WASHINGTON - Many companies were substantially overvalued even after the stock market's bubble had burst in early 2000 because investors were fooled by the contribution of pension funds to their bottom line.
Share prices of one-tenth of the companies in the Standard & Poor's 500 with defined benefit plans were overvalued by at least 20% at the end of 2001 - more than 18 months after the collapse of the stock market - because of the boost from pension earnings, according to a new study. About 130 companies in the index do not sponsor defined benefit pension plans.
The findings by Julia L. Coronado and Steven A. Sharpe, economists at the Federal Reserve Board, Washington, could spur calls for a complete overhaul of the accounting of pensions by requiring companies to simply report the true market value of pension assets and liabilities in their financial statements.
Ms. Coronado and Mr. Sharpe conducted the study by analyzing financial reports and stock prices of companies between 1993 and 2001. They attributed the overvaluation to the lag between the plunge in pension assets and the continuing high levels of pension earnings.
The lag, caused by the built-in smoothing mechanism in the current pension accounting rules, enabled companies to continue booking hefty gains from pension funds in their operating profits in 2000 and 2001, even though their pension assets had been decimated by the steep stock market decline.
But because of increased awareness of pension funds' contributions to corporate profitability, investors may not be fooled a second time into punishing companies on the downside as their pension expenses soar, Ms. Coronado said in an interview.
"We know that investors are a lot more focused on this issue," she said, suggesting that the misvaluation of pension earnings might not be true any more.
Under Financial Accounting Statement 87, the accounting rule governing pensions, corporate plan sponsors can report either the fair market value of their assets, or a smoothed value - amortizing any difference between the actual and the assumed long-term rate of return on assets- over a five-year period. As a result, there is a considerable lag in the value of pension assets that companies report in their financial statements.
By 2000, pension earnings accounted for 9% of projected earnings for the average corporation. But pension earnings averaged 22% in 2000, and 25% in 2001, for firms with the highest pension income, according to Ms. Coronado and Mr. Sharpe.
This and other flaws in the accounting rule for pensions may play a large role in the failure of investors to recognize the difference between a company's core operating earnings and earnings from pension funds, the study pointed out.
Thus, at the end of 2000, the ratio of net pension value (assets minus liabilities) to pension earnings for the S&P 500 companies was $187 billion divided by $20.6 billion, resulting in a multiple of 9.07. Instead, investors gavepension earnings a multiple of 20.7.
At the end of 2001, when the net pension value for these firms was a $2 billion loss, and pension earnings totaled $20.4 billion, investors still accorded a multiple of 18.3 to the pension earnings. Based on the authors' formula, the true multiple was -0.098.
"Our results suggest that investors do not distinguish between these two sources of earnings (core operating earnings and pension earnings), at least not in the way that one would expect in an efficient market. If anything, pension earnings appear to receive a higher valuation multiple than core earnings," Ms. Coronado and Mr. Sharpe noted in the study.
A statistical test performed as part of the study reached the stunning conclusion - investors gave pension earnings the same multiples as those of the underlying corporation.
High-growth companies, trading at a high price-earnings multiple, received a much higher multiple on their pension earnings than slow growth companies trading at low p/e ratios. "It would appear that investors implicitly extrapolate to the pension earnings of high p/e firms the prospect of more robust growth that these firms presumably enjoy in their core businesses," the authors noted.
The findings come as no surprise to those leading the charge for an overhaul of pension accounting rules.
"This paper proves that investors will lap up any numbers thrown out there," said Jeremy Gold, an independent actuary in New York. "It really adds fire to the fact that people are not thinking of pension plans separately, but just taking their earnings and multiplying them."
Kenneth Buffin, who heads an eponymous investment and actuarial consulting firm in Sparta, N.J., said: "Many of us were convinced that the market was indeed concentrating on reported earnings and consistently assigning unreasonably high multiples to the pension income element and failing to discriminate between core earnings and the anomalous accounting effect of FAS 87."
As a result, investors are just as likely to undervalue companies that report huge pension expenses in the down market years, said Lawrence Bader, a retired actuary from William M. Mercer Inc., New York, who has recommended changes in the current accounting rules in a paper with Mr. Gold.
"The obvious message is that investors should be looking at pension assets and liabilities, and not at FAS 87 pension expenses," he said.
Moreover, the Federal Reserve's study reinforces findings of ones in the 1980s by Alicia Munnell, MartinFeldstein, and other economists.
But the impact of pension earnings on corporate profits - and on stock prices - varied widely among the S&P 500 companies, so that the overvaluation in the aggregate was only 5%, or about one percentage point per year during the bull market between 1995 and 2000, the study found.
Consequently, "pension accounting made only a modest contribution to the aggregate stock market `bubble' during the 1990s," said the authors of the Federal Reserve study. This result is borne out by an earlier analysis of impact of pension funds on the S&P 500 companies by David Zion, an accounting analyst at Credit Suisse First Boston, New York.
"It's a big issue, but it's a big issue for select companies, so it doesn't surprise me that in the aggregate they don't see a very large impact," he said.