 that
were underfunded by more than 25% of their market cap. No surprises, all of the
companies on this year’s list are repeat performers.
Exhibit 33: Estimated 2004 Pension Underfunding > 25% of Equity Market Capitalization
US$ in millions
47
The Magic of Pension Accounting, Part III 07 February 2005
On the other hand, we estimate that 212 companies in the S&P 500 may have defined
benefit pension plans at the end of 2004 that are underfunded by less than 2.5% of their
equity market capitalization. We estimate that 39 companies may have overfunded
pension plans at the end of 2004, with only two companies—Sherwin-Williams and
AT&T Corp—that could have pension plans that are overfunded by more than 10% of
their equity market cap.
Next, we evaluate which companies have the largest potential exposure, the highest
leverage to defined benefit pension plans, by comparing the pension obligation to the
equity market capitalization. We estimate that the 20 companies in Exhibit 34 had
projected benefit obligations at the end of 2004 that exceeded their equity market
capitalization as of December 31, 2004. All but three of the companies were on a similar
list that we ran back in October 2003. The three newcomers are AT&T, Qwest
Communications, and Maytag. For 103 companies with defined benefit plans, the
estimated PBOs at the end of 2004 could represent less than 5% of their current equity
market cap.
Exhibit 34: Estimated 2004 Projected Benefit Obligation > Equity Market Capitalization
US$ in millions
2004E PBO/
12/31/04 Equity Market
Company Ticker 2004E PBO Market Cap Cap
Delta Air Lines Inc DAL $ 12,664 $ 968 1,308%
General Motors Corp GM 105,562 22,627 467%
Goodyear Tire & Rubber Co GT 7,167 2,571 279%
Ford Motor Co F 68,518 25,749 266%
Delphi Corp DPH 12,392 5,062 245%
Unisys Corp UIS 6,462 3,420 189%
Visteon Corp VC 2,384 1,266 188%
Lucent Technologies Inc LU 30,804 16,622 185%
United States Steel Corp X 8,004 5,827 137%
Navistar International NAV 4,042 3,071 132%
Dana Corp DCN 2,961 2,596 114%
AT&T Corp T 16,833 15,169 111%
Qwest Communication Intl Inc Q 8,879 8,063 110%
Pactiv Corp PTV 4,029 3,754 107%
Eastman Kodak Co EK 9,763 9,245 106%
Lockheed Martin Corp LMT 25,950 24,591 106%
Hercules Inc HPC 1,745 1,663 105%
Coors (Adolph) - Cl B RKY 2,784 2,737 102%
Allegheny Technologies Inc ATI 2,086 2,070 101%
Maytag Corp MYG 1,665 1,672 100%
Source: Company data, CSFB estimates.
48
The Magic of Pension Accounting, Part III 07 February 2005
make to their pension plans during the next fiscal year. In their 2003 10-Ks the
companies in the S&P 500 disclosed they would contribute $24 billion to their pension
plans during 2004. For companies that provide this disclosure we work this estimate into
our forecasts. For those companies that don’t provide it we make an estimate of
required contributions in 2004.
With 2004 now over, we look to 2005. To provide investors with an estimate of the
potential contributions to the pension plan in 2005, we take the information in the
pension footnote and transform it into the funded status according to the funding
requirements, using a few rules of thumb and some back-of-the-envelope calculations in
our pension-forecasting model.
We expect the companies in the S&P 500 to contribute an aggregate $14 billion to their
pension plans in 2005. We expect our forecast of required contributions to differ from
the companies’ actual funding requirements due to a number of our simplifying
assumptions, which we elaborated on in our September 8 report. In Exhibit 35, we
provide the aggregate pension contribution for the companies in the S&P 500 during
2000-03 and our estimates for 2004 and 2005.
9%
$80 8%
7%
$74
$60 6% 6%
5%
$40 4% 4%
$46
2% 3%
2%
2%
$20 2%
$27
$16 1%
$14 $14
$0 0%
2000 2001 2002 2003 2004E 2005E
Note: 2004E and 2005E percentages are based on trailing five-year average cash flow from operations.
Source: Company data, CSFB estimates.
49
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 36: Estimated 2005 Pension Contributions > 25% of Trailing Five-Year Average
Cash Flow from Operations, by Company
US$ in millions
Earnings Exposure
The aggregate benefit to earnings from the pension plans of the companies in the S&P
500 dried up in 2002. Although pension plans got healthier in 2003, their improvement
wasn’t enough to offset the losses that had been smoothed out of earnings in the
previous two years. The companies in the S&P 500 reported $28 billion of pension
expense in 2003, a dramatic jump up from $3 billion in 2002 and a complete turnaround
from 2000 when the companies reported net pension income of $12 billion. It looks like
the pension plan is no longer a profit center for most companies according to Exhibit 37.
$35
$31 $31
$30 $28
$26
$25
Pension (Income)/Expense
$20
$15 $12
$10 $9 $9 $7
$10
$6
$5 $3
$0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004E 2005E 2006E
($5)
($2)
($10) ($7)
($15) ($12)
50
The Magic of Pension Accounting, Part III 07 February 2005
We estimate that 232 companies will experience an increase in pension cost between
2004 and 2005, causing an earnings headwind that further reduces earnings in 2005.
We estimate that the 18 companies in Exhibit 38 will experience an increase in pension
cost between 2004 and 2005 of at least $0.10 per share.
Exhibit 38: Estimated Earnings Headwind from the Pension Plan > $0.10 per Share
2004E 2005E
Pension (Income)/ Pension (Income)/
Company Ticker Expense per Share Expense per Share Increase
Delta Air Lines Inc DAL $ 2.74 $ 3.07 $ 0.34
Intl Business Machines Corp IBM (0.26) 0.03 0.28
Boeing Co BA 0.13 0.36 0.24
General Motors Corp GM 3.74 3.96 0.22
Snap-On Inc SNA (0.31) (0.10) 0.21
FedEx Corp FDX 0.42 0.61 0.19
Lockheed Martin Corp LMT 0.98 1.14 0.16
ITT Industries Inc ITT 0.26 0.39 0.13
Parker-Hannifin Corp PH 0.38 0.51 0.13
Textron Inc TXT 0.01 0.13 0.12
Peoples Energy Corp PGL (0.02) 0.10 0.12
United Technologies Corp UTX 0.27 0.39 0.11
Eastman Chemical Co EMN 0.34 0.46 0.11
Eaton Corp ETN 0.32 0.43 0.10
Pactiv Corp PTV (0.01) 0.09 0.10
Black & Decker Corp BDK 0.24 0.34 0.10
Raytheon Co RTN 0.66 0.76 0.10
Ford Motor Co F 0.32 0.42 0.10
Source: Company data, CSFB estimates.
51
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 39: Estimated Earnings Headwind from the Pension Plan > 5% of 2005 First Call
Consensus Earnings Estimate
A
Pension Cost/Share B
Increase between 2004E 2005E First Call A/B
Company Ticker and 2005E Consensus Percentage
Unisys Corp UIS $ 0.07 $ 0.39 19%
Lucent Technologies Inc LU 0.03 0.21 13%
Snap-On Inc SNA 0.21 1.66 13%
CMS Energy Corp CMS 0.07 0.76 9%
Boeing Co BA 0.24 2.58 9%
Pactiv Corp PTV 0.10 1.59 6%
Source: Company data, CSFB estimates.
Using a few actuarial rules of thumb, we estimate the ABO at the end of 2004 for each
company in the S&P 500, then compare it with our estimate of the plan assets to arrive
at the minimum pension liability. Taking into account the minimum liability adjustments
from prior years, prior service cost, any remaining transition obligation, and assuming a
35% tax rate, we arrive at an estimate of the after-tax charge to equity of $11 billion that
the companies in the S&P 500 might have to make at the end of 2004. Based on these
assumptions, we estimate that 12 companies will have to take a charge to equity of over
$500 million in 2004; the 13 companies in Exhibit 40 might have to record a charge to
equity that would exceed 10% of their total equity at the end of 2003.
Exhibit 40: Estimated 2004 Minimum Pension Liability Charge to Equity > 10% of Total
Equity
US$ in millions
52
The Magic of Pension Accounting, Part III 07 February 2005
Sensitivity Analysis
We recommend using our model to perform what-if scenarios to gauge the potential
exposure of a company to its defined benefit pension plan. The model is available on
CSFB’s Research & Analytics Web site, and we leave it in your capable hands to come
up with a variety of scenarios.
For now, we provide you with a high-level sensitivity analysis, which we generated using
our model, to give an indication of how sensitive the aggregate funded status, pension
cost, and pension contributions are to changes in the actual return on plan assets and
the discount rate. A couple of reference points when examining these different
sensitivities:
• The median discount rate used for accounting purposes at the end of 2003 was
6.10%; our forecast assumes that the discount rate dropped 25 basis points in 2004.
• The actual return on plan assets in 2003 was 18.80%; our estimate for 2004 is that
the plan assets had a positive return of 9.38%.
• The discount rate used for funding purposes in 2003 was 6.59%; we assume a
discount rate of 6.14% in 2004.
53
The Magic of Pension Accounting, Part III 07 February 2005
Earnings Sensitivity
The matrix in Exhibit 42 provides the sensitivity of total estimated pension cost in 2005
for the companies in the S&P 500 to varying levels of discount rates and actual returns.
We estimate aggregate pension cost of $31 billion in 2005. To maintain pension cost at
its estimated 2004 level, $26 billion, discount rates would have to jump 50 basis points
with the 9.38% return we are assuming for 2004. Under no scenario presented would
we revert back to the good ol’ days of pension income.
54
The Magic of Pension Accounting, Part III 07 February 2005
Our base-case scenario for 2005 keeps the expected rate of return assumption constant
at the same level that each company used in 2003. In Exhibit 44, we display what
would happen to aggregate pension cost for the companies in the S&P 500 if we were
to ratchet down the 2005 expected return assumption by 50-basis-point increments
up to a 200-basis-point reduction. For example, if we drop the expected rate of return
by 100 basis points, we estimate aggregate pension cost for the S&P 500 would
increase by $12 billion, from $31 billion to $43 billion, reducing aggregate S&P 500
earnings by nearly 2%. If we drop the expected return assumption 200 basis points,
estimated pension cost increases by $24 billion and S&P 500 earnings drop by
approximately 3.5%.
55
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 44: S&P 500 Pension Cost Assuming Different Expected Rates of Return, 2005E
US$ in billions, unless otherwise stated
$60 8.5%
$55
8.0%
$49
1
Pension Cost as a % 2005E Earnings
$50 8.1%
$43 7.5%
$20 5.5%
5.5%
5.0%
$10
4.5%
4.6%
$0 4.0%
Current Level Down 50 bps Down 100 bps Down 150 bps Down 200 bps
Change in Expected Rate of Return
1
Equals 2005E pension cost/(1/11/05 First Call consensus estimate x 2003 diluted share count).
Source: Company data, CSFB estimates.
We drill down to the company level, focusing on those companies that could experience
the largest increase in pension cost per share from changing the expected rate of
return. We first calculate pension cost per share assuming expected returns stay flat
(we also assume a 35% tax rate for all companies and keep the share counts at 2003
levels), then we calculate pension cost assuming a 100-basis-point drop in the expected
return assumption; last we compare the two. For 52 companies we estimate pension
cost would increase by at least $0.10 per share, including the 17 companies in Exhibit
45 with an increase of at least $0.20 per share from a 100-basis-point decline in the
expected rate of return.
56
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 45: 2005E Pension Cost per Share Increases by > $0.20 from a 100-Basis-Point Drop in Expected
Rate of Return
Pension Cost per Share, 2005E
Expected Return Expected Return Increase
Company Ticker Remains Flat Drops 100 bps per Share
General Motors Corp GM $ 3.96 $ 5.01 $ 1.05
United States Steel Corp X 1.30 1.73 0.43
Coors (Adolph) - Cl B RKY 1.17 1.54 0.37
Delta Air Lines Inc DAL 3.07 3.44 0.36
Cummins Inc CMI 1.56 1.89 0.33
Boeing Co BA 0.36 0.67 0.31
Lockheed Martin Corp LMT 1.14 1.44 0.30
NCR Corp NCR 0.72 1.00 0.29
Intl Business Machines Corp IBM 0.03 0.31 0.29
Navistar International NAV 0.72 1.00 0.28
ITT Industries Inc ITT 0.39 0.65 0.26
Textron Inc TXT 0.13 0.36 0.23
General Dynamics Corp GD 0.38 0.60 0.22
Consolidated Edison Inc ED (0.25) (0.03) 0.21
Northrop Grumman Corp NOC 1.46 1.67 0.21
Ford Motor Co F 0.42 0.62 0.20
United Technologies Corp UTX 0.39 0.58 0.20
Source: Company data, CSFB estimates.
To measure whether an increase in pension cost due to a change in the expected return
assumption is meaningful to a particular company’s future earnings, we compare our
estimated increase in pension cost per share due to a 100-basis-point drop in the
expected rate of return to the First Call consensus earnings estimate. The resulting
increase in pension cost represents at least 5% of the consensus earnings estimate for
the 29 companies in Exhibit 46. In other words, a 100-basis-point drop in the expected
rate of return would reduce 2005 earnings by more than 5% for each of the 29
companies in Exhibit 46.
57
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 46: 2005E Pension Cost per Share as a % of Consensus Increases by > 5% from a 100-Basis-
Point Drop in Expected Rate of Return
A
Pension Cost/Share Increase
from a 100-Basis-Point B A/B
Company Ticker Drop in the Expected 2005E First Call Consensus Percentage
Rate of Return
Unisys Corp UIS $ 0.12 $ 0.39 30%
Lucent Technologies Inc LU 0.05 0.21 24%
General Motors Corp GM 1.05 4.78 22%
Goodyear Tire & Rubber Co GT 0.15 0.72 21%
NCR Corp NCR 0.29 2.30 12%
Boeing Co BA 0.31 2.58 12%
Allegheny Technologies Inc ATI 0.14 1.17 12%
Ford Motor Co F 0.20 1.91 10%
Electronic Data Systems Corp EDS 0.07 0.74 10%
Boise Cascade Corp BCC 0.14 1.56 9%
Pactiv Corp PTV 0.15 1.59 9%
Lockheed Martin Corp LMT 0.30 3.32 9%
Raytheon Co RTN 0.16 1.87 9%
AT&T Corp T 0.14 1.78 8%
Consolidated Edison Inc ED 0.21 2.87 7%
Eastman Kodak Co EK 0.19 2.55 7%
Hercules Inc HPC 0.08 1.14 7%
Coors (Adolph) - Cl B RKY 0.37 5.31 7%
Goodrich Corp GR 0.12 1.77 7%
Maytag Corp MYG 0.10 1.48 6%
Dana Corp DCN 0.10 1.62 6%
United States Steel Corp X 0.43 6.76 6%
Snap-On Inc SNA 0.10 1.66 6%
Northrop Grumman Corp NOC 0.21 3.53 6%
Navistar International NAV 0.28 4.79 6%
Cooper Tire & Rubber CTB 0.08 1.35 6%
Textron Inc TXT 0.23 3.97 6%
Meadwestvaco Corp MWV 0.11 1.92 6%
CMS Energy Corp CMS 0.04 0.76 6%
Source: Company data, CSFB estimates.
58
The Magic of Pension Accounting, Part III 07 February 2005
Key Metrics
We arrive at a weighted-average score for each industry group by selecting five key
metrics to include in a simple test:
1. Funded status/market capitalization—focuses on the size of the claim that the
pension plan has on the shareholders’ stake in the company, applying a score
based on how large the over- or underfunded status of the pension plan is as a
percent of equity market capitalization.
2. Projected benefit obligation/market capitalization—focuses on a company’s
potential exposure to its defined benefit pension plan, ignoring the plan’s assets and
instead applying a score based how large the pension obligation is as a percent of
equity market cap.
3. Contributions/cash flow from operations—focuses on the expected pension related
cash burden that companies could face, comparing estimated 2005 pension
contributions to trailing average five-year cash flows from operations.
4. Real pension cost adjustment/operating income—focuses on the impact on
operating income of stripping out reported pension cost and replacing it with service
cost, applying a score based on the magnitude of the adjustment to operating
income. We evaluate this adjustment as a percentage of operating income over the
past five years.
5. Off-balance-sheet pension liability adjustment/reported equity—focuses on the
impact that placing the funded status of the pension plan on balance sheet would
have on its reported equity, applying a score based on the magnitude of the hit to
shareholders’ equity.
We believe these are critical metrics for evaluating an industry’s exposure to defined
benefit pension plans, but we have a problem combining them in our analysis because
each metric is unique. We turn to a useful statistical tool to overcome this hurdle: the Z-
Score. Z-Scores help to “standardize” each metric with respect to the others. Once we
have Z-Scores across each key metric, we can combine them to come up with one
overall score for each industry. As we consider each of the five metrics equally
important, we apply equal 20% weights to each.
59
The Magic of Pension Accounting, Part III 07 February 2005
• Presentation. In our last pension report card, we provided a Z-Score for each metric
along with a weighted total Z-Score. We now only provide the weighted total Z-Score
along with actual amounts for each metric. For example, we now show you what the
actual industry amount is for the Funded Status/Market Cap metric instead of showing
you the Z-Score for that metric. However, our weighted total Z-Score calculation has
not changed.
Some Examples
Now that we’ve laid out the key metrics and described the changes to our report card,
let’s walk through some examples from Exhibit 47. We assigned scores to each industry
group and placed those with the highest exposure or lowest Z-Scores at the top of the
report card and those with the lowest exposure or highest Z-Scores at the bottom.
High Exposure
A Z-Score represents the number of standard deviations from the mean. It’s interesting
to note that only four industry groups received a Z-Score that was more than 1 standard
deviation below the mean, indicating that the highest exposure is concentrated among
them. Three of these industry groups should come as no surprise: Auto Components,
Automobiles, and Airlines; Aerospace and Defense companies are unique as they are
partially reimbursed for pension contributions that are associated with defense
department contracts. These industry groups stand out because, compared to others,
the degree of their pension plan underfunding is significant relative to market
capitalization, the size of their PBO relative to market capitalization exposes a
significant amount of leverage to their pension plan, the amount of their 2005 estimated
contribution to the pension plan is large in comparison to their cash flows, the
adjustment to operating income to better reflect pension cost is large, and their
shareholders’ equity would take a big hit as a result of recording the off-balance-sheet
pension liability. Of the 20 industry groups with a Z-Score below zero, 16 had a Z-Score
between 0 and (0.57). Although we wouldn’t necessarily categorize the exposure of
these 16 industry groups as insignificant, it’s certainly not as alarming as that of the
industry groups at the top of the list.
