Most people familiar with fair market valuation already know that the amount of government pension debt in the U.S. is a lot higher than indicated by official numbers. This is primarily because public pension plans usually use methods that undervalue their liabilities. What is the “real” amount of public pension debt? According to an updated study from Stanford finance professor Joshua Rauh —$3.846 trillion for 2015. This figure is up from his original report last year titled “Hidden Debt, Hidden Deficits,” which found the public pension debt for 2014 was $ 3.4 trillion.
The $3.8 trillion in unfunded pension liabilities figure stands in contrast to the $1.4 trillion that states reported themselves under the new GASB standards. GASB 67/68 requires public plans to value assets on a market basis and use a “blended” discount rate to value liabilities. This blended discount rate only kicks in, though, when plans recognize certain future inability to pay promised pensions and most plans have found a way to continue using their assumed rate of return as the proxy for a discount rate.
According to Rauh’s paper, the liability weighted average discount rate for all 696 plans in his survey was 7.36% in 2015, which was only marginally lower than the 7.6% liability weighted expected return. In fact, only 9% of the plan sample used a discount rate lower than their expected rate of return. Per the GASB standards, as long as a plan determines that its projected assets (including future contributions and investment returns) are enough to cover its projected liability, that plan can use the expected rate of return as its discount rate even though its funded ratio is extremely low.
So how does the Rauh paper determine the appropriate discount rate and the market value of pension liability? Taking advantage of the new GASB rule requiring the disclosure of interest rate sensitivity for pension liability, the paper calculates the duration and the convexity of the liability of each pension plan. The resulting duration is then matched with the Treasury yield curve to determine the proper discount rate for each plan. Finally, the pension liability is revalued using the new discount rate and the respective duration and convexity.
Not unexpectedly, the difference in total unfunded liabilities is enormous. The market valued liability weighted discount rate was only 2.77% (compared to the official 7.36%), reflecting the highly certain nature of public pension benefits. The aggregate funded ratio under Rauh’s method for state and local pension systems was thus only 48.3%, far below the official 72.3%. This was also a step down from the previous year when the unfunded market value liability funded ratio was 51.4%.
Using the lower discount rates, the paper also reveals that the more realistic annual cost to keep pension debt from rising is 12.7% of all state and local own revenue (or 18.2% of tax revenue). Again, this is far higher than the actual pension contribution in 2015, which was only 4.9% of all state and local revenue.
While the paper’s conclusion is not new, its careful methodology paints a more accurate picture of public pension debt. Rather than relying on a fixed bond yield and a predetermined duration, the author revalues each plan’s liability based on its respective discount rate and duration, and takes into account its own convexity. This not only results in more realistic estimates but also provides a good model for other researchers to emulate.