Another big problem, is unrealistic assumptions about returns on pension plan assets. This bias toward unrealistic assumptions occurs both in the private and public sectors. For purposes of illustration, below is the data for
IBM, which has an underfunded corporate pension plan:
The plan assumes an 8% annual return. To put this assumption into a larger context, the 30-year Treasury yield is currently around 3.1%. An assumption somewhat above the Treasury yield might not be too bad, but one that is more than 2.5 times that yield is inflated. It is also inconsistent with the conservatism principle that is at the heart of accounting, but the application of that principle falls short of the ideal.
The end result is that firms can readily maintain unrealistic pension plan assumptions in their financial statements (with the overly optimistic assumptions being confined to the notes to the financial statements--that's why due diligence is so important). In practical terms, the rosy assumptions rationalize a pension funding level that is consistently below what is actually required. Over time, a large unfunded pension liability accumulates.
IMO, the underfunded nature of public and private pensions argues strongly for pension reforms that would require full funding (capping the underfunded liability at a modest percentage of the overall pension obligation to allow for fluctuations in returns and costs) and place a ceiling on estimated returns (perhaps no more than 100 basis points above the average 30-year Treasury yield over a set period of time). More robust funding requirements and more realistic assumptions about returns would greatly reduce the risk of chronic underfunding and the accumulation of substantial unfunded liabilities. Moreover, if returns proved higher than expected, the difference between pension assets and pension obligations would allow for reduced funding for a period of time, hence funding levels would, over the longer-term, be appropriate with respect to the actual pension obligations.