The recent changes in pension accounting standards promulgated in 2006 relate to an old economists’ dilemma: whether salary trends can be predicted over a long term horizon. During 2006, the Financial Accounting Standards Board (FASB) issued Statement 158. For fiscal years ending after Dec.15, 2006, it requires all companies to reflect each defined benefit and other posretirement benefit plan’s funded status on the company balance sheet against the equity account of other comprehensive income.
FAS 158 does not change net income or comprehensive income for the previous years. Rather, it requires a one-time adjustment to accumulated other comprehensive income (AOCI) in shareholders’ equity.
FAS 158 reaffirmed the approach to overhauling pension accounting which would ask companies to include projected salary increases when reporting deficits or surpluses.
FAS 158 amends FAS 87. For nearly 20 years, FAS 87 required companies to show in the footnotes to their financial statements the amount owed to employees based on the projected obligation measure, thus not quantifying right away any future equity impact, and furthermore excluding salary projections.
So what does this change do to corporate balance sheets?
A number of actuarial consulting firms using existing disclosures from 2005 to 2007 estimated a combined pension deficit of about $90 billion. This projection was based on an analysis the expense structures and included projected salary increases for over one hundred large U.S.-based public companies. When they left the estimated salary increases out of the calculation, the deficit dropped to between $10-12 billion.
Does including the expected salary increases in the pension liability on the balance sheet make sense?
Some economists contend that salary increases are cyclical, strongly correlated to consumer price index moves (i.e. inflation or deflation), and often tied to labor productivity. In addition salary increases are inversely correlated to the marginal product of capital and equipment productivity.1
However, other economists contend that salary increases are industry specific and strongly related to bargaining strength, degree of industry protectionism, budgets and tax revenues for public employees, strength of labor unions for blue collar workers and other non-quantifiable market realities.
In response to FASB, numerous public companies have claimed that salary increases indeed aren’t guaranteed and therefore the estimates of a constantly growing salary liability should not be included in calculating the pension deficits.
While salary increases in the United States aren’t contractually guaranteed, they typically outpace inflation by a small margin over the long term, although the reverse is true sometimes over short time spans.
Salary increases for 2003 in the United States were approximately 3.4 percent, which was the lowest rate in more than 30 years. U.S. employers awarded average pay increases of 3.8 percent in 2007, just slightly more than they granted in 2006, 3.7 percent. The CPI was 2.7 percent in 2007, so the median wage hike exceeded inflation. Pay increases are projected to remain flat at 3.8 percent in 2008. (Source: Economics Research Institute)
This short term pattern proves that business cycles, which reverse over relatively short terms, indeed are related to salary increases.
Thirty years ago, the United States was going through a period of poignant inflation, driven in part by the gas crisis of the early and mid seventies. At that point in time, average raises were 7-10 percent, but inflation in the seventies was 7-8 percent, leaving an effective increase in buying power of 1-2 percent.2 Through the eighties, the economy improved, inflation dropped to 5-6 percent and salary increases started to trail the inflation rate. In fact, during the eighties, average salary increase rates dropped from 10 percent per year to just over 5 percent per year. In the first half of the 1990’s, with the very low U.S. inflation (3-4 percent), those same salary increase rates dropped to about 4-4.25 percent and they stayed there through 2001 and 2002.3
The positive evidence over the past 30 years suggests that salary increases are almost certain, thus supporting the claim that their inclusion in the projected obligation is a realistic measure of what a company would expect to pay in future.
Do we need to consider both short term and long term salary impacts?
Estimating cross-industry salary trends over longer periods tends to be complicated by the current credit crunch (i.e. persistence of illiquidity), the financial markets short-term downturn, and emerging industries which display highly volatile profit margins. All this makes estimating the projected pension liability more realistic if it is split into two components: A liability for employees closer to retirement which is LESS volatile because of LESS uncertainty from business cycles and a more UNstable one for employees further from retirement. The short tenor of the interest rate curve is similarly different than the long term displaying more volatility.
Thus due to the current condition of a credit crunch, the compression of expected returns makes the long term pension liability overstated due to short term conditions. Out of five actuarial valuation reports which I audited in 2008, all five assumed a rate on return on assets which was smaller than estimated salary increases, immediately creating a negative overall return since the rates are not dynamically adjusted yearly.
Are there industry specific issues to consider?
The emerging industries seem to need specific individual guidance: according to a salary study by online skills assessment specialist Brainbench, most IT workers received very little in the way of salary increases from 2002 to 2007, due to large price reductions and margin compression in IT equipment.
Brainbench’s salary study is based on data collected from more than 6,000 survey respondents, two-thirds of whom were male. The salary survey covers a variety of different trends, including year-over-year salary increase data from 2001 to 2007; male and female salary levels in key IT disciplines; salary increase rankings on a per-industry basis; the role of IT certifications in salary increases; and the typical size of compensation packages that augment salary increases.
Many IT firms didn’t offer additional compensation in 2007, but large organizations (those with more than $1 billion in revenue) and small organizations (less than $1 million) were more likely to offer such compensation. Workers in the computer industry consistently fared better than their colleagues in other industries in terms of salary increases when the IT equipment had better margins, before 2001. Before 2001, workers in health care, education or training, and financial services also fared better than normal.
How are companies’ credit ratings being impacted?
While moving benefit obligations on the balance sheet is likely to force many companies to recognize a big liability, there is uncertainty as to whether credit scores will be lowered for such companies, possibly jeopardizing their lending agreements, since the gap between salary increases and existing plan asset returns is primarily influenced by the competitive nature of the companies’ sales on a case by case bases, in which case the price increases may outpace inflation by a higher margin than the salaries.
In view of this assumption, I have selected a sample of 25 companies from a list of 35 companies with the largest reported FAS 158 projected pension deficits that were included by the actuarial consulting firms which estimated the $90 billion combined deficit from 2005 to 2007 in their study. I examined them for ratings changes by the main three ratings companies (i.e., Moody’s, Standard and Poor and Fitch). Of the sample, only three of them had been downgraded after the implementation of FAS 158, and for reasons other than the pension liability. Thus, the rating companies are prudent in including the future pension obligation as a right-away downgrade trigger.
1 See the neoclassical Solow-Swan steady-state growth model.
2 Source: Financial Trend Forecaster.
3 Source: Economics Research Institute.
Cicero Limberea is Director and Head of Accounting, Policy and Research at AIG Consumer Finance Group. He can be reached at Cicero.Limberea@aig.com
<< Previous Article |
| Print This Article