Low Exposure
The degree of exposure begins to drop off once we move beyond the top dozen or so
high-exposure groups as evidenced by the changing Z-Score, which rapidly begins to
approach 0 before becoming positive and moving into the low exposure group. We
found that 34 out of the 59 industry groups received a Z-Score of greater than zero
(highlighting their moderate-to-low exposure to their defined benefit pension plan),
including Software and Real Estate that appear to have the lowest exposure.
60
The Magic of Pension Accounting, Part III 07 February 2005
Note: Industry groups in this analysis include only companies with DB plans (i.e., analysis excludes industry groups that do not have companies with DB
plans, including Internet & Catalog Retail, Internet Software & Services, Trading Companies & Distributors, and Wireless Telecommunication Services).
This report card further excludes Office Electronics, which consists of only Xerox. We’ve also made additional adjustments. For example, we exclude
certain metrics for some industry groups (e.g., Contributions/CFO metric is not applied to Commercial Banks or Capital Markets because they either do
not report CFO or CFO was negative); additional weight is applied to remaining metrics in such cases.
Source: Company data, CSFB estimates.
61
The Magic of Pension Accounting, Part III 07 February 2005
For those of you not familiar with Rube Goldberg, Rube was known for taking something
relatively simple and turning it into something overly complicated; sounds a lot like
pension accounting. The cartoon in Exhibit 48 is a Rube Goldberg classic.
Passing man (A) slips on banana peel, (B) causing him to fall on rake (C). As handle of rake rises, it throws
horseshoe (D) onto rope, (E) which sags, thereby tilting sprinkling can (F). Water (G) saturates mop (H).
Pickle terrier (I) thinks it is raining, gets up to run into house and upsets sign (J), throwing it against nontipping
cigar ash receiver, (K) which causes it to swing back and forth and swish the mop against window pane,
wiping it clean. If man breaks his neck by fall, move away before cop arrives.
Source: Rube Goldberg is the ® and © of Rube Goldberg Inc.
In December 2003, the board took a small step toward fixing pension accounting by
providing investors with some new disclosures. Recently there have been some
rumblings on this topic again: an FASB board member indicated the board may consider
pensions in the not too distant future; pensions are on the short list of major projects
when it comes to convergence with the International Accounting Standards Board
(IASB); it’s among the risk-based initiatives for the SEC’s enforcement division; pension
reform is set to be a hot topic in Washington D.C. this year; and pensions will be
highlighted in the SEC’s upcoming report to Congress on off-balance-sheet activity.
With all that said, in our view it’s just a matter of time before the FASB takes a big step,
adding a major project on pension accounting to its agenda. In fact, we would not be
62
The Magic of Pension Accounting, Part III 07 February 2005
surprised to see it on the agenda before the end of this year. Especially now that
employee stock options have been taken care of.
However, after hearing company complaints about earnings volatility, we think the FASB
could once again come to a compromise, requiring companies to report the funded
status of the pension plan on the balance sheet as an asset for an overfunded plan and
as a liability for an underfunded plan (the net method). On the income statement, the
FASB could end up asking companies to run the service cost through operating income,
interest cost through interest expense and the actual returns on plan assets and
changes in the pension obligation due to changes in actuarial assumptions through
other comprehensive income as part of shareholders’ equity.
As the board may soon start work on pension accounting, we thought it was time to
update our analysis on what would happen to the earnings and balance sheets of the
companies in the S&P 500 if we eliminated the magic of pension accounting and instead
simply reflected the economics of the pension plans in the financial statements.
However, once past the intricate calculations, there are only two components to a
defined benefit pension plan: plan assets and an obligation. Think of a defined benefit
pension plan as an investment subsidiary with a collection of assets (equity, fixed
income, real estate, private equity, venture capital, timberlands, precious metals, etc.)
and an obligation (the promise to pay retirement benefits to the employees in the future,
similar to a series of zero coupon bonds). The most relevant information about this
investment subsidiary is the fair value of the assets (the market value of the stocks,
bonds, real estate, and other assets that are in the portfolio) and the fair value (present
value) of the pension obligation. We have reasonable estimates of each. If the assets
are greater than the obligation, the pension plan is overfunded. If the obligation is larger
than the assets, then the pension plan is underfunded. (See Exhibit 49.)
63
The Magic of Pension Accounting, Part III 07 February 2005
From an analytical perspective, and if we could rewrite the accounting rules, we would
treat the pension plan as an investment subsidiary by first reporting the pension plan
assets as an asset on the balance sheet and the pension obligation as a liability. We
would then report the changes in plan assets (actual returns) and the pension obligation
(service cost, interest cost, and actuarial gains and losses) in the appropriate place on
the income statement.
We applied this thought process to the companies in the S&P 500, eliminating the magic
of pension accounting and reflecting the underlying economics of the defined benefit
pension plan in the financial statements, by replacing the amounts reported under FAS
No. 87 with what we refer to as the “real” pension asset or liability and the “real” pension
expense or income. What follows in Exhibit 50 is the simple, three-step framework that
we used for each of the 370 companies in the S&P 500 with a defined benefit pension
plan.
Reported Adjusted
• Remove the amounts that are • Instead, put the pension plan assets
Step 1: currently reported on the balance on the asset side of the balance sheet
sheet for the pension plan. and the projected benefit obligation on
(Report the Pension Plan Assets & the liability side.
Obligation on the Balance Sheet)
• Take the expected return on plan • Replace them with the actual return on
Step 2: assets and amortization amounts and plan assets and gains and losses on the
throw them out the window. pension obligation. Continue to record
(Calculate “Real” Pension
interest cost and service cost to arrive at
Expense or Income)
the “real” pension expense or income.
64
The Magic of Pension Accounting, Part III 07 February 2005
• We estimate the off-balance-sheet pension liability for the S&P 500 was $1,137 billion
(yup, that’s over $1 trillion) at the end of 2003 and $964 billion for 2002. The off-
balance-sheet pension asset for the S&P 500 was $887 billion and $768 billion for
2003 and 2002, respectively.
• The 10 companies with the largest off-balance-sheet pension liabilities are included in
Exhibit 51, accounting for over 36% of the total for the S&P 500. It should be no
surprise that these same 10 companies also have the largest off-balance-sheet
pension assets, accounting for 34% of the total for the S&P 500.
65
The Magic of Pension Accounting, Part III 07 February 2005
Off-Balance-Sheet Off-Balance-Sheet
Pension Liability Pension Asset
Company Ticker 2003 2003
General Motors Corp GM $ 94,349 $ 52,841
Intl Business Machines Corp IBM 66,997 48,006
Ford Motor Co F 51,654 41,820
General Electric Co GE 36,000 26,841
Verizon Communications VZ 35,646 29,998
Boeing Co BA 32,164 23,975
Lucent Technologies Inc LU 28,736 25,473
SBC Communications Inc SBC 24,897 18,111
Lockheed Martin Corp LMT 23,264 19,183
Du Pont (E I) De Nemours DD 19,446 17,040
Source: Company data, CSFB estimates.
Many companies have more than one defined benefit pension plan. For example, a
company may have a different plan for each overseas subsidiary. Each pension plan is
accounted for separately. However, the information provided in the pension footnote is
for all a company’s pension plans combined. (Some companies do break down the data
between domestic and overseas plans.) To simplify our analysis, we treat the
information in the pension footnote as if it were all from one pension plan. The results
could be very different if we applied our methodology to each individual pension plan.
In the aggregate, placing the pension plan assets on the balance sheet would increase
total assets for the S&P 500 by 5% and 4%, respectively, for 2003 and 2002; treating
the PBO as debt would increase aggregate debt for the S&P 500 by 18% and 17%. The
difference between the increase in assets and the increase in liabilities would reduce
total equity for the S&P 500 by $163 billion, or 5%, at the end of 2003. Some specific
impacts include:
If the PBO is considered the equivalent of debt, debt increases by 10% or more for 294
companies. Debt would more than double for 84 companies, and more than triple for 43
companies, including the 24 companies in Exhibit 52 that had over $1 billion in debt
outstanding. The pension obligation is very similar to debt, except that the creditors are
the employees. In effect, a company is borrowing from its employees, which may not be
the most efficient way to raise capital.
66
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 52: Over 200% Increase in Total Debt by Treating PBO as Debt (>$1 Billion of Debt)
US$ in millions
Treating pension plan assets as an asset on the balance sheet would increase total
assets by over 10% for 158 companies, by over 25% for 53 companies, and by over
50% for the 14 companies in Exhibit 53. On an asset basis, some of these companies
look like equal part operating company and pension plan.
67
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 53: Over 50% Increase in Total Assets by Adding Pension Plan Assets
US$ in millions
If the pension plan’s funded status were reported on the balance sheet, shareholders’
equity would increase by 5% or more for nine companies. On the other hand, equity
would decline by 5% or more for 131 companies, including the 19 companies in Exhibit
54 where equity would drop by 25% or more. In other words, the shareholders’ equity
for these companies may be overstated by more than 25% due to pension accounting.
Exhibit 54: Reporting Pension Plan’s Funded Status on Balance Sheet Reduces Equity by
25% or More
US$ in millions
68
The Magic of Pension Accounting, Part III 07 February 2005
• Note that shareholders’ equity would fall by more than half for Campbell Soup,
Boeing, Delphi, and Ford, and would be wiped out for Maytag, Hercules, and General
Motors.
Why the dramatic disconnect between what’s on the balance sheet for the pension plan
and the real economics of the plan, its funded status? That’s the Magic of Pension
Accounting and all its wonderful smoothing mechanisms. In this case, it is primarily
unrecognized losses that result in GM’s assets being understated by $53 billion ($94
billion fair value of plan assets - $41 billion on balance sheet pension asset), its liabilities
understated by $94 billion ($102 billion projected benefit obligation - $8 billion on
balance sheet pension liability), and therefore its shareholders equity overstated by
$41.5 billion pretax, or $27 billion assuming a 35% tax rate.
69
The Magic of Pension Accounting, Part III 07 February 2005
• The debt-to-equity ratio would increase by over 1,000 basis points for 283 companies.
We isolate the 20 companies in Exhibit 56 that appear to have relatively little leverage
on a reported basis, with reported debt-to-equity ratios that are 50% or less. After
adjusting the balance sheet to reflect the economics of their pension plans, their debt-
to-equity ratios climb by over 6,000 basis points.
Exhibit 56: Reported Debt-to-Equity Ratio below 50% Climbs over 6,000 Basis Points
Debt-to-Equity Ratio Adj. Debt-to-Equity Ratio Difference
Company Ticker 2003 2003 (bps)
NCR Corp NCR 17% 476% 45,938
Rockwell Collins Inc COL 5% 452% 44,679
ITT Industries Inc ITT 33% 295% 26,287
United Technologies Corp UTX 45% 225% 17,927
3M Co MMM 38% 208% 17,010
Honeywell International Inc HON 48% 213% 16,465
Rockwell Automation ROK 49% 201% 15,248
FedEx Corp FDX 28% 158% 13,065
Northrop Grumman Corp NOC 36% 153% 11,667
Aon Corp AOC 47% 162% 11,525
Tektronix Inc TEK 14% 116% 10,191
SBC Communications Inc SBC 47% 135% 8,773
Brunswick Corp BC 46% 132% 8,580
Cigna Corp CI 33% 117% 8,384
Snap-On Inc SNA 33% 116% 8,304
Fluor Corp FLR 25% 107% 8,239
Agilent Technologies Inc A 41% 111% 7,046
United Parcel Service Inc UPS 26% 93% 6,703
Crane Co CR 50% 115% 6,450
Lilly (Eli) & Co LLY 50% 114% 6,382
Source: Company data, CSFB estimates.
Here we view an overfunded pension plan as an asset for the sponsoring company, one
that is difficult to monetize; however, it is still an asset. Capital that would have been
allocated to the pension plan can now be put to other uses: reinvestment, share
repurchases, debt paydown, etc. On the other hand, an underfunded plan can be
considered a liability that will result in increased future contributions to the pension plan
from the company, drawing capital away from other parts of the business. We run this
analysis because the current accounting views the pension plan on a net basis, many
investors view it this way as well, and this may be a compromise view for the FASB
when it starts working on pension accounting.
70
The Magic of Pension Accounting, Part III 07 February 2005
Under the net method, we removed whatever was reported on the balance sheet for the
pension plan and replaced it with the plan’s funded status. If the pension plan is
overfunded, we treat it as an asset; if it is underfunded, then it’s a liability. To balance
the balance sheet, the other side of each adjustment was reported in shareholders’
equity.
In the aggregate, placing only the funded status on the balance sheet would decrease
total assets for the S&P 500 by 1% while also increasing aggregate debt by 3% in both
2002 and 2003. The combination of a decrease in assets and an increase in liabilities
would reduce total equity for the S&P 500 by $163 billion, or 5%, at the end of 2003.
Notice the net effect on equity is the same as our first calculation. Some specific
impacts include:
• Using 2003 data, total assets would decrease by over 5% for 25 companies and by
over 10% for the five companies in Exhibit 57. On the other hand, total assets would
increase for just one company (FPL), and remain the same for 157 companies.
Exhibit 57: Decrease in Total Assets from Bringing the Funded Status onto the Balance
Sheet Under the “Net” Method
US$ in millions
Exhibit 58: Over 50% Increase in Total Debt by Adding Pension Underfunding (Shortfall)
under “Net” Method (>$1 Billion of Debt)
US$ in millions
71
The Magic of Pension Accounting, Part III 07 February 2005
Volatility is not necessarily a bad thing, unless it’s hidden. Pension accounting removes
the volatility in the pension plan from the financial statements, distorting both the
balance sheet and the income statement. The volatility that is inherent in the pension
plan is left for investors to discover when they read the pension footnotes. This makes
the process of understanding the economic status of the pension plan and the
company’s exposure to the plan that much more difficult. Volatility is a fact of doing
business; financial statements that don’t reflect that volatility are misleading.
Investors, market pundits, accountants, and Corporate America are all obsessed with
earnings and have strived to make earnings the ultimate measure of performance;
clearly, businesses are too complicated for that and so is the accounting. In particular,
some assets and liabilities are reported on the balance sheet at fair value today, with
the changes in fair value flowing through the income statement into earnings. To take a
mark-to-market result and place a multiple on it makes little sense, since you know the
fair value of the asset or liability; it is on the balance sheet. Instead, its valuation should
be taken into account separately in a valuation analysis. Then we should evaluate the
assumptions that management used to validate the valuation. If we want to isolate the
value of the pension plan, that value is best represented by the funded status.
72
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 59: Side-by-Side Comparison, Pension Accounting Cost Versus “Real” Pension
Cost—Aggregate S&P 500 Totals
US$ in millions
Net Pension Expense / (Income) $ 21,986 1 Net Pension Expense / (Income) $ 10,093
1
This $22 billion of pension expense reported in the 2003 excludes $6 billion of settlements, curtailments, and
other. After including settlements, curtailments, and other, total 2003 pension expense for all the companies in
the S&P 500 was $28 billion.
Source: Company data, CSFB estimates.
Applying this methodology to the companies in the S&P 500 would have increased
aggregate earnings (net income from continuing operations) by approximately 3% in
2003. Contrast that with the prior four years, when earnings would have dropped 67% in
2002, 67% in 2001, and 9% in 2000, while increasing 25% in 1999, as displayed in
Exhibit 60.
Exhibit 60: S&P 500 Reported Net Income versus Adjusted Net Income
US$ in billions
% Difference
25% $73 $81
$100 15%
$0 0%
(9%) 3%
($100) 1999 2000 2001 2002 2003 (15%)
($200) (30%)
($300) (45%)
($400) (60%)
(67%) (67%)
($500) (75%)
Extending this analysis to industry groups, the strong performance in pension plan
assets during 2003 would have caused earnings to increase for 29 of 59 industry
groups, including the seven industry groups in Exhibit 61 with an increase of more than
20%. Note the first two industry groups where earnings would more than double.
73
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 61: Reporting What Actually Happened to the Pension Plan Increases Net Income
by over 20% in 2003
US$ in millions
• Extending the analysis a bit further, 2003 earnings would increase for 149 companies.
Earnings would increase by more than 10% for 55 companies, more than 50% for 17
companies and more than double for the 11 companies in Exhibit 62.
Exhibit 62: Reporting What Actually Happened to the Pension Plan More than Doubles Net
Income in 2003
US$ in millions
Reported Net Income Adjusted Net Income Difference
Company Ticker 2003 2003 (%)
Williams Cos Inc WMB $ 15 $ 90 490%
NCR Corp NCR 58 325 461%
Boise Cascade Corp BCC 17 92 437%
Ford Motor Co F 921 3,779 310%
Goodrich Corp GR 39 106 176%
Northrop Grumman Corp NOC 808 2,029 151%
Eastman Kodak Co EK 238 520 119%
Weyerhaeuser Co WY 288 615 114%
Verizon Communications VZ 3,509 7,297 108%
General Motors Corp GM 2,862 5,819 103%
Transocean Inc RIG 18 37 100%
Source: Company data, CSFB estimates.
• On the other hand, earnings would fall for 143 companies as a result of reporting what
actually happened to the pension plan through earnings, including the six companies
in Exhibit 63 where net income would swing to a net loss.
Exhibit 63: Reporting What Actually Happened to the Pension Plan Turns Net Income into
Net Loss in 2003
US$ in millions
Reported Adjusted
Net Income Net Loss Dollar
Company Ticker 2003 2003 Decline
Boeing Co BA $ 718 $ (158) $ (876)
FedEx Corp FDX 830 (87) (917)
Parker-Hannifin Corp PH 196 (83) (279)
Penney (J C) Co JCP 371 (3) (374)
Rockwell Collins Inc COL 258 (54) (312)
Tektronix Inc TEK 35 (71) (106)
Source: Company data, CSFB estimates.
