We feel a little sheepish coming back so soon to accounting for pensions and other post-retirement benefits, but we keep finding new reasons to condemn SFAS 87 and 106.One recent spur is the confusion surrounding the bankruptcy of Delphi, the former subsidiary of General Motors. Among the key issues in this morass is the ultimate disposition of the company's pension and OPEB liabilities. The situation is complicated because GM guaranteed an as-yet-unannounced portion of Delphi's liabilities, while Delphi indemnified GM against any loss from that guarantee. (Yes, they each agreed to cover the others' position!) Caught in the middle are the employees, represented by the United Auto Workers, trying desperately to hold onto jobs and benefits. Adding to the intrigue is the Securities and Exchange Commission's subpoena of GM's documents about its pension activities.
Despite the compelling need for it, there is little useful public information about Delphi's pension and OPEB plans, because the financial statement numbers have no connection with real amounts that are relevant for sorting out the uncertain future for Delphi and GM.
Whose fault is it?
Once again, we point in the direction of the Financial Accounting Standards Board for tolerating these standards' painfully obvious flaws. They were concocted during the 1980s under political pressure when preparers were not just bellicose about keeping useful information off the financial statements, but also the source of most of the board's funding contributions. Out of this cauldron, generally accepted accounting principles for pensions and OPEB were cast in a weird configuration that still haunts today's capital markets.
The situation at Delphi - and elsewhere - proves that a new resolution is needed right away. In addition, the Office of the Chief Accountant issued a report last summer that soundly criticized GAAP for causing useful information about pensions and OPEB to remain off the balance sheet, and went on to suggest consolidating the statements of the parent and the funding entities.
With its decision on November 10 to put pension accounting on its agenda, it is clear that FASB has also seen the potential for a calamitous collapse. If the system were to crumble at a time when SFAS 87 was in effect, it wouldn't be just our finger pointing in the board's direction. Rather, everyone would be asking where the accountants were when the house of cards fell. We don't think many would be moved by an explanation that SFAS 87 was the best political compromise that could be forged 20 years ago.
Bass-O-Matic accounting
One of the all-time memorable sketches from Saturday Night Live involves running a fish through the "Bass-O-Matic" (a poorly disguised blender) and converting it to a nasty liquid. The revulsion is magnified when the announcer proclaims that this product eliminates the tiresome tasks of cleaning and cooking fish.
What we have come to comprehend is that GAAP for pensions and OPEB is Bass-O-Matic accounting. That is, all the plans' economic elements are figuratively thrown into a blender and processed so thoroughly that the truth is neither recognizable nor consumable, even by reading the footnotes.
As we see it, the reported annual cost aggregates so many real and nonsensical components that it's incomprehensible to all but a few:
* Service cost - the only authentic operating expense associated with these plans.
* Interest cost - a financing charge that's disguised as an operating expense.
* Expected return on plan assets - an imaginary result of investing that is offset against operating costs.
* Plan amendments - costs from increased benefits for past work are deferred and amortized as if there is a kind of employee goodwill; if benefits are cut (as they have been for most OPEB plans), the savings are also deferred and amortized, presumably to offset higher costs generated by disgruntled employees.
* Actuarial gains/losses - the obligations are so uncertain in amount (but not existence) that actuaries must estimate their market values; because changing conditions alter the real value, new estimates are usually needed every year; rather than confronting the fact that real volatility results from these unconscionably risky promises, the preparers bullied FASB into sheltering reported income from its effects.
* Corridor amortization - FASB basically assumed that unexpected asset returns and actuarial adjustments in some years would be counteracted in later years, thus hiding the volatility that clearly signals that large risks exist. So-called corridor amortization merely chips away at unreversed deferred gains and losses. Unfortunately, low interest rates and meager investment returns have overwhelmed this safety net, and most companies now have huge and growing deferred losses.
In addition to blending everything into a single, undecipherable expense, these standards also create deceptive balance sheets by offsetting assets against obligations and then blending the deferred debits and credits with that amount in a way that completely obscures the real situation.
How bad is it?
To see how bad the situation is, consider these 2004 numbers for the four largest U.S. companies with defined-benefit and OPEB plans that compare the real economic numbers with those that appear on the companies' balance sheets:
* ExxonMobil: Real net liability - $16.4 billion; recognized net liability - $6.0 billion.
* GM: Real net liability - $70.0; recognized net asset - $6.7 billion.
* Ford: Real net liability - $44.7 billion; recognized net liability - $12.9 billion.
* GE: Real net liability - $3.5 billion; recognized net asset - $10.7 billion.
For all four, the real combined net liability is $134.6 billion, but their balance sheets report a net debt of only $1.5 billion. Now, that's bad!
Garbage in, garbage out
The inevitable consequence of Bass-O-Matic accounting is that no one has a useful platform for making decisions or taking action.
Consider analysts who use EBITDA to assess a company's operating cash flows. If they use reported pension/OPEB expense in the earnings, they're using a number that is both unreliable and a nonsensical aggregation that includes interest and amortization.
Consider policy-makers who must decide whether to reform ERISA and its funding requirements. How can they develop an informed position if they use numbers derived from generally accepted accounting principles? If they were to look at the results above, it would appear that the big four are essentially fully funded, instead of being short by nearly $135 billion. No one can solve a problem they don't see.
Consider labor representatives charged with protecting their union members' benefits and job security. How can they get a handle on the real situation when GAAP accounting is so distorted?
Consider managers who are supposed to avoid being underfunded while coping with volatile real costs, investments and liabilities. In our heart of hearts, we hope that they don't actually believe the GAAP numbers that hide far more facts than they reveal.
What needs to come from FASB?
Now that FASB has added pensions to its agenda, we only hope that its members have the insight and courage to make immediate and radical changes in pension accounting.
The time for slow and cautious consideration has passed. The Bass-O-Matic standards must be replaced with simple rules along these lines: Report what happens, when it happens, with total transparency and openness. Volatile income and balance sheet stress are the real consequences of the risky practices of deferring unknown amounts of compensation while investing in marketable securities. Financial statements should be nonvolatile if and only if what they purport to describe is not actually volatile.
This organized deception has to stop, now. Our profession has nothing to gain and much to lose from our association with fishy financial statements that don't tell the truth.
