By many accounts, it has been a good year for corporate pension funds. For the first time since 1999, assets for most pension portfolios showed positive investment returns, reducing deficits by more than $50 billion. Meanwhile, discount-rate relief passed by Congress in April has lowered required contributions to many underfunded plans at least temporarily. But the question remains: is Corporate America’s $4 trillion-plus pension portfolio any better prepared to weather the next downturn?
A revision of FAS 132 was supposed to answer that question. Thanks to the new regulations, which went into effect last December, companies have had to disclose descriptions of investment strategies and targets, plan-asset allocations, expected benefits payments going forward five years, and estimated cash contributions for the coming fiscal year.
The new disclosures are reassuring, but they are far from conclusive.
On the positive side, corporate pension plans appear adequately diversified. According to Howard Silverblatt, a market equities analyst for Standard & Poor’s, the average S&P 500 company has 64.3 percent of its pension assets in equities, 30.1 percent in fixed income, 4.3 percent in real estate, and 1.8 percent in “other” investments.
Companies are also showing more restraint in projecting expected rates of return from pension portfolios — figures they were charged with inflating during the market meltdown. Jack Ciesielski, publisher of The Analyst’s Accounting Observer, says that 2004 10-Ks show “a big outbreak of rationality” associated with those rates.
FAS 132 will also allow analysts to better question them. “Now when a company says it’s expecting a 10 percent return, you can shoot your mouth off if the plan has 60 percent of its assets in fixed income,” he says.
Others say that FAS 132 barely scratches the surface. The information required “is a lot of fluff,” charges Mark Beilke, director of employee-benefits research at actuarial firm Milliman Inc. “The idea was to show how the payments and liabilities of a plan will unfold over the future — but it failed.” The mandate does not elicit the right kind of information about the types of assets and length of liabilities over the life of a plan, he charges.
With investors and regulators looking closely at pension performance, many believe FAS 132 is just the beginning of a push for more disclosure. Once the door is open and investors realize the impact of pension performance on corporate performance, “they start to ask for more,” says Erol Hakanoglu, managing director at Goldman Sachs’s Capital Market Strategies. He predicts disclosure trends for pensions will mirror those for stock options, with firms gradually providing more details on assets and the durations of liabilities.
One of the biggest complaints against FAS 132 is that the requirements are not specific enough. For example, the four categories in which assets must be grouped — equities, fixed income, real estate, and other — are too vague for analysts to adequately compare expected rates of return against market rates. “If you say you invest in real estate, is that a REIT [real estate investment trust], or a property, or what? And there is no separate disclosure for hedge funds or high-yield bonds, which will be more volatile,” says Hakanoglu. Moreover, there’s little guarantee that what is reported at year-end will hold true during the year.
Neither does FAS 132 clarify the discount-rate assumptions used to calculate pension liabilities — which companies have long been required to disclose but not explain. The “right” discount rate would be one that reflects zero-coupon bonds coming due at the same time as pension payouts. But most firms simply extrapolate that number based on the corporate AA bond rate. “It’s not clear that they’re picking a rate that is really tied to the characteristics of their plan,” says Mike Johnston, leader of Hewitt Associates’s retirement and financial-management practice. The resulting pension income or expense is a noncash item, but it affects net income. DuPont, for example, would face an additional $60 million in pension expenses if it were to lower its discount rate by half a percentage point, according to its recent 10-K.
What’s missing, says Johnston, is a requirement to reveal the duration of the liabilities, or how quickly the company must make good on its obligations. The mandate to show projected benefit payouts for each of the next 5 years and a total estimate for the next 10 years “doesn’t define what the composition of the liabilities are or what the cash call on the company might be,” he adds.
Analysts are also skeptical about the projected cash contributions that companies now disclose, perhaps the most sensitive number required under FAS 132. According to a study of the S&P 500, the aggregate planned 2004 contribution was pegged at $20 billion, down from the aggregate $71 billion contributed in 2003. “If corporate bond rates don’t go up significantly, that $20 billion could go up a lot — and that’s a cash-flow number,” says Silverblatt. He expects to track those estimates, since FAS 132 also requires firms to update the estimates quarterly if there is a “significant” change.
Windows to Their World
Some companies are disclosing even more than is required by the regulation. Discount rates have come down from an average of 6.7 percent in 2002 to 6.1 percent in 2003, according to Ciesielski. In addition, sensitivity analyses, or the impact on earnings if pension assumptions are changed, now appear more frequently in MD&As. DuPont, for example, offers the impacts on pension expense for potential changes in both the discount rate and assumed rate of return, while EDS discloses the assumed maturity of its liabilities upon which it bases its discount rate (19 years in 2003).