74
The Magic of Pension Accounting, Part III 07 February 2005
Cumulative 1999-2003
In addition to looking at 2003, we also ran the analysis on a cumulative basis over the
past five years. In the aggregate, net income for the S&P 500 in 1999-2003 would drop
by approximately 15%. We found that 356 companies would have experienced a decline
in net income, including 119 companies with a more than 15% drop in net income and
the 13 companies on the left hand side of Exhibit 64 where cumulative net income
would swing to a net loss. In addition, on the right hand side of Exhibit 64, we include
the 13 companies with the largest dollar decline in net income by removing the magic of
pension accounting and replacing it with what actually happened to the pension plan.
Note that Lockheed Martin and Raytheon are the only two companies to make both lists.
Exhibit 64: Reporting What Actually Happened to the Pension Plan Produces Large Decline in Cumulative Net Income in
1999–2003
US$ in millions
Note that net income would increase for only 14 companies, led by Avaya, McGraw Hill,
Synovus, and Safeco.
• The ROE declines for 130 companies. A positive ROE swings to a negative ROE for
the six companies in Exhibit 65.
75
The Magic of Pension Accounting, Part III 07 February 2005
The ROE increases for 234 companies, including the 12 companies in Exhibit 66 with
an increase in ROE exceeding 2,000 basis points.
76
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 67: S&P 500 Pension Accounting Cost versus “Real” Pension Cost
US$ in billions
$200
Pension (Expense)/Income
$0
Expected Returns
One of the main drivers as to why what is reported as pension cost in the income
statement can vary so dramatically from what actually happened to the pension plan is
the difference between the actual return and the expected return. The actual return on
pension plan assets is not recognized in accounting for pension cost; instead, it is
replaced by the expected return on plan assets. The expected return on plan assets is
calculated by multiplying the company’s expected rate of return assumption by
something called the market-related value of plan assets. The market-related value can
be either the fair value of plan assets or a calculated value of plan assets. (A five-year
moving average of the fair value of the plan assets would be the best analogy.)
Using the expected return as opposed to the actual return is the equivalent of your
depositing a paycheck for what you think you should get paid instead of what you were
actually paid. If you have ever tried doing this, you know that it does not work so well. Or
another way of thinking about it, it’s like replacing revenue on the income statement with
budgeted revenue. In other words, “We only sold 50 units this quarter; however, we
budgeted for 100, so let’s recognize revenue as if we sold 100 units.”
The difference between the actual return on plan assets and the expected return, which
we like to refer to as the “unexpected return,” is moved off balance sheet, along with
gains and losses on the projected benefit obligation, where they accumulate until they
get so big they have to be recognized. (See our first The Magic of Pension Accounting,
of September 27, 2002, for further discussion on the amortization of unrecognized gains
and losses.) We would instead simply report the actual return on the income statement.
77
The Magic of Pension Accounting, Part III 07 February 2005
Unexpected Returns
The actual return on defined benefit pension plan assets for the S&P 500 was 18.80%,
or $180 billion, during 2003. Performance was strong across the board, as 325
companies experienced a positive return on plan assets while only 34 had a loss
However, through the magic of pension accounting, S&P 500 companies collectively
recognized only $97 billion of the $180 billion in actual returns generated during 2003 in
the form of the expected return on plan assets. In other words, the companies expected
to see their pension plan assets increase by $97 billion. Instead, they actually increased
by $180 billion, creating an unexpected gain of $83 billion. Remember, only the
expected return is recognized in earnings, the unexpected gain or loss is accumulated
and may eventually get amortized to earnings over time.
Nearly the exact opposite result occurred in 2002, as pension plans were hit with poor
performance; the actual loss on plan assets was 8.85%, or $93 billion (338 companies
had a loss and only 16 had a gain). However, S&P 500 companies collectively
recognized $103 billion of income during 2002 in the form of the expected return on plan
assets. In other words, the companies expected to see their pension plan assets
increase by $103 billion. Instead, they actually declined by $93 billion, resulting in an
unexpected loss of $196 billion.
• Three hundred two companies had an unexpected gain in 2003. In other words, their
actual return on plan assets was greater than their expected return on plan assets.
The unexpected gain was over $500 million for 40 companies, including 17
companies that reported unexpected gains of over $1 billion. Fifty-six companies had
an unexpected loss. We list the 10 largest unexpected gains and losses in 2003 in
Exhibit 68.
Cumulatively from 1999 through 2003, the actual return on pension plan assets for the
companies in the S&P 500 was $223 billion versus an expected return of $485 billion,
generating an unexpected loss of $262 billion. Exhibit 69 compares the aggregate
actual and expected returns for the S&P 500 over the past five years.
78
The Magic of Pension Accounting, Part III 07 February 2005
$200 $179
$165
$150
$97 $103 $103 $97
$100 $86 $79 $83
Return Amount
$58
$50
$0
1999 2000 2001 2002 2003
($50) ($39)
In Exhibit 70, we show the relationship between expected and actual returns for the
pension plans of the companies in the S&P 500 during the cumulative period from 1999
to 2003. Each point in Exhibit 70 represents a different S&P 500 company.
The diagonal line running through the exhibit represents points at which the expected
return and actual return are equal. As you can plainly see not that many pension plans
outperformed their expected return from 1999 through 2003, in fact only 15 companies
were able to make this claim. On the other hand the actual return on pension plan
assets fell short of the expected return for 349 companies in the S&P 500.
A few outliers make Exhibit 70 difficult to read, strip them out and you get Exhibit 71.
Exhibit 70: A High Level View of the Difference between Exhibit 71: . . . and a Fresh Look after Stripping
Expected Returns versus Actual Returns across Companies away the Outliers
in the S&P 500 for the Cumulative Period 1999–2003 . . . US$ in millions, unless otherwise stated
US$ in millions, unless otherwise stated
$40,000 $5,000
$35,000
$4,000 EK
$30,000
Actual Return ($)
$25,000
Outperform $3,000 Outperform
$20,000 GM
$2,000
$15,000
LU VZ
IBM
$10,000 F LTR
SBC GE $1,000
T
LMT
BA
$5,000 Underperform Underperform
BLS
DD
$0 $0
$0 $5,000 $10,000 $15,000 $20,000 $25,000 $30,000 $35,000 $40,000 $0 $1,000 $2,000 $3,000 $4,000 $5,000
Source: Company data, CSFB estimates. Source: Company data, CSFB estimates.
79
The Magic of Pension Accounting, Part III 07 February 2005
• For 54 of the companies, the cumulative unexpected loss was greater than $1 billion,
including the 14 companies in Exhibit 72 where it was over $3 billion.
Exhibit 72: Actual Return < Expected Return for the Cumulative Period 1999–2003
US$ in millions
Exhibit 73: Actual Return > Expected Return for the Cumulative Period 1999-2003
US$ in millions
80
The Magic of Pension Accounting, Part III 07 February 2005
The only item that is truly a compensation cost is service cost; it is the value of the
future retirement benefits that the employees earned by working during the current year.
Think of it as a form of deferred compensation. We record it wherever the company
reports labor cost. It is the only item of pension cost that we report as an operating cost.
Therefore, working with either our concept of real pension cost or with net pension cost
per FAS No. 87, the impact on operating income is the same.
We divide up the remaining items and put them somewhere else on the income
statement. Interest cost can be thought of as a financing cost, and we record it as part
of interest expense. Interest cost is the increase in the benefit obligation due to the
passage of time; it is similar in concept to the accretion on a zero coupon bond. The
actual return on plan assets results from changes in the market prices of the stocks,
bonds, real estate, and whatever else is in the pension portfolio. The actuarial gain or
loss on the PBO is also a mark-to-market concept; here the changes in value are
generally due to changes in actuarial assumptions. We record them both wherever the
company reports investment gains and losses (i.e., other income, etc.).
The FASB may help us here, as the board has a joint project with the IASB on its
agenda on financial performance reporting. One of the topics of discussion is likely to be
pension accounting, and the location of pension cost on the income statement could be
made clear, possibly being split apart as we propose.
81
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 75: S&P 500 Aggregate Expected Returns as Percentage of Operating Income,
1999–2003
8%
6%
4%
2%
0%
1999 2000 2001 2002 2003
For specific companies, the expected return appears even more significant. It accounted
for over 10% of operating income for 153 companies in 2003, and over 25% for 65
companies, including the 30 companies in Exhibit 76 where the expected return on plan
assets was more than 50% of the companies’ operating income in 2003.
82
The Magic of Pension Accounting, Part III 07 February 2005
Expected Operating
Return Income
Company Ticker 2003 2003 Percentage
Rowan Cos Inc RDC $ 17 $2 793%
Navistar International NAV 232 65 357%
Avaya Inc AV 214 69 310%
NCR Corp NCR 331 131 253%
Boeing Co BA 3,403 1,748 195%
Cummins Inc CMI 157 93 169%
Meadwestvaco Corp MWV 303 199 152%
Qwest Communication Intl Inc Q 858 582 147%
Goodyear Tire & Rubber Co GT 311 253 123%
Unisys Corp UIS 501 428 117%
Dana Corp DCN 216 199 109%
Tektronix Inc TEK 55 57 96%
Delphi Corp DPH 647 676 96%
Lockheed Martin Corp LMT 1,748 1,976 88%
Northrop Grumman Corp NOC 1,195 1,538 78%
Pactiv Corp PTV 354 467 76%
Du Pont (E I) De Nemours DD 1,368 1,811 76%
Eastman Kodak Co EK 759 1,018 75%
Raytheon Co RTN 970 1,316 74%
General Motors Corp GM 6,947 10,165 68%
Electronic Data Systems Corp EDS 339 533 64%
Goodrich Corp GR 182 296 61%
Intl Business Machines Corp IBM 5,931 10,098 59%
Boise Cascade Corp BCC 102 175 59%
ITT Industries Inc ITT 327 559 58%
Ford Motor Co F 4,584 8,116 56%
Donnelley (R R) & Sons Co DNY 157 281 56%
Dow Chemical DOW 1,082 2,019 54%
Textron Inc TXT 432 822 53%
Rockwell Collins Inc COL 166 335 50%
Source: Company data, CSFB estimates.
If we strip out all the components of pension cost and leave behind service cost,
operating income for the S&P 500 would remain flat in 2003, while declining by 3% in
2002, and 4% in 2001. The five industry groups in Exhibit 77 would experience a decline
in operating income of 5% or more in 2003. We learn from this analysis that not all
operating income is truly “operating.”
83
The Magic of Pension Accounting, Part III 07 February 2005
Operating Adjusted
Income Operating Income
Industry 2003 2003 % Decline
Communications Equipment $ 6,589 $ 5,350 (18.8%)
Computers & Peripherals 17,682 15,979 (9.6%)
Aerospace & Defense 14,897 13,839 (7.1%)
Industrial Conglomerates 39,383 37,076 (5.9%)
Leisure Equipment & Products 2,412 2,289 (5.1%)
Source: Company data, CSFB estimates.
On the other hand, the seven industry groups in Exhibit 78 would experience an
increase in operating income of 5% or more in 2003.
Operating Adjusted
Income Operating Income
Industry 2003 2003 % Increase
Auto Components $ 2,092 $ 2,802 33.9%
Metals & Mining 3,278 3,730 13.8%
Automobiles 19,448 21,588 11.0%
Computers & Peripherals 17,682 15,979 9.6%
Office Electronics 1,801 1,968 9.3%
Construction & Engineering 272 290 6.7%
Electronic Equipment & Instruments 733 781 6.5%
Source: Company data, CSFB estimates.
84
The Magic of Pension Accounting, Part III 07 February 2005
A B C D=A+B-C
Reported Pension Service Adjusted %
Company Ticker Operating Inc. Cost/(Benefit) Cost Operating Inc. Difference
Avaya Inc AV $ 69 $ (8) $ 53 $8 (88%)
Meadwestvaco Corp MWV 199 (64) 71 64 (68%)
Qwest Communication Intl Inc Q 582 (149) 174 259 (55%)
Boeing Co BA 1,748 (67) 753 928 (47%)
Unisys Corp UIS 428 (23) 100 305 (29%)
Pactiv Corp PTV 467 (64) 33 370 (21%)
Intl Business Machines Corp IBM 10,098 (803) 1,113 8,182 (19%)
Tektronix Inc TEK 57 (0) 10 47 (18%)
Consolidated Edison Inc ED 1,418 (157) 102 1,159 (18%)
Donnelley (R R) & Sons Co DNY 281 (2) 48 230 (18%)
Textron Inc TXT 822 (32) 105 685 (17%)
Eastman Kodak Co EK 1,018 (8) 157 853 (16%)
Dana Corp DCN 199 26 53 172 (14%)
Dow Chemical DOW 2,019 (28) 242 1,749 (13%)
Xcel Energy Inc XEL 1,104 (70) 67 967 (12%)
Prudential Financial Inc PRU 2,852 (170) 180 2,502 (12%)
FPL Group Inc FPL 1,531 (123) 51 1,357 (11%)
Bellsouth Corp BLS 6,115 (486) 181 5,448 (11%)
SBC Communications Inc SBC 6,469 89 732 5,826 (10%)
FedEx Corp FDX 1,471 228 374 1,325 (10%)
Source: Company data, CSFB estimates.
• At the other end of the spectrum, operating income would increase by 5% or more for
48 companies. The 19 companies in Exhibit 80 would experience an increase of over
10%. Note that for the three companies at the top of the list— Rowan, Navistar, and
Goodyear—operating income would more than double.
85
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 81: CAGR Decline of over 200 Basis Points Attributable to Adjustments
Reported 5-Year Adjusted 5-Year
Company Ticker Operating Income CAGR Operating Income CAGR Change (bps)
Meadwestvaco Corp MWV (8%) (21%) (1,379)
Qwest Communication Intl Inc Q (29%) (38%) (948)
Unisys Corp UIS (12%) (16%) (353)
AT&T Corp T (18%) (21%) (270)
Synovus Financial Cp SNV 24% 21% (237)
Tektronix Inc TEK (22%) (24%) (219)
Dow Chemical DOW (2%) (4%) (209)
Source: Company data, CSFB estimates.
The growth rate improves for 249 companies, including the eight companies in Exhibit
82 where the growth rate improves by over 500 basis points.
86
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 82: CAGR Increase of over 500 Basis Points Attributable to Adjustments
Reported 5-Year Adjusted 5-Year
Company Ticker Operating Income CAGR Operating Income CAGR Change (bps)
NCR Corp NCR (3%) 16% 1,875
Rowan Cos Inc RDC (59%) (41%) 1,742
Northrop Grumman Corp NOC 10% 27% 1,685
Delphi Corp DPH 5% 19% 1,368
Goodyear Tire & Rubber Co GT (25%) (13%) 1,203
Navistar International NAV (34%) (23%) 1,111
Boise Cascade Corp BCC (4%) 2% 598
Georgia-Pacific Corp GP 16% 21% 517
Source: Company data, CSFB estimates.
Cumulative 1999-2003
In addition to looking at 2003, we ran the analysis on a cumulative basis over the past
five years. In the aggregate, operating income for the S&P 500 in 1999-2003 would drop
by approximately $119 billion, or 3%. We found that 250 companies would experience a
decline in operating income. In Exhibit 83, we list the 10 companies that would have the
largest percent decrease in operating income alongside the 10 companies with the
largest dollar decrease in operating income.
Exhibit 83: Largest Increases and Decreases in Operating Income from Adjustments during the Cumulative Period 1999–2003
Largest Cumulative % Decrease Largest Cumulative $ Decrease
Reported Adjusted Reported Adjusted
Operating Operating Operating Operating
Income Income 1999- Income Income 1999-
Company Ticker 1999-2003 2003 % Decrease Company Ticker 1999-2003 2003 $ Decrease
Allegheny Technologies Inc ATI $ 334 $ 17 (95%) General Electric Co GE $ 144,084 $ 131,592 $ (12,492)
Delta Air Lines Inc DAL 678 192 (72%) Intl Business Machines Corp IBM 49,718 38,842 (10,876)
NCR Corp NCR 1,048 374 (64%) Verizon Communications VZ 67,905 57,971 (9,934)
PG&E Corp PCG 2,442 1,334 (45%) SBC Communications Inc SBC 49,353 41,830 (7,523)
Meadwestvaco Corp MWV 1,744 1,041 (40%) Boeing Co BA 17,107 11,829 (5,278)
Lockheed Martin Corp LMT 8,920 5,690 (36%) Bellsouth Corp BLS 33,533 29,370 (4,163)
Unisys Corp UIS 2,635 1,723 (35%) Lucent Technologies Inc LU (2,675) (6,538) (3,863)
Northrop Grumman Corp NOC 6,000 3,955 (34%) Lockheed Martin Corp LMT 8,920 5,690 (3,230)
Prudential Financial Inc PRU 7,221 4,865 (33%) Ford Motor Co F 58,738 55,669 (3,069)
Pactiv Corp PTV 2,022 1,375 (32%) AT&T Corp T 39,611 36,643 (2,968)
Source: Company data, CSFB estimates.
Operating Margins
• The impact on operating margins for the entire S&P 500 is minimal. However,
operating margins would drop by over 100 basis points for 28 companies. The 17
companies in Exhibit 84 would experience a decline of 150 basis points or more in
2003. Only two companies, Lucent and Pactiv, would have had a more than 300-
basis-point decline for each of the past four years.
87
The Magic of Pension Accounting, Part III 07 February 2005
• Operating margins would improve for 218 companies, including the eight companies
in Exhibit 85 with improvements of more than 150 basis points in 2003.
Now a 150-basis-point improvement in operating margins may not sound like a lot;
however, it depends upon your starting point. For example, take a look at Goodyear with
reported operating margins of 2%: a 192-basis-point increase nearly doubles the
operating margins to 4%.
EV/EBITDA Analysis
Extending the analysis a bit further, we take a look at EV/EBITDA multiples. We adjust
the EBITDA for the last fiscal year in a similar manner to operating income, removing
pension cost and replacing it with service cost. We also adjust enterprise value by
treating the funded status of a defined benefit pension plan as net debt (enterprise value
is as of the fiscal year-end). We then compare EV/EBITDA multiples on a reported basis
with our adjusted multiples. The multiple for the S&P 500 using 2003 historical results
would rise from about 11.48 to 11.62 times, an increase of approximately 1%.
88
The Magic of Pension Accounting, Part III 07 February 2005
• The EV/EBITDA multiple would increase by over 10% for the three industry groups in
Exhibit 86. The implication of this adjustment is simple: Investors may not always be
getting what they think they’re paying for.
