Anyone who believes financial statements ought to provide truthful and useful information should be troubled by pension and other post-employment benefits accounting. The political process that produced SFAS 87 and 106 nearly three decades ago was among the worst in the history of the Financial Accounting Standards Board. Once the board suggested it favored truthful reporting for these deferred benefits, corporate constituents pounced, threatening to cut off contributions. Ever since FASB's unfortunate capitulation, financial statements have concealed truth, instead of revealing it.
The best defense that the members could offer was they achieved a "worthwhile improvement in financial reporting." Surely today's FASB knows that only a giant leap forward is good enough.
In 2005-6, the board created fleeting hope for reform by issuing SFAS 158 over managers' objections. Funded status now appears as a net liability or asset, while remnants of smoothing are reported as other comprehensive income in equity. FASB intimated that a second phase would fix the income manipulations, but paused to allow the International Accounting Standards Board to produce its standard. Alas, it didn't come out until 2011, and still left bad accounting in place.
OUR PREMISES
It doesn't take deep insight to see that finishing this task ought to be at the top of FASB's agenda. One company after another faces a seemingly unforeseen need to infuse its benefit plans with billions when the cash is urgently needed elsewhere. In effect, money that could be used to hire new workers and create profits goes to retirees. Better accounting would help eliminate surprises and avoid funding crises by divulging relevant facts that today's GAAP hides.
Our analysis is rooted in these premises:
• FASB is now unfettered and free to completely redo accounting in this area;
• FASB is not obliged to converge with the IASB's solution;
• Existing GAAP has nothing worth retaining; and,
• The new standard can be very simple.
THE PROBLEM
GAAP for retirement benefits became densely complex because of FASB's political vulnerability and a split based on members' conceptual differences. Some clung to matching, while others preferred the asset/liability theory.
Despite this cleavage, the members unanimously disliked the outside pressure and wanted to be rid of it. The outcome was badly compromised standards that keep useful truth out of financial statements through smoothing, offsetting and aggregating.
Smoothing occurs by deferring income effects of amendments, liability value changes and unexpected asset returns. Offsetting takes place between assets and liabilities and between income and expenses; also, current gains and losses are offset against past and hypothetical future losses and gains. Aggregating creates an annual GAAP cost that blends diverse income items into a single indecipherable mishmash labor cost.
In addition, comparable results aren't produced by uniformly implementing what then-FASB Chairman Bob Herz called a "Rube Goldberg" standard. In fact, it's impossible to rationally compare GAAP numbers between companies or for one company over several years.
ASSESSING THE DAMAGES
Useful information is lost by offsetting debt with collateral. First, it obscures the liability-based risk that is associated with this immense debt and largely beyond management's influence. Second, offsetting cloaks the drastically different asset-based risk and inhibits complete assessments of management's performance in earning satisfactory yet safe returns.
In the same way, nothing good is gained and a great deal is sacrificed by aggregating grossly dissimilar income elements into a single meaningless sum. Useful information is clearly not provided by a net benefit cost that is a senseless amalgam of labor cost, interest expense, expected investment returns, and fabricated allocations of past costs and savings. The damage is worsened because this monkey business makes it effectively impossible to assess managers' success in controlling the outcomes. The ultimate consequence is that employees' and shareholders' interests are placed at even greater risk.
Our solution would penetrate the dense fog surrounding pension and other benefit plans and otherwise dispel confusion by breaking these arrangements down into their elemental components. We think any objections will ring hollow because our proposals are already generally accepted for other types of compensation, debt and investments.
A FOUR-PART SOLUTION
Solving this problem is straightforward when the several components of benefit arrangements are accounted for individually, instead of lumping them together into an amorphous hodgepodge.
1. Annual compensation cost. The essence of any benefit plan is compensating employees' present efforts with deferred payments. The debit is obvious: compensation expense equal to the estimated fair value of the deferred amounts. This labor cost is most usefully treated like all other labor costs without offsetting it or combining it with other income items.
2. Benefit liability. Without question, postponing payments creates a liability. Useful accounting will not camouflage the fact that management created a hyper-risky obligation with unknown amounts to be paid to unidentifiable retirees until they die. These obligations are inarguably debt and their full balance belongs on balance sheets among other liabilities. Because uncertainty creates risk and makes precise measurement elusive, they should be continuously reported at their estimated fair value and described with comprehensible disclosures. Because of the volatility created by this debt-based risk, income can be reported usefully only without smoothing. For clarity, interest expense and gains and losses from value changes belong in financing income.
