New Pension Accounting: Volatility City?

If FASB makes companies run plan assets through their income statements, they can expect herky-jerky profits, a new study suggests.

Marie Leone, CFO.com | US
January 21, 2008

Nothing is official yet. But the next phase of pension-accounting rulemaking is likely to affect the income statement. That means that the Financial Accounting Standards Board is considering changing the way companies report profits. If that happens, it's sure to cause a stir.

Under the current rule, FAS 158, Employers' Accounting for Defined Benefit Pension and Other Postretirement Plans, companies measure changes in the value of pension assets and liabilities caused by stock market upswings and downturns. They record those changes as "other comprehensive income" in the shareholder's equity section of their financials.

After that, the changes in value are allotted into the income statement over several years, using the so-called "corridor" approach. The method is a way of amortizing the changes to smooth the effects on net income.

But FASB members, staffers, and other accounting experts are discussing the idea of requiring companies to record changes in value on the income statement immediately. Such a move would be in keeping with FASB's recent role as a fair-value advocate. "I think [FASB] will do it because it moves pension accounting to full fair value," says Charles Mulford, the director of the Georgia Institute of Technology's Financial Analysis Lab.

Further, according to new research Mulford released this month, the effects of the accounting change could be big enough to dwarf the impact of the company's core business and "render the income statement useless," Mulford predicted.

For example, consider a year in which stock market returns are strong. A company records a $1 billion increase in the value of its pension assets above what it had estimated. If the billion-dollar change flows through the income statement, the company looks more profitable than it would have if the stock market had been down. The billion-dollar gain would offset that year's pension expense.

Conversely, a $1 billion drop in pension-asset value flowing through the company's income statement making it look, at first blush, that profits are down. "FASB's argument [for considering the revision], which I don't disagree with, is that the change reflects what is happening. And that ought to show up in the financial statements," Mulford told CFO.com.

The first phase of FAS 158 made that happen on the balance sheet. FASB issued the rule in 2006, forcing companies to recognize overfunded or underfunded pension plans, and record any changes in value on their balance sheets at fair value. The rule revision caused some companies to rework debt covenants based on debt-to-asset and other affected ratios. Others had to revise internal strategies. And while many companies didn't like taking the short-term hit, it didn't cause earth-shattering changes to results unless a company was already on the brink of insolvency and its covenants were also teetering.

In fact, Moody's Investors Service, the credit-rating agency, was underwhelmed by the effect of the pension rule. Moody's noted that even though FAS 158 would boost total liabilities "significantly," the change would shake neither company fundamentals nor credit ratings.

While the same will probably be true for the second phase of pension accounting, adjustments to net income are simply harder to swallow than balance-sheet changes. For its part, FASB hasn't released any official discussion papers about the second phase, agreeing to work in tandem with the International Accounting Standards Board to come up with a proposal. For now, FASB is leaving the door open to criticism of the new idea.


To that end, Mulford, along with graduate student associates Erin Quinn and Ryan Swanson, produced a study issued this month that seeks illustrate the impact of running changes in pension-plan value through the income statement. In a nutshell, the accounting change would spur a big rise in volatility stemming from the percentage changes in pension expense and in income from continuing operations. (The latter refers to net income before such one-time events as extraordinary items and discontinued operations.)

The study looked at 24 of the 30 companies represented in the Dow Jones Industrial Average. Six companies were eliminated from the report because they either didn't have pension plans or pension data wasn't clear or available. The researchers applied the full fair value treatment to changes in asset values for each company for a five-year stretch (2002 to 2006). The study compared actual returns on pension plan assets to expected returns, reflecting the market swings of bull and bear markets.

In 2002, for example, all 24 companies — including 3M, Boeing, Caterpillar, Citigroup, Dupont, JP Morgan Chase, and Merck — predicted positive returns. But weak markets led to actual losses on plan assets for all two-dozen companies.

Thus, the pension expense for median companies that year climbed by 887 percent. Run through the income statement, and evaluated by an untrained eye, that increase in pension expense could make the median company look quite a bit less profitable. Further, company profits would look extremely volatile compared to 2006, when the median company adjusted pension expense dropped by 109 percent.

To see the effects the change in accounting treatment might have on a single company, consider Dupont's pension expense over the five years (see chart). The chemical company's adjusted expense figure swung from a 2,077 percent deviation between expected and actual results in 2002 to a negative deviation of 374 percent in 2006.

The study's results for percentage change to income from continuing operations was similarly "all over the map," said Mulford. Using the revised accounting treatment, the median company would have reported a 51 percent drop in income from continuing operations in 2002, while income would have risen by 9.8 percent in 2006.

When it comes to the volatility of individual companies, Boeing stands out as being especially affected by fair-value pension accounting (see chart). The aircraft manufacturer would have recorded a 177 percent dip in income in 2002, but a 168 percent rise in 2006. Alcoa would have suffered from a big swing in volatility as well, going from a 306 percent drop in income in 2002 to a 9.8 percent rise in 2006. Caterpillar's fair-value adjustments would have been recognized as a 171 percent decrease in income in 2002 and an 11 percent jump in 2006.


Pension expense swings
Percentage change in adjusted pension expense if recorded at fair value. (%)
Company name* 2006 2005 2004 2003 2002
3M -150.5 32.0 -72.0 232.0 1182.3
Alcoa -64.5 4.8 53.2 61.2 1000.4
Altria Group -91.5 45.6 168.5 93.8 716
Boeing -145.1 -121.7 -35.0 195.9 5314.9
Caterpillar -93.6 199.6 0.9 271.6 1318.0
Citigroup -184.9 -81.9 247.3 98.2 1338.9
Dupont EI de Nemours -474.8 -56.2 68.4 -65.8 2077.0
Johnson & Johnson -18.4 27.8 85.9 138.7 993.9
JP Morgan Chase 64.1 71 136.6 22.2 1068.5
Merck & Co. -134.2 -12.4 -26.7 26.1 779.6
United Technologies -215 -2.5 33.5 107.7 2471.3
Verizon Communications -258.8 -49.7 -200.4 98.6 N/M**
*Sampling of Dow 30 companies. **N/M = Not meaningful. If the expense changed from a negative number (a benefit) to a positive number (a cost), the percent change is not calculated.
Source: "The Effects of Enacted and Proposed Pension Accounting Changes on Leverage, Profitability and Earning Volatility," 2008, Georgia Institute of Technology.

Swings in Income from Continuing Operations
Percentage change in adjusted ICO if recorded at fair value. (%)
Company name* 2006 2005 2004 2003 2002
3M 10.7 -2.8 6.5 -15.3 -77.6
Alcoa 9.8 -1.2 -11.3 -13.9 -306.4
Altria Group -7.4 -3.5 -10.9 -4.0 -24.0
Boeing 67.8 -54.5 14.1 -131.1 -277.1
Caterpillar 11.3 -30.9 -0.2 -60.5 -171.0
Citigroup 1.7 1.1 -3.5 -0.9 -10.0
Dupont EI de Nemours 30.8 11.0 -18.0 -34.3 -124.5
Johnson & Johnson 1.0 -1.3 -4.2 -6.4 -18.4
JP Morgan Chase -0.6 -0.8 -3.3 -0.5 -57.0
Merck & Co. 10.6 0.9 1.5 -1.2 -12.9
United Technologies 19.8 0.2 -2.8 -7.4 -74.0
Verizon Communications 56.3 9.1 34.2 -53.8 -183.0
*Sampling of Dow 30 companies.
Source: "The Effects of Enacted and Proposed Pension Accounting Changes on Leverage, Profitability and EArning Volatility," 2008, Georgia Institute of Technology.