Exhibit 86: Industry EV/EBITDA Multiples Rising over 10% Due to Adjustments
Industry EV / EBITDA Adjusted EV / EBITDA % Change in Multiple
Aerospace & Defense 11.03x 12.65x 15%
Auto Components 6.00x 6.71x 12%
Communications Equipment 17.31x 19.36x 12%
Source: Company data, CSFB estimates.
• The EV/EBITDA multiple would increase by 10% or more for 29 companies, and by
20% or more for the nine companies in Exhibit 87.
Exhibit 87: Company EV/EBITDA Multiples Rising over 20% Due to Adjustments
Company Ticker EV / EBITDA Adjusted EV / EBITDA % Change in Multiple
Circuit City Str Crct Cty Gp CC 0.20x 0.36x 80%
Avaya Inc AV 14.51x 21.73x 50%
Boeing Co BA 11.76x 17.43x 48%
Delphi Corp DPH 4.32x 5.77x 34%
Unisys Corp UIS 6.82x 9.06x 33%
Pactiv Corp PTV 7.86x 10.29x 31%
Tektronix Inc TEK 14.34x 18.75x 31%
Aon Corp AOC 3.40x 4.16x 22%
Rockwell Collins Inc COL 9.79x 11.94x 22%
Source: Company data, CSFB estimates.
Next we compare the EV/EBITDA multiple for each company with the multiple for its
industry group. We split the companies into two camps, those with premium multiples
versus the group and those with discounted multiples versus the group. We then split
the companies into the same two camps using our adjusted EV/EBITDA multiple. We
compare the two sets of results and find only 20 companies that switched camps.
• The EV/EBITDA multiple went from less than the industry average to greater than the
industry average for eight companies (moving from less expensive then the group on
a reported basis to more expensive than the group after making our adjustments) as
seen in Exhibit 88.
Exhibit 88: Companies Trading below Group Multiple That Trade Higher after Adjustment
Co. EV / Group Adj. Co. EV / Adj. Group
Company Ticker Industry Group EBITDA Multiple EBITDA Multiple
Avaya Inc AV Communications Equip. 14.51x 17.31x 21.73x 19.36x
Black & Decker Corp BDK Household Durables 7.55x 7.79x 8.40x 7.99x
Brown-Forman -Cl B BF.B Beverages 12.38x 12.83x 13.35x 12.99x
Donnelley (R R) & Sons DNY Commercial Serv. & Suppl. 7.18x 7.82x 8.01x 7.90x
Exelon Corp EXC Electric Utilities 7.45x 7.50x 7.87x 7.73x
Meadwestvaco Corp MWV Paper & Forest Products 9.40x 9.41x 10.09x 9.73x
Phelps Dodge Corp PD Metals & Mining 12.43x 13.34x 13.00x 12.96x
Unisys Corp UIS IT Services 6.82x 8.11x 9.06x 8.40x
Source: Company data, CSFB estimates.
89
The Magic of Pension Accounting, Part III 07 February 2005
• The EV/EBITDA multiple went from more than the industry average to less than the
industry average for 12 companies (moving from more expensive than the group on a
reported basis to less expensive than the group) as seen in Exhibit 89.
Exhibit 89: Companies Trading above Group Multiple That Trade below after Adjustment
Co. EV / Group Adj. Co. EV / Adj. Group
Company Ticker Industry Group EBITDA Multiple EBITDA Multiple
Centex Corp CTX Household Durables 7.91x 7.79x 7.91x 7.99x
Cisco Systems Inc CSCO Communications Equip. 19.24x 17.31x 19.24x 19.36x
CMS Energy Corp CMS Electric Utilities 7.55x 7.50x 7.11x 7.73x
Cooper Industries Ltd CBE Electrical Equipment 11.96x 11.82x 11.89x 12.20x
Cooper Tire & Rubber CTB Auto Components 6.36x 6.00x 6.54x 6.71x
Gannett Co GCI Media 12.73x 12.68x 12.60x 12.73x
Goodrich Corp GR Aerospace & Defense 12.01x 11.03x 12.07x 12.65x
Honeywell International HON Aerospace & Defense 11.10x 11.03x 11.63x 12.65x
Qualcomm Inc QCOM Communications Equip. 17.78x 17.31x 17.78x 19.36x
Sungard Data Systems SDS IT Services 8.25x 8.11x 8.25x 8.40x
United States Steel Corp X Metals & Mining 20.05x 13.34x 8.40x 12.96x
Visteon Corp VC Auto Components 6.43x 6.00x 5.55x 6.71x
Source: Company data, CSFB estimates.
Summary
Removing the smoothing mechanisms and stripping out what a company believes
should have happened with the pension plan and replacing it with what actually did
happen better reflects the company’s true exposure to its pension plan and the
underlying volatility in the financial statements. Following this methodology for the entire
S&P 500 would increase earnings by 3% in 2003, while reducing earnings by 67% in
both 2002 and 2001; this methodology would also reduce earnings by 9% in 2000 and
increase earnings by 25% in 1999. In the aggregate, placing the pension plan assets on
the balance sheet would increase total assets for the S&P 500 by 5% and 4%,
respectively, in 2003 and 2002; treating the PBO as debt would increase aggregate debt
for the S&P 500 by 18% and 17%. The difference between the increase in assets and
the increase in liabilities would reduce total equity for the S&P 500 by $163 billion, or
5%, at the end of 2003.
90
The Magic of Pension Accounting, Part III 07 February 2005
Analysis of Assumptions
With a defined benefit pension plan, a company has entered into an obligation where
the future cash flows to its retirees—the pension benefits that it will pay in the future—
are unknown. Therefore, a company and its actuaries must make a number of
assumptions about the future, including the discount rate and salary inflation
assumptions. When applying pension accounting, arguably the most controversial of the
assumptions is the expected rate of return on the pension plan assets. In addition to the
three assumptions that you typically see disclosed, there are lots of other important
pension assumptions, including mortality, employee turnover, and retirement age.
Overview
With a heightened level of investor interest in this topic as a result of the SEC inquiry
into pension accounting, we decided to take another close look at the pension
assumptions for the companies in the S&P 500, including an entire section on whether
or not the assumptions make sense. (Note that for companies that provide a range of
assumptions we use the median. For those that provide a different set of assumptions
for their domestic and international pension plans we calculate a weighted average.)
In 2003, we saw the median pension assumptions of companies in the S&P 500 move
to levels not seen in the last 10 years. For instance, the median expected rate of return
dropped 81 basis points from its peak of 9.31% in 2000 to 8.5% in 2003, its lowest level
of the decade. Meanwhile, the median discount rate hit 6.1%, also a 10-year low.
Finally, the median salary inflation rate, which hovered around 5% in the early 1990s,
fell to 4% in 2003. (See Exhibit 90.)
1
Exhibit 90: S&P 500 Pension Assumptions
Description 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
Expected Rate of Return (%)
Median 9.00 9.00 9.00 9.00 9.10 9.25 9.25 9.31 9.24 8.80 8.50
High 13.00 13.00 11.50 11.50 11.50 11.50 11.50 12.00 12.00 12.00 13.07
Low 5.00 5.00 5.00 3.00 3.00 4.00 6.00 5.75 5.65 3.35 3.75
Discount Rate (%)
Median 7.50 8.25 7.50 7.50 7.25 6.75 7.50 7.50 7.25 6.75 6.10
High 11.00 9.38 9.00 8.50 8.30 11.00 8.25 8.50 12.00 12.29 10.84
Low 5.75 6.00 6.10 4.50 4.50 5.00 4.75 4.75 4.65 4.30 3.63
Salary Inflation Rate (%)
Median 5.00 5.00 5.00 5.00 4.75 4.50 4.53 4.50 4.50 4.15 4.00
High 9.00 8.00 8.00 8.00 8.00 8.00 8.00 7.50 9.00 11.55 8.49
Low 2.90 3.00 3.00 2.33 2.00 2.40 1.70 1.70 1.18 1.70 1.50
1
Note that the highs and lows for each assumption may be substantially different from the medians because of companies with international plans, which
may provide ranges of assumptions that are sometimes materially different from U.S. plan ranges.
Source: Company data, CSFB estimates.
91
The Magic of Pension Accounting, Part III 07 February 2005
below those market levels, and the mix of plan assets. As a starting point, the SEC
indicated that companies should use historical returns of similarly allocated portfolios. At
the 30th Annual AICPA National Conference on Current SEC Developments in
December 2002, the SEC staff made clear the range in which it believes expected
return assumptions should fall by highlighting a study from 1926 through the third
quarter of 2002 that indicated 10% and 6% historical average annual returns on equity
and fixed income portfolios, respectively. It was at this conference that the SEC staff
indicated it could challenge anything over 9%.
The expected rate of return is used to calculate the expected return component of
pension cost, a surrogate for the actual return on the pension plan assets. The expected
return represents the product of a company’s expected rate of return and the market-
related value of the company’s plan assets at the beginning of the period. Market-
related value can either be the fair value of the plan assets or a calculated value that
recognizes changes in fair value over a period no greater than five years (e.g., a five-
year moving average). An increase in the expected return will lower pension cost,
increasing earnings, while a decrease in the expected return will increase pension cost,
reducing earnings. That is why companies are loath to reduce the expected rate of
return.
Companies have a significant amount of discretion when setting their expected rates of
return. The highest rate of return used for 2003 belonged to AES at 13.07%. KLA-
Tencor held the distinction of having the lowest rate at 3.75%. Exhibit 91 shows the
distribution of expected rates of return across the companies in the S&P 500 during
2003. Even as companies have ratcheted down their expected return assumptions over
the past few years, 273, or almost 75% of the companies with defined benefit plans in
the S&P 500, still use expected rates of return of 8% or higher. The median expected
rate of return was 8.50% in 2003, down from 8.80% in 2002 and 9.24% in 2001.
Exhibit 91: Distribution of Expected Rates of Return across S&P 500 Companies, 2003
Conservative Aggressive
240
206
200
Number of Companies
160
120
80 63 61
40 16
10 5 1
-
Below 6 6.00-6.99 7.00-7.99 8.00-8.99 9.00-9.99 10.00-10.99 11.00 or Above
92
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 92: Median Expected Rate of Return Assumptions across Industry Groups, 2003 (%)
Industry % Industry % Industry %
Air Freight & Logistics 9.21 Commercial Banks 8.55 Oil & Gas 8.14
Airlines 9.00 Consumer Finance 8.50 Personal Products 8.11
Multiline Retail 9.00 Containers & Packaging 8.50 Diversified Financial Services 8.10
Multi-Utilities & Unregulated Power 9.00 Diversified Telecom Services 8.50 Capital Markets 8.00
Wireless Telecom Services 9.00 Machinery 8.50 Specialty Retail 8.00
Distributors 8.95 Paper & Forest Products 8.50 Energy Equipment & Services 7.96
Electric Utilities 8.90 Pharmaceuticals 8.50 Household Products 7.90
Aerospace & Defense 8.75 Industrial Conglomerates 8.48 Building Products 7.84
Commercial Services & Supplies 8.75 IT Services 8.48 Hotels Restaurants & Leisure 7.75
Food & Staples Retailing 8.75 Chemicals 8.46 Real Estate 7.75
Gas Utilities 8.75 Health Care Providers & Services 8.38 Electronic Equip. & Instruments 7.58
Metals & Mining 8.75 Media 8.38 Thrifts & Mortgage Finance 7.50
Road & Rail 8.70 Textiles Apparel & Luxury Goods 8.33 Software 6.62
Tobacco 8.69 Beverages 8.30 Semiconductors & Semiconductor Equip. 6.60
Automobiles 8.64 Office Electronics 8.30 Construction & Engineering 6.50
Communications Equipment 8.63 Household Durables 8.30 Biotechnology 5.86
Leisure Equipment & Products 8.60 Construction Materials 8.25 Internet & Catalog Retail N/A
Food Products 8.59 Insurance 8.25 Internet Software & Services N/A
Electrical Equipment 8.58 Computers & Peripherals 8.20 Trading Companies & Distributors N/A
Auto Components 8.56 Health Care Equipment & Supplies 8.20
Source: Company data, CSFB estimates.
Companies have lots of discretion when it comes to setting these expected rates of
return. We can see that clearly when we look at the range of assumptions within an
industry group. The 14 industry groups in Exhibit 93 had a spread of over 225 basis
points between the highest and lowest expected rates of return. For each group, we
show the company that used the highest expected rate of return and the one with the
lowest. For example, General Mills used a 10.4% expected rate of return, while Sara
Lee used 7.2%; both companies are in the Food Products industry group. General Mills
has since ratcheted down its expected return assumption to 9.6%.
93
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 93: Widest Spread in 2003 Expected Rate of Return Assumptions (%)
Highest Expected Return Lowest Expected Return
Industry Median Range Company Ticker % Company Ticker %
Metals & Mining 8.75 6.00 Freeport McMoran Cop & Gld FCX 10.00 Worthington Industries WOR 4.00
Multi-Utilities & Unregulated Power 9.00 4.66 AES Corp. (The) AES 13.07 Duke Energy Corp DUK 8.41
Semiconductors & Semiconductor Equip 6.60 4.25 Teradyne Inc TER 8.00 KLA-Tencor Corp KLAC 3.75
Capital Markets 8.00 3.75 Northern Trust Corp NTRS 8.75 Merrill Lynch & Co MER 5.00
Machinery 8.50 3.64 Navistar International NAV 9.00 Pall Corp PLL 5.36
Health Care Equipment & Supplies 8.20 3.51 Baxter International Inc BAX 8.81 Stryker Corp SYK 5.30
Communications Equipment 8.63 3.28 Scientific-Atlanta Inc SFA 10.00 Andrew Corp ANDW 6.72
Food Products 8.59 3.20 General Mills Inc GIS 10.40 Sara Lee Corp SLE 7.20
Hotels Restaurants & Leisure 7.75 3.15 Darden Restaurants Inc DRI 10.40 Hilton Hotels Corp HLT 7.25
Media 8.38 2.64 Knight-Ridder Inc KRI 9.00 Interpublic Group Of Cos IPG 6.36
Energy Equipment & Services 7.96 2.50 Rowan Cos Inc RDC 9.00 Nabors Industries Ltd NBR 6.50
Oil & Gas 8.14 2.50 Ashland Inc ASH 9.00 Apache Corp APA 6.50
Insurance 8.25 2.43 Hartford Finl Svcs Grp Inc HIG 9.00 AFLAC Inc AFL 6.57
Specialty Retail 8.00 2.25 Boise Cascade Corp BCC 8.50 Office Depot Inc ODP 6.25
Source: Company data, CSFB estimates.
Exhibit 94: Number of Companies Reporting Expected Rates of Return above or below 9%
344
350
259
250
# of Companies
200 185
160 161
147
150 132 136 134
126 130
119 118 122 118
109
100 101 102 101
95
100
50
17
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
The number of companies using expected rates of return above 9% hit a peak in 2001
at 185 companies; that has since fallen off a cliff with only 17 companies using expected
rates of return over 9% in 2003. However, 50 companies still use a 9% expected rate of
return, and nearly 75% of the companies with pension plans in the S&P 500 use an
expected rate of return above 8%.
94
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 95 includes the 17 companies that applied expected rates of return above 9% in
2003, with six companies—AES, Darden, General Mills, FedEx, Freeport McMoran, and
Scientific Atlanta—above 10%. Note that the majority of the AES pension plan is in
Brazil and that Freeport McMoran Copper & Gold has an Indonesian defined benefit
pension plan covering most of its Indonesian workforce. As a result, their expected rates
of return are higher than those of other companies included in our analysis because the
Brazilian and Indonesian plans have significantly different risk profiles.
Of the 17 companies on this list, 10 have since announced a reduction in their expected
rates of return: Darden Restaurants, General Mills, FedEx, Scientific Atlanta, Bemis,
Kroger, Supervalu, Xcel Energy, United Parcel Service, and Air Products. FedEx, which
reduced its expected rate of return from 10.1% to 9.1%, described its decision as
follows:
Establishing the expected future rate of investment return on our pension assets is a
judgmental matter. Management considers the following factors in determining this
assumption:
• The duration of our pension plan liabilities, which determines the investment
strategy we can employ with our pension plan assets.
• The types of investment classes in which we invest our pension plan assets and
the expected compound return we can reasonably expect those investment
classes to earn over the next 10- to 15-year time period (or such other time period
that may be appropriate).
95
The Magic of Pension Accounting, Part III 07 February 2005
We review the expected long-term rate of return on an annual basis and revise it as
appropriate. Also, we periodically commission detailed asset/liability studies
performed by third-party professional investment advisors and actuaries. These
studies project our estimated future pension payments and evaluate the efficiency of
the allocation of our pension plan assets into various investment categories. These
studies also generate probability-adjusted expected future returns on those assets.
The asset/liability study for 2004 supports a long-term return on assets of at least
9.10%. Our actual compound return on assets was 9.10% for the 15-year period
ended March 31, 2003. Based on these factors, we selected 9.10% as our estimated
future rate of return on pension assets for 2004.
FedEx also laid out the negative effect on earnings of reducing its expected rate of
return:
The 100-basis-point decrease in the expected long-term rate of return for 2004 will
negatively affect our 2004 pension cost by approximately $65 million. Our 2003
pension cost was negatively affected by approximately $48 million by the 80-basis-
point decrease in the expected long-term rate of return to 10.10% for 2003 from
10.90% for 2002.
Of the nine remaining companies with expected rates of return above 9%, a few actually
try to defend their assumptions; three examples are included in Exhibit 96.
96
The Magic of Pension Accounting, Part III 07 February 2005
At the other end of the spectrum are the 10 companies in Exhibit 97 with expected rates
of return below 6% in 2003. KLA-Tencor is at the bottom of the S&P 500 using an
expected rate of return of 3.75%.
Discount Rate
The health of a defined benefit pension plan, its funded status, is extremely sensitive to
changes in the discount rate assumption. For example, depending upon the duration of
the PBO, a 100-basis-point swing in the discount rate could easily move the PBO by
10% or more. The discount rate drives the value of the projected benefit obligation,
which in turn affects the funded status and is used to calculate interest cost and service
cost. Remember, a higher discount rate results in a smaller pension obligation,
improving the funded status of the plan. A higher discount rate also ends up reducing
net pension cost, increasing earnings. A lower discount rate has the exact opposite
impact, increasing the size of the pension obligation and increasing net pension cost.