Although there's nothing new in our proposal that isn't already in GAAP, we encourage FASB to issue rules that will force managers to report everything clearly and completely.
3. Funding. By contract or law, corporations earmark assets as collateral against their benefit obligations. Rather than either buying insurance to pay the debt or constructing a portfolio that would hedge the liability and mitigate real volatility, managers create high-risk equity portfolios to minimize current cash outflows and boost expected yields, hoping they'll grow to provide cash when the debts come due. This strategy creates additional risk that spawns more, not less, year-to-year income variability. If reported income doesn't reveal this asset-based volatility, managers aren't held accountable for controlling it. We propose that employers simply report the fund as a stand-alone investment at fair value and present its actual returns in investing income. No smoothing, no offsetting, just plain truth, like everybody does for trading securities.
4. Plan amendments. Managers routinely amend plans to resolve collective bargaining issues and cope with financial exigency. If future cash outflows are increased, the liability should be credited with an equal debit reported among the year's expenses as a current cost of doing business, instead of speciously spreading it over the future as additional labor cost. If an amendment produces smaller future cash outflows, the liability should be debited with an equal credit reported as an immediate financing gain from debt forgiveness, instead of mindlessly spreading these savings over future years as a reduction in labor cost.
FOR EXAMPLE
To show how GAAP produces misleading outcomes, we analyzed data from the most recent annual reports for ExxonMobil, AT&T and IBM. The accompanying table shows the results.
The first two lines of the table show amounts for total liabilities, first under GAAP and then as modified to eliminate asset offsets. It then shows debt-to-equity ratios calculated using both the nominal GAAP numbers and modified debt measures (assuming unchanged equity). In these cases, the ratio increased by approximately 20 percent, 10 percent and 85 percent, revealing risk that was partially concealed. The results also show that GAAP inhibits comparability between companies.
[IMGCAP(1)]The next two lines in the table illustrate how smoothing misrepresents annual reported pension costs. The modified cost equals service cost plus interest less actual return plus/minus actuarial liability changes, plan amendments, and other real events.
The last three lines report the measurement error equal to the difference between the reported and modified costs. The relative size of the error compared to the GAAP cost ranged from very large (approximately 120 percent) to huge (approximately 270 percent) to astronomical (approximately 1,400 percent). We also scaled this error against the companies' reported pretax income. For ExxonMobil, the error was material at 6 percent; for AT&T, it was 55 percent (which means modified income was less than half the reported amount); and for IBM, the error was slightly higher than 20 percent.
The non-comparability of uniformly applying GAAP is manifested twice, first in the significant impact for each company, and second in varied degrees of distortion across companies.
This analysis clearly suggests that GAAP suppresses useful information about risk by smothering facts about the real costs and volatility of benefit plans. This situation should be intolerable to absolutely everyone. That includes management because the capital markets' uncertainty about the real effects surely causes them to impose higher capital costs and bid stock prices lower.
Our efforts also reminded us that current footnotes are frustratingly difficult to decipher, primarily because managers prefer obfuscation over illumination.
OUR PRESCRIPTION
As everywhere else, reporting plain and simple truths will always produce more useful information. GAAP for pension and other benefits falls atrociously short of plain truth and comes nowhere close to simple.
What, then, should FASB do? Here's our prescription:
1. Report service cost as the only current labor cost, interest as a financing cost, and fund returns as investing income;
2. Put the full benefit liability among the debts on the balance sheet and continuously mark it to market;
3. Put fund investments with the assets on the balance sheet and continuously mark them to market;
4. Report all gains and losses on the income statement without deferring, offsetting, aggregating, smoothing or surreptitiously slipping them directly into equity; and,
5. Mandate clear footnotes.
The result will be the truth, the whole truth, and nothing but the truth.
It can't get any simpler, or better, than that.
Paul B. W. Miller is a professor at the University of Colorado at Colorado Springs and Paul R. Bahnson is a professor at Boise State University. The authors' views are not necessarily those of their institutions. Reach them at paulandpaul@qfr.biz.