The Goodyear Tire & Rubber Co. even revised its methodology for choosing a discount rate when it restated earnings in May. The company switched from using a six-month average of corporate-bond rates — which gave it a higher-than-average discount rate of 8 percent in 2001 — to the more standard AA rate, bringing its 2001 rate down to 7.5 percent and lowering 2002 and 2003 rates by a half of a percentage point each. The result? A $161 million increase in pension liabilities and a $100.1 million decrease to earnings in the period, thanks to additional noncash charges and higher taxes.
If FAS 132 is behind these moves, companies aren’t giving it credit. In fact, few firms say they feel pressured to match the norms that emerge from the new data. Even EDS, which has one of the highest equity holdings at 80 percent, stands by its approach. “This will not change our investment strategy,” says spokesperson Sean Healy, noting the firm has explained its reasoning (a young workforce) in its 10-K.
Others, such as Ryder System Inc. treasurer Dan Susik, add that FAS 132 means “just disclosing what we were already doing internally.” The duty to update its projected contribution in quarterly statements, he says, caused little consternation when the firm decided to boost the figure from the $41 million estimated in the 10-K to $65 million, for example. “You have to be able to justify it to the board, so making the disclosure after that wasn’t difficult,” says Susik.
Appealing for Caution
In part, the new openness is coming from the Securities and Exchange Commission. In fact, Milliman’s Beilke says that many of the numbers he’d like to see from FAS 132 are appearing in MD&As “because the SEC has pressured companies to give more information about assumptions that could affect income.” The commission called for many such changes in its review of the Fortune 500 during 2002.
SEC deputy chief accountant Scott A. Taub even took a shot at pension discount rates in a recent speech. With average rate assumptions running one-quarter to one-half of a percentage point above the AA bond index rate, Taub said he was unimpressed with how firms justified them. “Unfortunately, we’re often given responses that merely refer to rates used by other companies,” he said, “or an auditor’s approval.” Instead, he suggested firms construct a mock bond portfolio in which maturities mirror liabilities, a step few have taken, says Hewitt’s Johnston.
How accurate any pension projections will be, of course, is unclear. “You’re never going to get to a perfect number, since you can never exactly match your liabilities to the published bond rates,” says Beilke. Guesses about stock-market performance are equally suspect. And disclosures about expected cash contributions may be hampered by congressional wrangling over the Pension Stability Act. While Congress recently allowed companies to switch from the defunct 30-year Treasury bill rate to the higher corporate bond rate for calculating a pension plan’s funded status, the move is only temporary. And the uncertainty over what to use long term, says Aliya Wong, director of pension policy for the U.S. Chamber of Commerce, “really messes up financial planning.”
Given the inherent instability of most assumptions and projections, many are hoping that the Financial Accounting Standards Board will give up trying to make pension-fund performance hit the bottom line harder than it already does. The silver lining of FAS 132, say some, may be that it forestalls more onerous changes to FAS 87 that would eliminate smoothing (see “Good-bye Smoothing?”). “If the argument for accounting changes is that analysts don’t have information, this should take care of it,” says Judy Schub, managing director for the Committee on Investment of Employee Benefit Assets, an arm of the Association for Financial Professionals.
Alix Nyberg is a freelance writer based in Boston.
In April, the Financial Accounting Standards Board informally decided to follow the International Accounting Standards Board’s lead on pension accounting issues — a decision that has many U.S. companies concerned. And while no date or agenda has yet been set, observers believe that such a move will inevitably end “smoothing,” the practice of amortizing pension gains and losses over time.
Convergence along these lines is “almost inevitable,” says John Ehrhardt, principal and consulting actuary for Milliman Inc. The current exposure draft of the IASB standard for pension accounting — IAS 19 — allows for either smoothing or marking to market, as British companies are now required to do. However, since the policy shift in the United Kingdom occurred when current IASB head Sir David Tweedie was in charge of its accounting standards board, Ehrhardt expects the international version to rule out smoothing as well.
A move to the British method would force companies to use the market value of their assets in calculating the noncash pension expense, rather than look at any smoothing method. Without smoothing, companies would have to expense or record any discrepancies between the plan’s long-term liabilities and the fair value of plan assets as measured at the end of a given reporting period. This could create wild swings in market value and revenue.
The full impact on U.S. income statements is hard to assess, since FASB envisions the measure proceeding hand in glove with a larger project to reshape income statements with new categories. Nonetheless, the prospect of losing smoothing is mobilizing certain industry groups. The Committee on Investment of Employee Benefit Assets, for example, commissioned two studies that predict pension-plan sponsors would move assets from equities to the less-volatile fixed-income securities, which could suck anywhere from $250 billion to $650 billion out of the stock market. —A.N.
What They’ll Contribute
Highest contributions based on a survey of the 2003 10-Ks of the 100 largest pension plans.
|Company||Expected ’04 contribution||Funded status*|
|Ford Motor||$1.1 billion||82.1%|
|SBC Communications||$1 billion||101.9%|
|Delphi Technologies||$650 million||65.2%|
|Source: Milliman Inc.
*Market value of assets/projected benefit obligation