The discount rate is also the assumption that management has the least control over, as
it is supposed to represent the rate at which the pension plan could be effectively
settled. Companies are allowed to look to current prices of annuity contracts or rates on
high-quality corporate bonds when setting the discount rate. Most companies use
corporate bond yields, a point clarified by the SEC in a 1993 letter to the EITF. Upset
that discount rate assumptions were not reflecting market yields for interest rates, which
at that point were at 10-year lows, the SEC made clear that the discount rate should
reflect the current level of interest rates. The SEC staff specifically suggested using
high-quality corporate debt yields, “ . . . for example, a fixed-income security that
receives a rating of Aa or higher from Moody’s would be considered high quality.” Ever
since this letter, the median discount rate for the S&P 500 has tracked very closely with
the Moody’s Aa yield as seen in Exhibit 98, with the spread between the two rates never
widening more than 40 basis points over the past 11 years. The median pension
discount rate used by the companies in the S&P 500 was 6.10% at the end of 2003,
compared with 6.75% in 2002 and 7.25% in 2001.
97
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 98: Moody's Aa Corporate Bond Rate versus S&P 500’s Median Discount Rate
Difference (bps) Median Discount Rate Moody's Aa Corporate Bond Rate
70 9.00%
60 8.50%
8.00%
50
Difference (bps)
7.50%
40
Rate
7.00%
30
6.50%
20
6.00%
10 5.50%
- 5.00%
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
Given the SEC guidelines, you might expect that all companies would use the exact
same discount rate assumption. Exhibit 99, which shows the distribution of discount
rates across the companies in the S&P 500, would prove that assumption wrong.
Conservative Aggressive
300
254
Number of Companies
250
200
150
100 90
50
1 11 8 2
-
Below 4.00 4.00-4.99 5.00-5.99 6.00-6.99 7.00-7.99 8.00 or Above
Yes, there does appear to be a tight grouping of companies around the median rate of
6.1%; in fact 69% of the companies with defined benefit pension plans in the S&P 500
use discount rates between 6.0% and 7.0%. However, you may wonder why they are
not all exactly the same and why do we see such a wide range—from KLA-Tencor’s
3.63% to AES’s 10.84%. There are several legitimate reasons why we see a variety of
different discount rates across companies, including:
• International plans. Companies with international pension plans will use different
discount rates. The discount rate used on a company’s Japanese plan, for example,
will differ from the discount rate used on its Turkish pension plan, which will differ from
the discount rate on its U.S. plan, etc.
98
The Magic of Pension Accounting, Part III 07 February 2005
• Timing. Companies with different fiscal year-ends will choose different discount rates
based on different measurement dates. In some cases, a pension plan may have a
different fiscal year than its sponsor, as long as it falls within 90 days prior to the
pension plan sponsor’s year-end.
• Demographics. Companies with an older workforce will generally use a lower discount
rate than those with a younger workforce due to the term of their respective
obligations.
Exhibit 100: Median Discount Rate Assumptions across Industry Groups, 2003 (%)
Industry % Industry % Industry %
Multiline Retail 6.88 Health Care Providers & Services 6.18 Paper & Forest Products 6.00
Air Freight & Logistics 6.75 Tobacco 6.15 Thrifts & Mortgage Finance 6.00
Automobiles 6.50 Machinery 6.14 IT Services 5.99
Food & Staples Retailing 6.38 Airlines 6.13 Electrical Equipment 5.98
Auto Components 6.30 Communications Equipment 6.13 Personal Products 5.94
Construction Materials 6.25 Gas Utilities 6.13 Health Care Equip. & Supplies 5.93
Containers & Packaging 6.25 Road & Rail 6.13 Industrial Conglomerates 5.85
Distributors 6.25 Specialty Retail 6.13 Computers & Peripherals 5.85
Diversified Financial Services 6.25 Chemicals 6.11 Building Products 5.84
Diversified Telecom Services 6.25 Beverages 6.10 Office Electronics 5.80
Electric Utilities 6.25 Household Products 6.05 Construction & Engineering 5.75
Hotels Restaurants & Leisure 6.25 Household Durables 6.05 Pharmaceuticals 5.70
Insurance 6.25 Textiles Apparel & Luxury Goods 6.05 Electronic Equipment & Instruments 5.40
Media 6.25 Capital Markets 6.00 Biotechnology 5.22
Metals & Mining 6.25 Commercial Services & Supplies 6.00 Semiconductors & Semiconductor Equip. 4.90
Multi-Utilities & Unregulated Power 6.25 Consumer Finance 6.00 Software N/A
Real Estate 6.25 Energy Equipment & Services 6.00 Internet & Catalog Retail N/A
Wireless Telecom Services 6.25 Food Products 6.00 Internet Software & Services N/A
Commercial Banks 6.23 Leisure Equipment & Products 6.00 Trading Companies & Distributors N/A
Aerospace & Defense 6.18 Oil & Gas 6.00
Source: Company data, CSFB estimates.
Even though the discount rate is technically the assumption that management has the
least control over, we still see a wide range of assumptions even within industry groups.
The 12 industry groups in Exhibit 101 each had a spread of over 150 basis points
between the highest and lowest discount rates. For each group, we show the company
that used the highest discount rate and the one with the lowest. For example, Allstate
used a 7% discount rate, while AFLAC used 4.64%; both companies are in the
Insurance industry group. The reason for AFLAC’s low discount rate? Nearly half of its
pension plan is in Japan, where the company is using a discount rate of 2.5%.
99
The Magic of Pension Accounting, Part III 07 February 2005
In 2003, KLA-Tencor employed the lowest discount rate in the S&P 500 at 3.63%, while
National Semiconductor held this distinction in 2002. Exhibit 103 lists the 12 companies
that used discount rates less than 5% in 2003. Remember, all else equal, a lower
discount rate is more conservative.
100
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 103: Companies Employing Discount Rates of Less Than 5%, 2003
Discount Rates (%)
Company Ticker 2003 2002 2001
KLA-Tencor Corp KLAC 3.63 N/A N/A
Andrew Corp ANDW 4.06 6.25 N/A
Molex Inc MOLX 4.07 4.67 N/A
National Semiconductor Corp NSM 4.30 4.30 4.65
Texas Instruments Inc TXN 4.58 4.91 5.25
Jabil Circuit Inc JBL 4.60 6.20 N/A
Pall Corp PLL 4.62 5.41 N/A
AFLAC Inc AFL 4.64 4.72 4.90
Applera Corp Applied Biosys ABI 4.81 7.25 7.50
Ace Limited ACE 4.84 5.06 N/A
Tupperware Corp TUP 4.86 5.44 5.62
Zimmer Hldgs Inc ZMH 4.92 6.05 6.01
Source: Company data, CSFB estimates.
A higher salary inflation assumption will result in a higher PBO, reducing the funded
status and increasing pension cost. Exhibit 104 displays the distribution of salary
inflation assumptions among the companies in the S&P 500. The median salary inflation
assumption used by the companies in the S&P 500 was 4.00% in 2003, compared with
4.15% in 2002 and 4.50% in 2001.
101
The Magic of Pension Accounting, Part III 07 February 2005
Aggressive Conservative
200
179
180
Number of Companies
160
140
124
120
100
80
60
40 38
20 7 8
-
Below 3.00 3.00-3.99 4.00-4.99 5.00-5.99 6.00 or Above
Exhibit 105: Median Salary Inflation Rate Assumptions across Industry Groups, 2003 (%)
Industry % Industry % Industry %
Thrifts & Mortgage Finance 5.00 Commercial Services & Supplies 4.00 Household Products 3.81
Tobacco 4.80 Consumer Finance 4.00 Computers & Peripherals 3.79
Multi-Utilities & Unregulated Power 4.69 Electric Utilities 4.00 Personal Products 3.76
Diversified Financial Services 4.57 Energy Equipment & Services 4.00 Textiles Apparel & Luxury Goods 3.75
Capital Markets 4.50 Gas Utilities 4.00 Hotels Restaurants & Leisure 3.75
Real Estate 4.50 Health Care Providers & Services 4.00 Semiconductors & Semiconductor Equip. 3.70
Aerospace & Defense 4.49 Leisure Equipment & Products 4.00 Road & Rail 3.70
Oil & Gas 4.39 Machinery 4.00 Electronic Equipment & Instruments 3.65
Food Products 4.35 Media 4.00 Paper & Forest Products 3.65
IT Services 4.27 Metals & Mining 4.00 Construction & Engineering 3.50
Beverages 4.25 Multiline Retail 4.00 Distributors 3.25
Diversified Telecom Services 4.25 Pharmaceuticals 4.00 Biotechnology 3.19
Insurance 4.25 Food & Staples Retailing 3.98 Building Products 3.10
Specialty Retail 4.25 Health Care Equip. & Supplies 3.96 Construction Materials 2.80
Wireless Telecom Services 4.25 Electrical Equipment 3.95 Airlines 1.89
Automobiles 4.23 Auto Components 3.92 Software N/A
Industrial Conglomerates 4.13 Office Electronics 3.90 Internet & Catalog Retail N/A
Air Freight & Logistics 4.00 Containers & Packaging 3.89 Internet Software & Services N/A
Chemicals 4.00 Household Durables 3.86 Trading Companies & Distributors N/A
Commercial Banks 4.00 Communications Equipment 3.85
Source: Company data, CSFB estimates.
102
The Magic of Pension Accounting, Part III 07 February 2005
Finding a wide range of salary inflation assumptions across different industries should
not really be much of a surprise. However, we even found a wide range of assumptions
within industry groups. The five industry groups in Exhibit 106 each had a spread of
over 330 basis points between the highest and lowest salary inflation rates. For each
group, we show the company that used the highest salary inflation assumption and the
one with the lowest. For example, Federated used a 5.8% salary inflation rate, while
Dillards used 2.5%; both companies are in the Multiline Retail industry group.
Exhibit 106: Widest Range in 2003 Salary Inflation Rate Assumptions (%)
Highest Salary Inflation Rate Lowest Salary Inflation Rate
Industry Median Range Company Ticker % Company Ticker %
Energy Equipment & Services 4.00 5.12 Halliburton Co HAL 8.49 Schlumberger Ltd SLB 3.37
Capital Markets 4.50 4.50 Franklin Resources Inc BEN 8.00 Mellon Financial Corp MEL 3.50
Metals & Mining 4.00 4.25 Freeport McMoran Cop & Gld FCX 8.00 Allegheny Technologies Inc ATI 3.75
Aerospace & Defense 4.49 3.93 General Dynamics Corp GD 7.50 Goodrich Corp GR 3.57
Multiline Retail 4.00 3.30 Federated Dept Stores FD 5.80 Dillards Inc -Cl A DDS 2.50
On the other end of the spectrum, KLA-Tencor used the lowest salary inflation
assumption in 2003, 1.5%, dethroning American Standard, which had used the lowest
salary inflation rate assumption in 2002, 1.7%. The seven companies in Exhibit 108
employed salary inflation rates of less than 3% in 2003.
Exhibit 108: Companies Employing Salary Inflation Rates of Less Than 3%, 2003
Salary Inflation Rate (%)
Company Ticker 2003 2002 2001
KLA-Tencor Corp KLAC 1.50 N/A N/A
American Standard Cos Inc ASD 1.70 1.70 1.70
Delta Air Lines Inc DAL 1.89 2.67 4.67
Dillards Inc -Cl A DDS 2.50 2.50 N/A
Zimmer Hldgs Inc ZMH 2.71 3.46 3.30
Chiron Corp CHIR 2.75 3.00 3.00
Vulcan Materials Co VMC 2.80 4.00 4.25
Source: Company data, CSFB estimates.
103
The Magic of Pension Accounting, Part III 07 February 2005
In this section of the report, we dig deeper to try and answer a simple question: Do the
assumptions make sense (specifically the expected return and discount rate)? In other
words, we are looking for support of the pension assumptions, which appears to be the
very same thing that the SEC is looking for (and apparently in some cases is having
difficulty finding). The inability to support the assumptions should cause you to pause; if
a company can’t support its pension assumptions, then you start to wonder which of the
myriad other accounting assumptions it can’t support.
104
The Magic of Pension Accounting, Part III 07 February 2005
reasonable. These are insufficient responses. No matter whom a company seeks out
for advice in making assumptions like the discount rate, the company should have
support for its assumptions that goes beyond a comparison to others. . . . In any
event, we believe that size of pension obligations and the significance of the discount
rate assumption to their calculation indicate that significant thought should be given
to the rate’s selection.
Further guidance on the discount rate assumption was provided by SEC Associate
Chief Accountant Jane Poulin in a speech at the AICPA National Conference on Current
SEC Developments in December (for more on the conference, see our December 13,
2004, report, Accountingfest 2004):
It’s important to note that both FAS 87 and FAS 106 indicate that, “. . . conceptually,
the selection of the assumed discount rate should be based on the single sum that, if
invested at the measurement date, would generate the necessary cash flows to pay
the benefits when due. Each company should have support for the rate selected
based on an evaluation, not just a comparison to other companies. For example, in
some cases, a company may construct a hypothetical portfolio of high quality
instruments with maturities that mirror the benefit obligation in order to accurately
estimate the discount rate relevant to their particular plan.
Remember that this is not the first time the SEC has poked around the pension
assumptions. Back in 1993 the SEC—upset that discount rate assumptions were not
reflecting market yields for interest rates—made clear that the discount rate should
reflect the current level of interest rates on high-quality corporate bonds. In December
2002, the SEC staff urged companies to pay close attention to their expected return
assumptions, indicating that the SEC could challenge anything over 9%.
105
The Magic of Pension Accounting, Part III 07 February 2005
If there is a lack of support for the pension assumptions, they may have to change; for
example, expected returns could get ratcheted down. So how do you figure out the
impact on the health of the pension plan and on pension cost from a change in one of
the pension assumptions? We recommend getting your hands on the sensitivity analysis
to changes in pension assumptions that some companies are now providing in either
the pension footnote or the Management Discussion and Analysis (MD&A) section of
the 10-K. For example, the sensitivity analysis in Exhibit 110 comes from Ford’s MD&A:
106
The Magic of Pension Accounting, Part III 07 February 2005
The foregoing indicates that changes in the discount rate and return on assets can have a significant effect on the funded
status of our pension plans and Stockholders' Equity. As stated above, we base the discount rate assumption on
investment yields available at year-end on corporate long-term bonds rated AA. We cannot predict these bond yields or
investment returns and, therefore, cannot reasonably estimate whether adjustments to our Stockholders' Equity for
minimum pension liability in subsequent years will be significant.
Source: Ford Motor's 12/31/03 10-K.
• Expected returns. The economic reality of the pension plan is unaffected by the
expected return. In other words, the expected return affects earnings but does not
impact the funded status of a pension plan. This thought process holds for all plans—
not just Ford’s.
If the company does not provide a sensitivity analysis, you can run varying assumptions
through our pension forecast model, or check out the sensitivity analysis and impacts of
changing the expected return that we have included in the Who Has Exposure? section
of this report.
107
The Magic of Pension Accounting, Part III 07 February 2005
Expected Return
We start off with the expected return assumption, easily the most contentious of the
pension assumptions and the one that investors question most often. As most investors
we speak to are in the business of generating these returns, they wonder how some
companies arrive at the assumptions; and the higher the return assumption, the more
skeptical these investors become.
Many investors focus solely on the absolute level of the pension assumptions to
determine whether or not they make sense. For example, a 10% expected rate of return
assumption on its own might not appear to make sense. However, after examining the
asset allocation of the pension plan and where the plan is located, you could arrive at a
different conclusion (or not).
To try and figure out whether these assumptions appear to make any sense, we start off
with historical returns for broad categories of asset classes. Then we look to the pension
disclosures for information on the plan asset allocations to see whether or not the
expected return assumptions make sense versus historical returns for similarly allocated
portfolios. Next we take a look at how the expected return assumptions have changed
and how they compare with the actual returns generated by the plan.
Historical Returns
Remember, the expected rate of return assumption is supposed to be a long-term
concept (10-plus years) that will vary depending on a company’s belief about future
market performance, its ability to generate rates of return either above or below those
market levels, and the mix of plan assets. As a starting point, the SEC indicated that
companies should use historical returns of similarly allocated portfolios. We will start in
the same place.
Critics of this approach argue that the world is much different today than it was in 1926,
so it’s inappropriate to use returns that go that far back. In Exhibit 111, we assess the
impact on long-term historical equity and fixed-income market returns if we begin with
1926 and begin to knock off decades of return data one-by-one until we arrive at the
present. In other words, we progress from returns over a 78-year period, to returns over
the past 8 years.
108
The Magic of Pension Accounting, Part III 07 February 2005
1
Exhibit 111: U.S. Equity and Fixed Income Compounded Annual Rates of Return
14%
12.6% 12.4%
12% 11.5%
11.0%
10.4% 10.6% 10.4%
9.9%
10% 9.6%
8.9%
7.9%
8% 7.5%
6.5%
6% 5.4% 5.5% 5.6%
4%
2%
0%
1926-2003 1936-2003 1946-2003 1956-2003 1966-2003 1976-2003 1986-2003 1996-2003
1
Growth rate calculations are based on geometric mean.
Source: Ibbotson.
In Exhibit 112, we slice the same return data a bit differently, looking at 10-year rolling-
average returns for both equities and fixed income.
1
Exhibit 112: 10-Year Rolling-Average U.S. Equity and Fixed-Income Market Returns
20.0%
18.0%
16.0%
14.0%
12.0%
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
12/1936 12/1946 12/1956 12/1966 12/1976 12/1986 12/1996 12/2003
(2.0%)
(4.0%)
(6.0%)
1
10-year rolling average return calculations are based on geometric mean.
Source: Ibbotson.
A slightly different picture begins to form when you go out farther in time. Exhibit 113
shows that the rolling-average 20-year equity market returns have for the most part
stayed above 10%.
109
The Magic of Pension Accounting, Part III 07 February 2005
1
Exhibit 113: 20-Year Rolling-Average U.S. Equity and Fixed-Income Market Returns
20.0%
Equity Fixed Income
17.5%
15.0%
12.5%
10.0%
7.5%
5.0%
2.5%
0.0%
12/1946 12/1956 12/1966 12/1976 12/1986 12/1996 12/2003
1
20-year rolling average return calculations are based on geometric mean.
