The Society of Actuaries recently published a study looking at the relationship between the funded status of defined benefit retirement systems and the discount rates they used. The report focused on the biggest public and private defined benefit plans in the U.S. between 2009 and 2014, and concluded there is no clear causal relationship between the unfunded liabilities of a plan and how the plan values its liabilities. However, public sector pension plans should not surmise from this finding that all is well with the discount rate practices they are using.
A primary finding from the Society of Actuaries report is that there is substantial variance in the discount rates used by defined benefit plans. The differences in discount rates were in part due to the difference kinds of plans analyzed. In 2014 about 60% of private sector multiemployer (or so-called “union plans”) had discount rates between 7.5% and 7.99%, where as 75% of public sector plans used discount rates in that range. By contrast, private sector single-employer plans (which are termed “non-union” plans in the report) almost never use a discount rate above 7.5%.
Public sector plans in particular were more likely to use high discount rates of 8% or greater — 11 of the 160 of the public plans analyzed by the report — and less likely to be using a 6% discount rate or lower. (Only three public plans were in this latter category: the Wisconsin Retirement System, Pennsylvania Municipal Retirement System, and Portland Fire and Police Disability Retirement Fund.)
The variance in discount rate approach reflects an ongoing regulatory dilemma for defined benefit plans in America. While all private sector defined benefit plans are subject to the Employee Retirement Income Security Act (known as ERISA), only private single employer plans are required to use market-valued discount rates. Public sector plans — governed by the Government Accounting Standards Board — and private sector multi-employer plans are allowed by their regulators to use the expected rate of return on assets as the discount rate.
Market-valued discount rates (using the risk of liabilities to measure the value of liabilities) tend to follow bond yields and are almost always going to be lower than a discount rate based on the asset allocation of a pension plan (using the risk of assets to measure the value of liabilities). Which of the two is appropriate for use by public sector plans in particular has been the subject of a recent actuary civil war in which members of the Society of Actuaries are falling on the market-valuation side.
Unfortunately, the ability for public plans and private multi-employer plans to use rate of return assumptions as the discount rate for valuing accrued liabilities tends to create a perverse incentive to apply rates that artificially lower the present value of promised pension benefits, thus understating the amount of recognized pension debt for a plan.
The conclusion of the study is that there is no statistical correlation linking plans with high discount rates to plans with low funded ratios, or vice versa. However, at the same time the Society of Actuaries study finds that the category of plans that currently use highest discount rates (public plans) have the lowest funded ratios. Unsurprisingly, the category that has the lowest discount rates (private single employer plans) enjoys the highest funded ratios, generally speaking.
A likely explanation for this phenomenon is that plans using high discount rates are probably also using other aggressive actuarial assumptions that have had a more proximate relationship to the growth of unfunded liabilities and decline of funded ratios. For example, with public plans we know that high discount rates generally mean high assumed rates of return, which in turn suggests an asset allocation that has trended away from safer, fixed income investments, and is thus more exposed to market volatility and investment return underperformance. Similarly, a plan that is willing to use a high discount rate may also be willing to use an aggressive payroll growth assumption or use an open amortization method.
Thus, while the Society of Actuaries report does not find a direct relationship between funding status and discount rate level, that doesn’t mean there isn’t an observed indirect relationship.
Moreover, as public plans continue to age and the baby boomer generation adds more and more people to retirement member rosters, the undervaluing of those liabilities may very well start to wreak havoc on pension plan solvency in a direct way. Just because there was no direct relationship from 2009 to 2014 does not mean there won’t be a direct, causal relationship between high discount rates and low funded ratios in the future.
To read the full paper of SOA, go here.