Guest Article

(From the February 4, 2008 issue of Deloitte's Washington Bulletin, a periodic update of legal and regulatory developments relating to Employee Benefits.)

SFAS No. 158: Effect of Pension Accounting Changes on Profitability, Leverage and Volatility

SFAS 158, Employers' Accounting for Defined Benefit Pension and Other Postretirement Plans, significantly changes the way pension plan liabilities are recognized in the plan sponsor's financial statements. The Georgia Tech Financial Analysis Lab analyzed the Forms 10-K filed by the companies comprising the Dow Jones Industrial Average to determine the impact of SFAS 158. Generally, the changes resulted in a decrease in total assets, a decrease in total shareholder equity, and an increase in total liabilities. For most, the recognition of larger pension liability artificially improved ROA and ROE, and increased leverage as measured by total liabilities to total shareholders' equity. Georgia Tech Financial Analysis Lab, College of Management, Georgia Institute of Technology, "The Effects of Enacted and Proposed Pension Accounting Changes on Leverage, Profitability and Earnings Volatility" (January 2008).

The SFAS 158 Study

Georgia Tech examined the 10-Ks filed by the companies comprising the Dow Jones Industrial Average. /1/ Using the balance sheet, statement of shareholders' equity and the benefit plan notes, it was able to back into the pre-SFAS 158 values for total assets, total liabilities and total shareholders' equity to perform the analysis.

Impact of SFAS 158

In terms of the balance sheet, Georgia Tech determined that SFAS 158 had the following impact.

In terms of leverage and profitability measures, SFAS 158 had the following impact.

Future Changes in Pension Accounting

While SFAS 158 corrects the balance sheet inconsistencies previously allowed by the delayed recognition of pension plan and OPEB components under SFAS 87, the full changes in fair value of pension assets and liabilities do not yet flow through income. Fair value accounting would call for such changes to flow though income in the year realized. For this reason, SFAS 158 is possibly only the start of pension accounting reform.

Georgia Tech examined the same companies to determine the effect on pension expense and on income from continuing operations if the full pension costs were recognized in the year realized. Under the analysis, net periodic benefit costs for pensions and OPEB reported by the company was replaced with a revised measure. Net periodic benefit cost includes service cost, interest costs, expected return on plan assets, and the amortization of certain costs for the year. The revised cost includes service cost and interest cost, but includes actual return on plan assets and the full actuarial gain or loss incurred during the year. The revised cost approximates what the cost would be if actual returns were used to calculate pension expense, and if all expenses were recognized in net income in the year incurred.

The five-year period used in the analysis, 2002 to 2006, included years of both good and bad economic performance. The analysis examined the percent change in pension expense between the reported and revised net periodic benefit cost, as well as the percent change in income from continuing operations assuming the revised expense replaced the reported amount. A combined federal and state marginal income tax rate of 38 percent was assumed in calculating revised income from continuing operations.

Effect on Pension Expense

The analysis showed the following median changes in pension expense over the 5-year period, where the revised pension expense is based on the full pension costs in any given year and on the actual return rather than the expected return on assets.

The largest changes are seen in 2006, showing a 109.0 percent decrease in pension expense during a year the markets were strong -- and in 2002, showing an increase in pension expense of 887.4 percent during a year the markets were weak and in which all 24 companies experienced actual losses on plan assets when their expected returns were positive.

Effect on Income from Continuing Operations

The analysis showed the following median changes in income from continuing operations over the 5-year period as a result of the pension expense adjustment.

The median company would have reported 51 percent lower income from continuing operations in 2002 using the adjusted pension expense.

Volatility

A common argument against use of full pension costs is that it increases the volatility of income from continuing operations because the costs include unrealized actuarial gains and losses. The variability of pension expenses was examined by looking at the coefficient of variation before and after the adjustment. The median coefficient before the adjustment was 0.38, and after the adjustment was 1.22 -- an increase of 221 percent.

Similarly, with regard to the variability of income from operating expenses, the median coefficient before the adjustment was 0.26, and after the adjustment was 0.51 -- an increase of 96 percent. This indicates that changes in accounting for pension expense that include full costs would significantly increase pension expense volatility. Such changes would also increase the volatility of income from continuing operations, but to a lesser degree.

/1/ Of the 30 companies in the DJIA, six were eliminated for data reasons. Two provided only defined contribution plans, another two do not have defined benefit plans for their US operations and provided only limited disclosure for their international plans, and the final two have fiscal years that ended prior to the required adoption of SFAS 158. Therefore, the analysis was completed on the remaining 24 companies in the DJIA.

The information in this Washington Bulletin is general in nature only and not intended to provide advice or guidance for specific situations. If you have any questions or need additional information about articles appearing in this or previous versions of Washington Bulletin, please contact: Robert Davis 202.879.3094, Elizabeth Drigotas 202.879.4985, Mary Jones 202.378.5067, Stephen LaGarde 202.879-5608, Erinn Madden 202.572.7677, Bart Massey 202.220.2104, Mark Neilio 202.378.5046, Martha Priddy Patterson 202.879.5634, Tom Pevarnik 202.879.5314, Sandra Rolitsky 202.220.2025, Tom Veal 312.946.2595, Deborah Walker 202.879.4955. Copyright 2008, Deloitte.