Source: Ibbotson.
110
The Magic of Pension Accounting, Part III 07 February 2005
1
Exhibit 114: U.S. Real Estate Compounded Annual Rates of Return
Real Estate
16%
14%
12%
9.7%
10% 9.2% 9.2%
7.7%
8%
6.7%
6%
4%
2%
0%
1978-2003 1983-2003 1988-2003 1993-2003 1998-2003
1
Growth rate calculations are based on geometric mean.
Source: National Council of Real Estate Investment Fiduciaries.
In Exhibit 115, we slice the same return data a bit differently, looking at 10-year rolling-
average returns for real estate.
1
Exhibit 115: 10-Year Rolling-Average U.S. Real Estate Market Returns
14.0%
Real Estate
12.0%
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
12/1987 11/1989 11/1991 11/1993 11/1995 11/1997 11/1999 11/2001 11/2003
1
10-year rolling average return calculations are based on geometric mean.
Source: National Council of Real Estate Investment Fiduciaries.
111
The Magic of Pension Accounting, Part III 07 February 2005
is as far back as data are available from the National Council of Real Estate Investment
Fiduciaries. For other assets, we use an 8.33% return, which is simply an equal-
weighted blend of historical equity, fixed–income, and real estate returns. We decided to
use a blended return for the other assets, as it is a bit of a wildcard. Some companies
have private equity and venture capital in their other assets category, while other
companies have short-term money market instruments.
Now that we have historical returns, the next step is to look at the pension plan’s asset
allocation. We gathered the asset allocation data for each company in the S&P 500 that
is now disclosed in the pension footnote. (See Asset Allocation section of this report for
more detail.)
Armed with the historical returns and asset allocations for each company, we calculated
our own estimated returns by multiplying each company’s asset allocations by the
historical return for each asset class. For example, if the company’s asset allocation
were 50% equities and 50% fixed income, the estimated return would be 7.90%—(50%
x 10.40%) + (50% x 5.40%). As a first step to assess the reasonableness of the
expected return assumptions we compare them with the estimated return.
We applied this process to every company in the S&P 500 with a defined benefit
pension plan and found that 270 of the 370 companies with pension plans actually
employed an expected rate of return that falls within 100 basis points of our estimate,
while 170 companies employed an expected rate of return that falls within 50 basis
points of our estimates. Not a surprise, it appears as if historical returns are driving the
expected return assumption.
The expected rate of return is higher than our estimate for 87 companies, including the
four in Exhibit 116 that used expected rates of return more than 100 basis points higher
than our estimated return.
Exhibit 116: Expected Return per Company > Estimated Return per Asset Allocations (%)
A B C=A–B
Expected Estimated Difference
Company Ticker Return (%) Return (%) (bps)
AES Corp. (The) AES 13.07% 7.09% 598
Freeport McMoran Cop & Gld FCX 10.00% 6.91% 309
Allegheny Energy Inc AYE 9.00% 7.51% 149
Starwood Hotels & Resorts Wrld HOT 7.30% 6.02% 129
Source: Company data, CSFB estimates.
On the surface, the expected return assumptions for these companies appear
aggressive when compared to historical returns. However, in some cases, there may be
an explanation; for example, companies like AES and Freeport McMoran that have
pension plans overseas, where the expected returns differ. AES calculates “. . . the
pension cost for the Brazilian plan based upon a number of actuarial assumptions,
including an expected long-term rate of return on plan assets of 14% in 2003.”
On the other hand, we found 272 companies where the expected return used by the
company appears to be conservative when compared with historical returns, as it is
lower than our estimated return, including the 12 companies in Exhibit 117 that used
expected rates of return over 200 basis points below our estimated return.
112
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 117: Expected Return per Company < Estimated Return per Asset Allocations (%)
A B C=A-B
Expected Estimated Difference
Company Ticker Return (%) Return (%) (bps)
Stryker Corp SYK 5.30% 8.98% (368)
Intel Corp INTC 6.87% 10.40% (353)
Biogen Idec Inc BIIB 5.00% 7.98% (298)
Interpublic Group Of Cos IPG 6.36% 9.27% (291)
Texas Instruments Inc TXN 6.33% 9.10% (277)
Zimmer Hldgs Inc ZMH 5.88% 8.65% (277)
Fluor Corp FLR 6.50% 8.81% (231)
Cincinnati Financial Corp CINF 8.00% 10.28% (228)
Fannie Mae FNM 7.50% 9.70% (220)
Sigma-Aldrich SIAL 7.22% 9.35% (213)
MGIC Investment Corp/Wi MTG 7.50% 9.55% (205)
Agilent Technologies Inc A 7.36% 9.40% (204)
Source: Company data, CSFB estimates.
Looking Forward
We first looked backward, at historical returns, to get an idea as to whether the
expected rates of return were reasonable, a good first step. However, the expected
return is a forward-looking concept, focused on what type of return management
expects the portfolio to earn in the future.
Everyone has their own view as to what the markets will return to investors in the future.
With that said, we constructed the matrix in Exhibit 118, assuming a simplified portfolio
allocation of 65% equity and 35% fixed income to help assess the reasonableness of
the expected rate of return assumptions, with a forward-looking slant. Say that you think
equities are going to return 8% going forward and that you expect fixed income to
generate a 5% return, you would expect an approximately 6.95% return on the portfolio.
Now compare that with what the company is expecting. The cells highlighted in the
matrix below represent the median, high, and low expected rates of return for the entire
S&P 500.
113
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 118: Expected Pension Plan Asset Portfolio Returns Associated with Different
Equity and Fixed-Income Market Returns
Actual Equity Market Return
15% 2% 4% 6% 8% 10% 12% 14% 16% 18%
10% 4.80% 6.10% 7.40% 8.70% 10.00% 11.30% 12.60% 13.90% 15.20%
= = =
The diagonal line running through the exhibit represents points at which the expected
return and actual return are equal. As you can plainly see, not that many pension plans
outperformed their expected returns from 1999 through 2003; in fact, only four
companies were able to make this claim. On the other hand, the actual return on
pension plan assets fell short of the expected return for 207 of the 211 companies in the
S&P 500 for which we have a complete set of data for the past five years.
114
The Magic of Pension Accounting, Part III 07 February 2005
1
Exhibit 119: Actual Return versus Expected Returns
(6%)
1
Based on geometric mean of returns over the five-year period from 1999 to 2003 for 211 companies.
Source: Company data, CSFB estimates.
The actual return on plan assets for the S&P 500 was 18.80% in 2003, (8.85%) in 2002,
and (7.38%) in 2001. Exhibit 120 shows the top 10 companies in terms of actual positive
or negative returns for 2003. Note that all 10 companies with the worst returns in 2003
have fiscal year-ends of June 30 or earlier. Therefore, the actual returns for the pension
plans of these companies would not reflect the strong stock market performance in the
second half of 2003 and would be burdened with the poor performance from 2002.
Exhibit 120: Highest Positive and Negative Actual Rates of Return, 2003
Highest Positive Return Highest Negative Return
Company Ticker 2003 2002 Company Ticker 2003 2002
Merck & Co MRK 33.27% (8.54%) National Semiconductor Corp NSM (22.14%) (7.00%)
Eastman Chemical Co EMN 32.93% (13.77%) Circuit City Str Crct Cty Gp CC (19.98%) (5.64%)
Time Warner Inc TWX 32.56% (14.63%) Brown-Forman -Cl B BF.B (18.56%) (13.18%)
Ambac Financial Gp ABK 32.11% (12.20%) Meredith Corp MDP (18.30%) 10.83%
EMC Corp/Ma EMC 31.19% (15.63%) Worthington Industries WOR (15.27%) (9.50%)
Zimmer Hldgs Inc ZMH 29.97% (20.41%) Scientific-Atlanta Inc SFA (15.17%) (3.36%)
Suntrust Banks Inc STI 29.96% (11.32%) Heinz (H J) Co HNZ (12.36%) (0.65%)
Weyerhaeuser Co WY 29.78% (13.98%) FedEx Corp FDX (12.03%) (3.40%)
Wellpoint Hlth Netwrk -Cl A WLP 29.53% (13.39%) Archer-Daniels-Midland Co ADM (11.73%) 2.04%
Progress Energy Inc PGN 28.67% (13.61%) Sara Lee Corp SLE (11.33%) 2.22%
Source: Company data, CSFB estimates.
The vast majority of defined benefit plans had positive returns in 2003. In fact, 134
companies had positive returns in excess of 20%. Only 34 companies’ pension plan
assets lost value in 2003. Exhibit 121 shows the distribution of actual rates of return
across the companies in the S&P 500 during 2003.
115
The Magic of Pension Accounting, Part III 07 February 2005
140
120
120
102
Number of Companies
100
80
59
60
40 32
27
17
20
7
Exhibit 122 shows the companies that experienced the widest positive and negative
differences or spreads between actual and expected returns in 2003. There were 305
companies that earned actual rates of return on plan assets in excess of their expected
rates of return in 2003, compared with only 4 in 2002 and 7 in 2001.
Exhibit 122: Widest Positive and Negative Spreads between Expected Return Rates and Actual Return Rates, 2003
Widest Positive Spreads (%) Widest Negative Spreads (%)
Company Ticker Expected Rate Actual Return Company Ticker Expected Rate of Actual Return
of Return Return
Merck & Co MRK 7.70% 33.27% Circuit City Str Crct Cty Gp CC 8.25% (19.98%)
Time Warner Inc TWX 8.00% 32.56% National Semiconductor Corp NSM 5.40% (22.14%)
Zimmer Hldgs Inc ZMH 5.88% 29.97% Brown-Forman -Cl B BF.B 8.75% (18.56%)
Eastman Chemical Co EMN 8.94% 32.93% Meredith Corp MDP 8.25% (18.30%)
Ambac Financial Gp ABK 8.75% 32.11% Scientific-Atlanta Inc SFA 10.00% (15.17%)
EMC Corp/Ma EMC 8.20% 31.19% FedEx Corp FDX 10.10% (12.03%)
Suntrust Banks Inc STI 8.75% 29.96% Heinz (H J) Co HNZ 8.90% (12.36%)
Wellpoint Hlth Netwrk -Cl A WLP 9.00% 29.53% Albertsons Inc ABS 8.50% (10.94%)
Weyerhaeuser Co WY 9.50% 29.78% Worthington Industries WOR 4.00% (15.27%)
Coca-Cola Co KO 7.75% 27.89% Archer-Daniels-Midland Co ADM 7.50% (11.73%)
Source: Company data, CSFB estimates.
116
The Magic of Pension Accounting, Part III 07 February 2005
return assumptions where the range had changed. Or in one instance, what appears to
be an error in the disclosure. Exhibit 123 shows ACE’s 2003 disclosure of its pension
assumptions and Exhibit 124 shows its 2002 disclosures, both exhibits come directly
from the company’s 10-K. Note that the expected rate of return appears to have jumped
334 basis points, from 3.35% in 2002 (see Exhibit 124) to 6.69% in 2003 (see Exhibit
123). However, it also appears possible that ACE’s expected rate of return and
compensation inflation rates were entered into the wrong rows in 2002, particularly
since ACE’s rate of compensation increase in 2002 exceeds both its discount rate and
expected rate of return according to Exhibit 124 (a very rare occurrence) and that the
2002 amounts in the 2003 disclosures appear to have changed, as illustrated in Exhibit
123. This provides a great example of the importance of scrutinizing the pension
assumptions without taking anything at face value.
Exhibit 123: ACE’s 2003 Pension Plan Disclosures Exhibit 124: ACE’s 2002 Pension Plan Disclosures
2003 2002 2002 2001
---------------------------------------------------------- ---------------------------------------------------------
Weighted-average assumptions as of December 31 Weighted-average assumptions as of December 31
Discount rate 4.84% 5.22% Discount rate 5.06% 6.53%
Expected rate of return on plan assets 6.69% 6.86% Expected rate of return on plan assets 3.35% 4.22%
Rate of future compensation increase 4.40% 3.36% Rate of compensation increase 7.81% 8.16%
Source: Annual report and 10-K. Source: Annual report and 10-K.
There were 107 companies that dropped their expected rates of return by over 50 basis
points from 2002 to 2003, including the 16 companies in Exhibit 125 that dropped their
rates by over 150 basis points.
Exhibit 125: Decreased Expected Rate of Return by over 150 Basis Points, 2003
Expected Return (%)
Company Ticker 2003 2002 Change (bps)
Worthington Industries WOR 4.00 8.36 (436)
Franklin Resources Inc BEN 5.31 8.00 (269)
Merck & Co MRK 7.70 10.00 (230)
Applera Corp Applied Biosys ABI 5.86 8.13 (226)
U S Bancorp USB 8.90 10.90 (200)
Freeport McMoran Cop & Gld FCX 10.00 12.00 (200)
Intel Corp INTC 6.87 8.84 (197)
Zimmer Hldgs Inc ZMH 5.88 7.64 (176)
Federal Home Loan Mortgage Corp FRE 7.25 9.00 (175)
Moodys Corp MCO 8.10 9.75 (165)
Coors (Adolph) -Cl B RKY 7.92 9.50 (158)
Tiffany & Co TIF 7.50 9.00 (150)
Conagra Foods Inc CAG 7.75 9.25 (150)
McKesson Corp MCK 8.25 9.75 (150)
Mellon Financial Corp MEL 8.50 10.00 (150)
Wachovia Corp WB 8.50 10.00 (150)
Source: Company data, CSFB estimates.
117
The Magic of Pension Accounting, Part III 07 February 2005
Discount Rates
If the expected return assumption is the most contentious, the discount rate is arguably
the most powerful pension assumption. The health of the pension plan, its funded
status, is very sensitive to changes in the discount rate. Even though this is the
assumption that management technically has the least control over, we noted in the
previous section of this report a wide range in discount rates, and a number of reasons
as to why that could occur.
To try and figure out whether these assumptions appear to make any sense, we look to
interest rate levels and the duration of the pension obligation to support the discount
rate.
118
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 127: Month-End Spot Rates for Moody’s Aa Corporate Bond Index and Pension
Discount Rates, 2003
6.50%
6.5% 6.42%
6.45%
6.13%
6.0%
6.01%
5.5%
5.50%
5.0%
01/2003 02/2003 03/2003 04/2003 05/2003 06/2003 07/2003 08/2003 09/2003 10/2003 11/2003 12/2003
# of Companies: 11 6 3 1 4 17 2 2 52 9 8 257
In general, the monthly median discount rate appeared to track the Moody’s Aa
corporate bond index spot rate at the end of each month reasonably closely throughout
2003. However, it is interesting to note that the median discount rate was higher than
the Moody’s rate for eight out of twelve months during the year—May, July, August, and
November were the only exceptions. Remember, a higher discount rate results in a
smaller pension obligation and lower pension cost. We found that 250 companies’
discount rates were within 25 basis points of the Moody’s Aa rate. However, there were
other companies where the discount rate appeared to bear no relation to the Moody’s
Aa rate, including the 10 companies in Exhibit 128 with the largest positive spreads
between the Moody’s Aa rate and the discount rate and the ten companies with the
largest negative spreads.
119
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 128: Largest Difference between Moody’s Aa Corporate Bond Rate and a Company’s Discount Rate
Discount Rate > Moody’s Aa Rate Discount Rate < Moody’s Aa Rate
Discount Moody’s Aa Difference Discount Moody’s Aa Difference
Measurement Rate Corporate (Basis Measurement Rate Corporate (Basis
Company Ticker Date1 2003 Bond Rate2 Points) Company Ticker Date1 2003 Bond Rate2 Points)
AES Corp. (The) AES 12/31 10.84% 6.01% 483 KLA-Tencor Corp KLAC 6/30 3.63% 5.83% (221)
Halliburton Co HAL 9/30 10.01% 5.86% 415 Pall Corp PLL 7/31 4.62% 6.42% (180)
Freeport McMoran FCX 12/31 7.75% 6.01% 174 Andrew Corp ANDW 9/30 4.06% 5.86% (180)
Conagra Foods Inc CAG 2/28 7.25% 6.23% 102 Molex Inc MOLX 6/30 4.07% 5.83% (176)
Penney (J C) Co JCP 10/31 7.10% 6.11% 99 Jabil Circuit Inc JBL 8/31 4.60% 6.24% (164)
Target Corp TGT 10/31 7.00% 6.11% 89 National Semiconductor NSM 5/31 4.30% 5.73% (143)
Allstate Corp ALL 10/31 7.00% 6.11% 89 Texas Instruments Inc TXN 12/31 4.58% 6.01% (143)
United Parcel Service UPS 9/30 6.75% 5.86% 89 AFLAC Inc AFL 9/30 4.64% 5.86% (122)
Dillards Inc -Cl A DDS 1/31 7.25% 6.45% 80 Ace Limited ACE 12/31 4.84% 6.01% (117)
Supervalu Inc SVU 2/28 7.00% 6.23% 77 Tupperware Corp TUP 12/31 4.86% 6.01% (115)
1
Used fiscal year-end as measurement date in cases where company did not provide one.
2
Note that the Moody’s Aa Corporate Bond Rate used in this analysis is based on the rate at the end of each month during 2003, which we set to
correspond with each S&P 500 company’s measurement date. For example, we applied the September month end rate to all companies with September
measurement dates and went on to apply this same methodology throughout the year.
Source: Company data, CSFB estimates.
There are a couple of good reasons as to why these wide spreads could occur. One is
the presence of international pension plans. For example, AES’s discount rate is largely
influenced by geography—the majority of its pension plans are overseas, where the
company uses a discount rate of 11.8%; the discount rate for its U.S. plans is 6%.
Another reason may be demographics, which we tackle next.
120
The Magic of Pension Accounting, Part III 07 February 2005
In Exhibit 129, we plot out the ratio of benefits paid to the pension obligation versus the
discount rates used by each company in the S&P 500. In theory, a downward sloping
line should magically appear; as we move from a longer-duration pension obligation
toward a shorter-duration pension obligation, we would expect to see the discount rates
drop. Instead, we find what is nearly a straight line across; in other words, no matter the
estimated duration of the pension obligation, companies appear to be using around the
same discount rate.
12%
AES
HAL
10%
Discount Rate (%)
8% FCX GM
DDS HCR
FDX
MAY
SBC
6%
2%
0%
0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% 16% 17% 18% 19% 20%
Longer Ratio of Benefits Paid/PBO (%) Shorter
Term Term
OK, so that didn’t work as well as we had planned. Does it go to show that companies
simply take the Moodys Aa rate, ignoring the duration of their pension obligation? We
are not sure. From the previous analysis, there is clearly a much stronger relationship
between discount rates and the yield on the Moody’s Aa index than between discount
rates and whether the pension obligation appears to be longer duration or shorter
duration.
The changes in a company’s discount rate assumption will clearly depend upon the
pension plan’s year-end. For example, the Moodys Aa rate at December 31, 2002, was
6.52%; it dropped by 51 basis points to 6.01% at December 31, 2003. Contrast that with
the month of May, where the rate decreased by 143 basis points year over year. Once
again, using the pension plan year-ends, we took a look at changes in discount rate
assumptions in relation to year-over-year changes in the Moody’s Aa rate for that
particular month-end. We found that the change in the discount rate assumption was
121
The Magic of Pension Accounting, Part III 07 February 2005
within 25 basis points of the year-over-year change for the Moodys Aa rate for that
month for 247 companies.
One other thing to keep in mind when evaluating changes in discount rate assumptions
is of course the absolute level of discount rates used by the companies; for example, six
companies in Exhibit 130—Procter & Gamble, Sara Lee, Sealed Air, National
Semiconductor, Molex, and Teradyne—use discount rates that are lower than the yield
on the Moody’s Aa index, indicating a more conservative assumption.
122
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 130: Year-over-Year Drop in the Moody’s Aa Corporate Bond Rate Exceeds Year-over-
Year Change in the Discount Rate by 50 Basis Points (Changes Based on Fiscal Year-End)
Year Over Year
Year Over Year Change in
Change in Moody’s Aa
Measurement Discount Rate Corporate Difference
Company Ticker Date1 2003 Bond Rate (Basis Points)
AES Corp. (The) AES 12/31 278 (51) 329
National Semiconductor Corp NSM 5/31 0 (143) 143
Parker-Hannifin Corp PH 6/30 0 (122) 122
Ingersoll-Rand Co Ltd IR 11/30 25 (70) 95
Cardinal Health Inc CAH 3/31 (30) (106) 76
Hewlett-Packard Co HPQ 10/31 11 (62) 73
Procter & Gamble Co PG 6/30 (50) (122) 72
Freeport McMoran Cop & Gld FCX 12/31 19 (51) 70
Sealed Air Corp SEE 12/31 15 (51) 66
Leggett & Platt Inc LEG 9/30 0 (65) 65
Illinois Tool Works ITW 9/30 0 (65) 65
United Parcel Service Inc UPS 9/30 0 (65) 65
American International Group AIG 12/31 14 (51) 65
Becton Dickinson & Co BDX 9/30 (1) (65) 64
Sara Lee Corp SLE 6/30 (60) (122) 62
Allstate Corp ALL 10/31 0 (62) 62
Molex Inc MOLX 6/30 (60) (122) 62
FedEx Corp FDX 2/28 (11) (72) 61
Safeway Inc SWY 1/31 0 (59) 59
Dillards Inc -Cl A DDS 1/31 0 (59) 59
AFLAC Inc AFL 9/30 (8) (65) 57
McKesson Corp MCK 3/31 (50) (106) 56
Computer Sciences Corp CSC 3/31 (50) (106) 56
Heinz (H J) Co HNZ 6/30 (70) (122) 52
Archer-Daniels-Midland Co ADM 6/30 (70) (122) 52
Teradyne Inc TER 12/31 0 (51) 51
Lincoln National Corp LNC 12/31 0 (51) 51
Bemis Co BMS 12/31 0 (51) 51
Union Pacific Corp UNP 12/31 0 (51) 51
Weyerhaeuser Co WY 12/31 0 (51) 51
1
Used fiscal year-end as measurement date in cases where company did not provide one.
Source: Company data, CSFB estimates.
On the other hand, the change in discount rates for a number of companies appears to
be more conservative than the change in Moody’s Aa rate. For 117 companies, the
year-over-year decline in the Moody’s Aa rate was less than the year-over-year change
in discount rate reported by the company. In other words, the change in the discount
rate appears more conservative than the change in the Moody’s Aa rate for these 117
companies. In Exhibit 131, we highlight the 16 companies where the year-over-year
decline in the yield on the Moody’s Aa index was less than the year-over-year change in
the discount rate reported by over 50 basis points.
123
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 131: Year-over-Year Drop in the Moody’s Aa Corporate Bond Rate Exceeds Year-over-
Year Change in the Discount Rate by 50 Basis Points (Changes Based on Fiscal Year-End)
Year Over Year
Year Over Year Change in
Change in Moody’s Aa
Measurement Discount Rate Corporate Difference
Company Ticker Date1 2003 Bond Rate (Basis Points)
Halliburton Co HAL 9/30 (228) (65) (163)
Andrew Corp ANDW 9/30 (219) (65) (154)
Occidental Petroleum Corp OXY 12/31 (188) (51) (137)
Intel Corp INTC 12/31 (188) (51) (137)
Applera Corp Applied Biosys ABI 6/30 (244) (122) (122)
Jabil Circuit Inc JBL 8/31 (160) (56) (104)
RJ Reynolds Tobacco Hldgs RJR 12/31 (125) (51) (74)
Fifth Third Bancorp FITB 12/31 (125) (51) (74)
Wendy's International Inc WEN 12/31 (125) (51) (74)
Hartford Finl Svcs Grp Inc HIG 12/31 (125) (51) (74)
Disney (Walt) Co DIS 9/30 (135) (65) (70)
Citizens Communications Co CZN 12/31 (113) (51) (62)
Zimmer Hldgs Inc ZMH 12/31 (112) (51) (61)
CSX Corp CSX 9/30 (125) (65) (60)
Tektronix Inc TEK 5/31 (200) (143) (57)
Ryder System Inc R 12/31 (106) (51) (55)
1
Used fiscal year-end as measurement date in cases where company did not provide one.
Source: Company data, CSFB estimates.
124
The Magic of Pension Accounting, Part III 07 February 2005
Funded Status
The funded status represents the economic value of the pension plan. An overfunded
pension plan is an economic asset for the sponsoring company, one that is difficult to
monetize but still an asset. Capital that would have been allocated to the pension plan
can now be put to other uses—reinvestment, share repurchases, debt paydown,
dividends, etc. On the other hand, an underfunded plan can be considered an economic
liability that could result in increased future contributions to the pension plan from the
company, drawing capital away from other parts of the business.
The funded status indicates the health of the pension plan; an overfunded plan is clearly
healthier than an underfunded plan. We start by examining the health of defined benefit
pension plans on a macro level. We combine the pension plan assets and the projected
benefit obligation for each company in the S&P 500 with a defined benefit pension plan.
We then determine the aggregate funded status for the S&P 500 for each of the last 10
years, which is depicted in Exhibit 132.
Exhibit 132: Pension Plan Assets, PBO, and Funded Status for the S&P 500
US$ in billions
$1,400
Pension Assets PBO Funded Status $1,314
$1,300
$1,177 $1,171
$1,200 $1,132 $1,142
$1,100 $1,050 $1,040
$975 $954 $955
$1,000
$867 $877 $884
$900
$800 $755
$719
$700 $652 $639 $658
$500
$400
$300 $248
$222
$200
$113
$61
$100 $13
($10) $8 $99 $10
$0
($100) 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
($200)
($172)
($216)
($300)
Defined benefit pension plans for the companies in the S&P 500 were the picture of
good health in 1999, when they were $248 billion overfunded (128% funded) in total. At
the end of 2003, we see a dramatically different picture, as the plans ended the year
$172 billion underfunded (87% funded). In other words, the health of the pension plans
had deteriorated by $420 billion over that five-year period. What caused the
deterioration? Simple, the pension plan assets could not keep pace with the pension
obligations. Note that in Exhibit 133, from 1999 to 2003 the pension plan assets grew by
$10 billion, a compound annual growth rate of less than 1%, while the pension
obligations grew by $430 billion, a compound annual growth rate of roughly 10%.
125
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 133: Pension Plan Assets, Obligations, and Funded Status for the S&P 500
US$ in billions
Pension Plan Stats 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
Plan Assets (US$ in billions) $550 $544 $652 $719 $868 $975 $1,132 $1,177 $1,050 $955 $1,142
Projected Benefit Obligation (PBO) (US$ in billions) $560 $536 $639 $658 $755 $877 $884 $954 $1,040 $1,171 $1,314
Funded Status $ (US$ in billions) (Under) Over ($10) $8 $13 $61 $113 $99 $248 $222 $10 ($216) ($172)
Funded Status % 98% 102% 102% 109% 115% 111% 128% 123% 101% 82% 87%
Source: Company data, CSFB estimates.
Exhibit 134: Breakdown of Pension Assets across Industries Exhibit 135: Breakdown of PBO across Industries
US$ in billions, unless otherwise stated US$ in billions, unless otherwise stated
Source: Company data, CSFB estimates. Source: Company data, CSFB estimates.
The 10 largest pension plans among the companies in the S&P 500 are listed in Exhibit
136. These 10 plans account for 38% of the total plan assets and 36% of the total PBO
for the S&P 500.
126
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 136: 10 Largest Defined Benefit Pension Plans in the S&P 500 Based on 2003 PBO
US$ in millions
Exhibit 137 shows the number of companies in the S&P 500 with either over- or
underfunded defined benefit pension plans in each year from 1993 through 2003.
Exhibit 137: Number of Companies with Over/Underfunded Defined Benefit Pension Plans
400
335 319 350
300
254
238 237
226 250
195 187
166 171 200
168
140 142 142 121 144
122 150
110
96 88 100
35 51
50
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
• Fewer companies had underfunded pension plans in the S&P 500 in 1999 than at any
other time in the last 10 years. However, since 1999 the pendulum has swung in the
opposite direction; the number of companies with underfunded plans had almost
quadrupled to 335 by the end of 2002, the largest number of underfunded pension
plans in the S&P 500 in the past 10 years. In 2003, the overall health of plans
improved a bit, as 319 companies finished off the year with underfunded plans—
despite the slight improvement, companies with underfunded plans still represented
86% of all companies with defined benefit plans in the S&P 500.
Focusing on 2003, the level of pension funding by company runs the gamut, from
Sherwin-Williams with the healthiest pension plan at 210% funded, to KLA-Tencor at the
other end of the spectrum with a pension plan at only 15% funded. However, there are
some concentrations in funded status, as is made clear in Exhibit 138, which contains a
distribution of the funded status of the pension plans of the companies in the S&P 500.
127
The Magic of Pension Accounting, Part III 07 February 2005
For example, 164, or 44% of the companies that offer defined benefit plans, have
pension plans that are 70-89% funded.
100
90
80 74
# of Companies
60 57 55
51
40
25
20
10
0
<50% 50% to 59% 60% to 69% 70% to 79% 80% to 89% 90% to 99% >100%
Funded Status
Underfunded
Of the 311 companies in the S&P 500 with underfunded defined benefit pension plans
at the end of 2003, 241 were less than 90% funded for two consecutive years. Exhibit
139 includes the 10 companies with the most underfunded pension plans in terms of
percent funded.
Exhibit 139: Most Underfunded Based on Plan Assets/PBO: <50% Funded in 2003
US$ in millions
At the end of 2003, 44 companies were underfunded by over $1 billion, including the 17
companies in Exhibit 140 that were underfunded by over $2 billion. Only six companies
were underfunded by more than $2 billion in each of the past three years—Ford Motor,
Exxon Mobil, General Motors, Delta Air Lines, Delphi, and United Technologies.
128
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 140: Defined Benefit Pension Plans Underfunded by at Least $2 Billion in 2003
US$ in millions
Overfunded
Of the 51 companies with overfunded defined benefit pension plans at the end of 2003,
nine were over 120% funded; the most overfunded plans are included in Exhibit 141. At
210% funded, Sherwin-Williams has the highest percentage funded status in the S&P
500.
The pension plans for the five companies in Exhibit 142 were overfunded by at least $1
billion at the end of 2003 compared with 43 companies in 1999.
129
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 142: Defined Benefit Pension Plans Overfunded by at Least $1 Billion in 2003
US$ in millions
130
The Magic of Pension Accounting, Part III 07 February 2005
Quality of Earnings
Another reason pension accounting causes so much confusion is that most investors
view a defined benefit pension plan as a retirement benefit provided to the employees
that has a cost to the company. Therefore, they believe that cost should result in the
company reporting an expense on its income statement. In the aggregate, companies in
the S&P 500 reported net pension expense from 1991 through 1998. The tide shifted in
1999 as S&P 500 companies started to, in aggregate, report pension income through
2001. Then it shifted again, with the S&P 500 companies reporting, in aggregate,
pension expense beginning in 2002. Exhibit 143 shows the evolution of total pretax net
pension cost from 1993 through 2003.
Exhibit 143: Pretax Net Pension Cost for the S&P 500, 1993–2003
US$ in billions
$35
$30 $28
$25
Pension (Income)/Expense
$20
$15
$12
$10 $9 $9
$10 $7
$6
$5 $3
$0
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
($5)
($15) ($12)
131
The Magic of Pension Accounting, Part III 07 February 2005
300
260 266
245 253 253
243 235 235
250
203
193
200
148 149
150
100 94
100 91
54 54 67
43 53 45
50
0
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003
In Exhibit 145, we find that over 80% of the companies in the S&P 500 report pension
cost that is in-line with the funded status of their defined benefit pension plans; 290
companies with underfunded plans reported pension expense, while 21 companies with
overfunded plans reported pension income. Note that this relationship does not always
hold, as 24 companies with underfunded pension plans reported pension income in
2003 and 29 companies with overfunded plans reported pension expense.
Exhibit 145: Funded Status of Companies Reporting 2003 Pretax Net Pension
(Income)/Expense
21
The 10 companies with the largest amounts of pension income during 2003 are listed in
Exhibit 146 alongside the 10 with the largest amounts of pension expense. If you
wonder why we continue to put brackets around pension income when you are used to
seeing brackets around losses, it’s because that’s how most companies report it.
132
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 146: Companies Reporting the Largest Pretax Net Pension (Income) or Expense in 2003
US$ in millions
Net Income
Impact of After-Tax Net Pension Income on the Company’s Bottom Line
For each company in the S&P 500, we estimate the after-tax pension cost using a 35%
tax rate for all companies. We then compare the after-tax pension cost to reported net
income from continuing operations (excludes discontinued operations, extraordinary
items, and the cumulative effect of a change in accounting principle). The 11 companies
in Exhibit 147 relied on pension income for over 5% of their net income in 2003.
Exhibit 147: Deriving over 5% of 2003 Net Income from After-Tax Pension (Income)
% of Net Income from Pension (Income)
Company Ticker 2003 2002 2001
Pactiv Corp PTV (21%) (32%) (45%)
Consolidated Edison Inc ED (19%) (24%) (23%)
Xcel Energy Inc XEL (9%) N/M (6%)
FPL Group Inc FPL (9%) (10%) (9%)
Bellsouth Corp BLS (9%) (20%) (20%)
Prudential Financial Inc PRU (8%) (100%) N/M
Textron Inc TXT (7%) (18%) (38%)
Intl Business Machines Corp IBM (7%) (15%) (12%)
Boeing Co BA (6%) (11%) (21%)
General Mills Inc GIS (6%) (12%) (7%)
Unisys Corp UIS (6%) (42%) N/M
Source: Company data, CSFB estimates.
• We estimate that over 15% of the net income for Pactiv and Consolidated Edison has
come from their pension plans in each of the last three years.
• The contribution of after-tax pension income to net income has been on the decline in
the last several years. When we ran this analysis using 2001 data, we found 19
companies where pension income accounted for over 20% of net income. For
example, Pactiv’s after-tax pension income represented 45% of its net income in 2001
but declined to 32% in 2002 and 21% in 2003.
133
The Magic of Pension Accounting, Part III 07 February 2005
Pension expense represented at least 10% of net income for 74 companies, including
the 20 companies in Exhibit 148 where it was over 30%.
Exhibit 148: After-Tax Pension Expense Exceeds 30% of 2003 Net Income
Pension Expense/(Income) % of Net Income
Company Ticker 2003 2002 2001
Boise Cascade Corp BCC 292% 171% N/M
Williams Cos Inc WMB 170% N/M 1%
Goodrich Corp GR 157% 15% 3%
NCR Corp NCR 118% (38%) (36%)
General Motors Corp GM 77% 68% 59%
Transocean Inc RIG 77% N/M 2%
Cummins Inc CMI 73% 17% N/M
Georgia-Pacific Corp GP 68% N/M N/M
Ford Motor Co F 67% 22% N/M
Symbol Technologies SBL 65% N/M N/M
Ashland Inc ASH 60% 37% 7%
Phelps Dodge Corp PD 55% N/M N/M
Northrop Grumman Corp NOC 46% (9%) (51%)
Cooper Tire & Rubber CTB 41% 19% 95%
Maytag Corp MYG 40% 27% 16%
Raytheon Co RTN 38% (7%) (3,718%)
Ryder System Inc R 38% 17% (3%)
Du Pont (E I) De Nemours DD 36% (8%) (6%)
Lockheed Martin Corp LMT 30% (20%) (291%)
Verizon Communications VZ 30% (24%) (204%)
Source: Company data, CSFB estimates.
• Strip out pension expense and net income more than doubles for four companies in
2003—Boise Cascade, Williams Cos., Goodrich, and NCR.
Operating Income
Impact of Pretax Net Pension Income on a Company’s Operating Income
Pension cost is reported in operating income wherever the company reports labor cost;
for example, cost of goods sold or SG&A. In this section, we further examine the impact
that pension accounting has had on operating income (as defined by Compustat) for the
companies in the S&P 500. Exhibit 149 includes the 11 companies where 5% or more of
operating income was pension related during 2003.
134
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 149: Companies Deriving 5% or More of 2003 Operating Income from Pension
Income
% of Operating Income from Pension Income
Company Ticker 2003 2002 2001
Meadwestvaco Corp MWV (32%) (26%) (41%)
Qwest Communication Intl Inc Q (26%) (74%) (16%)
Pactiv Corp PTV (14%) (24%) (28%)
Avaya Inc AV (12%) N/M 131%
Consolidated Edison Inc ED (11%) (18%) (16%)
FPL Group Inc FPL (8%) (9%) (8%)
Intl Business Machines Corp IBM (8%) (14%) (16%)
Bellsouth Corp BLS (8%) (13%) (12%)
Xcel Energy Inc XEL (6%) (6%) (4%)
Prudential Financial Inc PRU (6%) (80%) (109%)
Unisys Corp UIS (5%) (34%) (63%)
Source: Company data, CSFB estimates.
• Five companies would have seen operating income slashed by over 10% in 2003 if
not for pension income. For example, Meadwestvaco reported 2003 operating income
of $199 million and pretax net pension income of $64 million. By stripping pretax net
pension income out of operating income in 2003, Meadwestvaco would be left with
$135 million, which is 32% lower than reported.
Exhibit 150 includes the 13 companies where pension expense represented 30% or
more of operating income during 2003.
• Rowan, Navistar International, and Goodyear Tire & Rubber would have reported
operating income in 2003 that would have been more than twice as high if not for
pension expense.
135
The Magic of Pension Accounting, Part III 07 February 2005
Earnings Headwind
Even as pension plans got healthier in 2003, the powerful smoothing mechanisms built
into pension accounting worked their magic, driving pension costs up and earnings
down. Between 2002 and 2003, net income from continuing operations for the S&P 500
increased by $202 billion. In the aggregate, pretax pension costs increased by $25
billion ($17 billion after tax), from $3 billion in 2002 to $28 billion in 2003, slowing the
growth in earnings for most companies with defined benefit pension plans. Extending
our analysis a bit, we assess the extent to which higher pension expense or lower
pension income put pressure on company earnings.
Note that of the 17 companies in Exhibit 151, we found that the pension plans’ health,
its funded status, actually improved between 2002 and 2003 for 11 of them. This
anomaly provides another example of the disconnect between pension accounting and
the economics of the pension plan: as the plans grew stronger, pension costs
increased.
To measure whether or not the change in pension cost was meaningful to a particular
company’s earnings, we first compare the earnings headwind from the pension plan in
2003 to 2002 reported earnings. We found that the earnings headwind created a drag of
at least 5% for 89 companies and at least 10% for 50 companies, including the 20
companies in Exhibit 152 that suffered an earnings drag of at least 20%.
136
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 152: Pension Related Earnings Headwind of at least 20% of Diluted EPS
A B
Pension Cost/Share Diluted Earnings A/B
Increase between per Share Earnings
Company Ticker 2002 and 2003 2002 Drag
United States Steel Corp X 3.45 0.62 556%
Northrop Grumman Corp NOC 1.53 0.34 450%
Lockheed Martin Corp LMT 0.93 1.11 84%
RJ Reynolds Tobacco Hldgs RJR 1.06 1.45 73%
General Motors Corp GM 1.81 3.51 52%
Verizon Communications VZ 0.77 1.49 51%
Penney (J C) Co JCP 0.12 0.26 47%
Boeing Co BA 0.27 0.61 44%
Weyerhaeuser Co WY 0.45 1.09 42%
Temple-Inland Inc TIN 0.40 1.02 40%
Delphi Corp DPH 0.22 0.61 36%
Unisys Corp UIS 0.24 0.69 35%
Ryder System Inc R 0.50 1.50 33%
Cummins Inc CMI 0.65 2.13 31%
Raytheon Co RTN 0.62 2.04 30%
Principal Financial Grp Inc PFG 0.11 0.41 26%
Goodrich Corp GR 0.28 1.14 25%
Deere & Co DE 0.29 1.33 22%
Eastman Chemical Co EMN 0.17 0.79 21%
Pactiv Corp PTV 0.18 0.92 20%
Source: Company data, CSFB estimates.
Earnings Tailwind
On the other hand, pension accounting is still providing an earnings tailwind for some
companies. We found that 29 companies experienced an increase in pension income or
a drop in pension expense between 2002 and 2003, helping to fuel earnings growth.
Focusing on the decrease in pension cost on a per share basis, we found that for eight
companies, pension cost dropped by more than $0.05 per share, including three
companies—American Standard, Hershey Foods, and Ameren—where pension cost
dropped by more than $0.10 per share.
Exhibit 153 highlights the eight companies where over 5% of the growth in net income
between 2002 and 2003 can be attributed to lower pension costs.
137
The Magic of Pension Accounting, Part III 07 February 2005
Exhibit 153: Year-over-Year Growth in 2003 Net Income Attributable to Decrease in Pension
Cost (Arising from a Rise in Pension Income or a Decline in Pension Expense)
US$ in millions, unless otherwise stated
Key Observations
• Let’s examine the figures for Donnelley & Sons more closely to illustrate this analysis.
Donnelley & Sons reported after-tax net pension income of $1 million in 2003 versus
$8 million of pension expense in 2002. That $9 million decline in pension cost
accounts for approximately 27% of the company’s $34 million year-over-year rise in
net income. If pension expense had remained constant at $8 million in 2003, net
income would have increased by only $25 million, a 17.6% growth rate versus the
23.9% growth rate in reported earnings.
• It is not just companies that report pension income that can have a benefit to earnings
from the pension plan. Companies that report pension expense can also benefit. For
example, Cinergy’s net income increased by $38 million year over year. Its reported
net pension expense declined by $4 million, from $44 million in 2002 to $40 million in
2003. If Cinergy’s pension expense had remained at $44 million during 2003, its net
income would have increased by only $34 million. In other words, the decline in
pension expense accounted for 11% of the growth in Cinergy’s net income.
138
The Magic of Pension Accounting, Part III 07 February 2005
Appendix A
Exhibit 154: S&P 500 Companies That Don't Have a Defined Benefit Pension Plan
Company Ticker Company Ticker Company Ticker
ADC Telecommunications Inc ADCT Equity Office Properties Tr EOP Nvidia Corp NVDA
Adobe Systems Inc. ADBE Equity Residential EQR Omnicom Group OMC
Advanced Micro Devices AMD Express Scripts Inc ESRX Oracle Corp ORCL
Alberto-Culver Co -Cl B ACV Family Dollar Stores FDO Paychex Inc PAYX
Altera Corp ALTR Federated Investors Inc FII Peoplesoft Inc PSFT
American Pwr Cnvrsion APCC Fiserv Inc FISV PMC-Sierra Inc PMCS
Amgen Inc AMGN Forest Laboratories -Cl A FRX Price (T. Rowe) Group TROW
Apartment Invt & Mgmt -Cl A AIV Gap Inc GPS Progressive Corp-Ohio PGR
Apollo Group Inc -Cl A APOL Gateway Inc GTW Prologis PLD
Apple Computer Inc AAPL Golden West Financial Corp GDW Providian Financial Corp PVN
Applied Materials Inc AMAT Grainger (W W) Inc GWW Pulte Homes Inc PHM
Applied Micro Circuits Corp AMCC Harrahs Entertainment Inc HET Qlogic Corp QLGC
AT&T Wireless Services Inc AWE HCA Inc HCA Qualcomm Inc QCOM
Autodesk Inc ADSK Health Management Assoc HMA Quest Diagnostics Inc DGX
Autonation Inc AN Home Depot Inc HD Radioshack Corp RSH
Bear Stearns Companies Inc BSC Humana Inc HUM Reebok International Ltd RBK
Bed Bath & Beyond Inc BBBY Intl Game Technology IGT Robert Half Intl Inc RHI
Best Buy Co Inc BBY Intuit Inc INTU Sanmina-Sci Corp SANM
Biomet Inc BMET Janus Capital Group Inc JNS Siebel Systems Inc SEBL
Block H & R Inc HRB JDS Uniphase Corp JDSU Simon Property Group Inc SPG
BMC Software Inc BMC KB Home KBH Solectron Corp SLR
Boston Scientific Corp BSX King Pharmaceuticals Inc KG Southwest Airlines LUV
Broadcom Corp -Cl A BRCM Kohls Corp KSS Staples Inc SPLS
Calpine Corp CPN Limited Brands Inc LTD Starbucks Corp SBUX
Capital One Finl Corp COF Linear Technology Corp LLTC Sun Microsystems Inc SUNW
Carnival Corp CCL Liz Claiborne Inc LIZ Sungard Data Systems Inc SDS
Centex Corp CTX Lowes Cos LOW Symantec Corp SYMC
Charter One Finl Inc CF LSI Logic Corp LSI Synovus Financial Cp SNV
Ciena Corp CIEN Marriott Intl Inc MAR Tellabs Inc TLAB
Cintas Corp CTAS Marshall & Ilsley Corp MI Tenet Healthcare Corp THC
Cisco Systems Inc CSCO Maxim Integrated Products MXIM Toys R Us Inc TOY
Citrix Systems Inc CTXS MBIA Inc MBI Union Planters Corp UPC
Clear Channel Communications CCU McDonalds Corp MCD Unitedhealth Group Inc UNH
Computer Associates Intl Inc CA Medimmune Inc MEDI Univision Communications Inc UVN
Compuware Corp CPWR Mercury Interactive Corp MERQ Veritas Software Co VRTS
Comverse Technology Inc CMVT Micron Technology Inc MU Walgreen Co WAG
Concord EFS Inc CE Microsoft Corp MSFT Wal-Mart Stores WMT
Costco Wholesale Corp COST Monster Worldwide Inc MNST Watson Pharmaceuticals Inc WPI
Dell Computer Corp DELL Network Appliance Inc NTAP Winn-Dixie Stores Inc WIN
Deluxe Corp DLX Nextel Communications NXTL Xilinx Inc XLNX
Dollar General Corp DG Nike Inc -Cl B NKE XL Capital Ltd XL
Ebay Inc EBAY Novell Inc NOVL Yahoo Inc YHOO
Electronic Arts Inc ERTS Novellus Systems Inc NVLS
EOG Resources Inc EOG Nucor Corp NUE
139
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B
Exhibit 155: Historical and Projected Funded Status for the S&P 500
US$ in millions
140
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
141
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
142
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
143
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
144
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
145
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
146
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
147
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
148
The Magic of Pension Accounting, Part III 07 February 2005
Appendix B continued
149
The Magic of Pension Accounting, Part III 07 February 2005
150
The Magic of Pension Accounting, Part III 07 February 2005
151
The Magic of Pension Accounting, Part III 07 February 2005
152
The Magic of Pension Accounting, Part III 07 February 2005
153
The Magic of Pension Accounting, Part III 07 February 2005
154
The Magic of Pension Accounting, Part III 07 February 2005
155
The Magic of Pension Accounting, Part III 07 February 2005
156
The Magic of Pension Accounting, Part III 07 February 2005
Disclosure Appendix
Important Global Disclosures
David Zion, CFA, CPA & Bill Carcache, CPA each certify, with respect to the companies or securities that
he or she analyzes, that (1) the views expressed in this report accurately reflect his or her personal views
about all of the subject companies and securities and (2) no part of his or her compensation was, is or will
be directly or indirectly related to the specific recommendations or views expressed in this report.
The analyst(s) responsible for preparing this research report received compensation that is based upon
various factors including CSFB's total revenues, a portion of which are generated by CSFB's investment
banking activities.
157
The Magic of Pension Accounting, Part III 07 February 2005
158
Disclaimers
This report is not directed to, or intended for distribution to or use by, any person or entity who is a citizen or resident of or located in any locality, state, country or other
jurisdiction where such distribution, publication, availability or use would be contrary to law or regulation or which would subject Credit Suisse First Boston or its
subsidiaries or affiliates (collectively "CSFB") to any registration or licensing requirement within such jurisdiction. All material presented in this report, unless specifically
indicated otherwise, is under copyright to CSFB. None of the material, nor its content, nor any copy of it, may be altered in any way, transmitted to, copied or distributed to
any other party, without the prior express written permission of CSFB. All trademarks, service marks and logos used in this report are trademarks or service marks or
registered trademarks or service marks of CSFB.
The information, tools and material presented in this report are provided to you for information purposes only and are not to be used or considered as an offer or the
solicitation of an offer to sell or to buy or subscribe for securities or other financial instruments. CSFB may not have taken any steps to ensure that the securities referred
to in this report are suitable for any particular investor. CSFB will not treat recipients as its customers by virtue of their receiving the report. The investments or services
contained or referred to in this report may not be suitable for you and it is recommended that you consult an independent investment advisor if you are in doubt about
such investments or investment services. Nothing in this report constitutes investment, legal, accounting or tax advice or a representation that any investment or strategy
is suitable or appropriate to your individual circumstances or otherwise constitutes a personal recommendation to you. CSFB does not offer advice on the tax
consequences of investment and you are advised to contact an independent tax adviser. Please note in particular that the bases and levels of taxation may change.
CSFB believes the information and opinions in the Disclosure Appendix of this report are accurate and complete. Information and opinions presented in the other sections
of the report were obtained or derived from sources CSFB believes are reliable, but CSFB makes no representations as to their accuracy or completeness. Additional
information is available upon request. CSFB accepts no liability for loss arising from the use of the material presented in this report, except that this exclusion of liability
does not apply to the extent that liability arises under specific statutes or regulations applicable to CSFB. This report is not to be relied upon in substitution for the exercise
of independent judgment. CSFB may have issued, and may in the future issue, a trading call regarding this security. Trading calls are short term trading opportunities based
on market events and catalysts, while stock ratings reflect investment recommendations based on expected total return over a 12-month period relative to the relevant
coverage universe. Because trading calls and stock ratings reflect different assumptions and analytical methods, trading calls may differ directionally from the stock rating. In
addition, CSFB may have issued, and may in the future issue, other reports that are inconsistent with, and reach different conclusions from, the information presented in this
report. Those reports reflect the different assumptions, views and analytical methods of the analysts who prepared them and CSFB is under no obligation to ensure that such
other reports are brought to the attention of any recipient of this report. CSFB and its affiliate companies are involved in many businesses that relate to companies mentioned
in this report. These businesses include specialized trading, risk arbitrage, market making, and other proprietary trading. CSFB may, to the extent permitted by law, act upon
or use the information or opinions presented herein, or the research or analysis on which they are based, before the material is published.
Past performance should not be taken as an indication or guarantee of future performance, and no representation or warranty, express or implied, is made regarding
future performance. Information, opinions and estimates contained in this report reflect a judgement at its original date of publication by CSFB and are subject to change
without notice. The price, value of and income from any of the securities or financial instruments mentioned in this report can fall as well as rise. The value of securities
and financial instruments is subject to exchange rate fluctuation that may have a positive or adverse effect on the price or income of such securities or financial
instruments. Investors in securities such as ADR’s, the values of which are influenced by currency volatility, effectively assume this risk.
Structured securities are complex instruments, typically involve a high degree of risk and are intended for sale only to sophisticated investors who are capable of
understanding and assuming the risks involved. The market value of any structured security may be affected by changes in economic, financial and political factors
(including, but not limited to, spot and forward interest and exchange rates), time to maturity, market conditions and volatility, and the credit quality of any issuer or
reference issuer. Any investor interested in purchasing a structured product should conduct their own investigation and analysis of the product and consult with their own
professional advisers as to the risks involved in making such a purchase.
Some investments discussed in this report have a high level of volatility. High volatility investments may experience sudden and large falls in their value causing losses
when that investment is realised. Those losses may equal your original investment. Indeed, in the case of some investments the potential losses may exceed the amount
of initial investment, in such circumstances you may be required to pay more money to support those losses. Income yields from investments may fluctuate and, in
consequence, initial capital paid to make the investment may be used as part of that income yield. Some investments may not be readily realisable and it may be difficult
to sell or realise those investments, similarly it may prove difficult for you to obtain reliable information about the value, or risks, to which such an investment is exposed.
This report may provide the addresses of, or contain hyperlinks to, websites. Except to the extent to which the report refers to CSFB’s own website material, CSFB has
not reviewed the linked site and takes no responsibility for the content contained therein. Such address or hyperlink (including addresses or hyperlinks to CSFB’s own
website material) is provided solely for your convenience and information and the content of the linked site does not in any way form part of this document. Accessing
such website or following such link through this report or CSFB’s website shall be at your own risk.
This report is issued and distributed in Europe (except Switzerland) by Credit Suisse First Boston (Europe) Limited, One Cabot Square, London E14 4QJ, England, which
is regulated in the United Kingdom by The Financial Services Authority (“FSA”). This report is being distributed in the United States by Credit Suisse First Boston LLC; in
Switzerland by Credit Suisse First Boston; in Canada by Credit Suisse First Boston Canada Inc.; in Brazil by Banco de Investimentos Credit Suisse Boston S.A.; in Japan
by Credit Suisse First Boston Securities (Japan) Limited; elsewhere in Asia/Pacific by whichever of the following is the appropriately authorised entity in the relevant
jurisdiction: Credit Suisse First Boston (Hong Kong) Limited, Credit Suisse First Boston Australia Equities Limited, Credit Suisse First Boston (Thailand) Limited, CSFB
Research (Malaysia) Sdn Bhd, Credit Suisse First Boston Singapore Branch and elsewhere in the world by the relevant authorised affiliate of the above. Research on
Taiwanese securities produced by Credit Suisse First Boston, Taipei Branch has been prepared by a registered Senior Business Person.
In jurisdictions where CSFB is not already registered or licensed to trade in securities, transactions will only be effected in accordance with applicable securities legislation,
which will vary from jurisdiction to jurisdiction and may require that the trade be made in accordance with applicable exemptions from registration or licensing
requirements. Non-U.S. customers wishing to effect a transaction should contact a CSFB entity in their local jurisdiction unless governing law permits otherwise. U.S.
customers wishing to effect a transaction should do so only by contacting a representative at Credit Suisse First Boston LLC in the U.S.
Please note that this report was originally prepared and issued by CSFB for distribution to their market professional and institutional investor customers. Recipients who
are not market professional or institutional investor customers of CSFB should seek the advice of their independent financial advisor prior to taking any investment
decision based on this report or for any necessary explanation of its contents. This research may relate to investments or services of a person outside of the UK or to
other matters which are not regulated by the FSA or in respect of which the protections of the FSA for private customers and/or the UK compensation scheme may not be
available, and further details as to where this may be the case are available upon request in respect of this report.
Copyright Credit Suisse First Boston, and its subsidiaries and affiliates, 2005. All rights reserved.
ASIA/PACIFIC: +852 2101-6000 EUROPE: +44 (20) 7888-8888 UNITED STATES OF AMERICA: +1 (212) 325-2000
AT0084.doc
AT0084.doc