GASB (Governmental Accounting Standards Board) passed some new standards relating to valuation of public pensions recently. From their own press release:
GASB Improves Pension Accounting and Financial Reporting Standards
The Governmental Accounting Standards Board (GASB) today voted to approve two new standards that will substantially improve the accounting and financial reporting of public employee pensions by state and local governments.
I'm going to look at some specific changes here.
The Statement calls for immediate recognition of more pension expense than is currently required. This includes immediate recognition of annual service cost and interest on the pension liability and immediate recognition of the effect on the net pension liability of changes in benefit terms. Other components of pension expense will be recognized over a closed period that is determined by the average remaining service period of the plan members (both current and former employees, including retirees). These other components include the effects on the net pension liability of (a) changes in economic and demographic assumptions used to project benefits and (b) differences between those assumptions and actual experience. Lastly, the effects on the net pension liability of differences between expected and actual investment returns will be recognized in pension expense over a closed five-year period.
- Projections of Benefit Payments. Projections of benefit payments to employees will be based on the then-existing benefit terms and incorporate projected salary changes and projected service credits (if they are factors in the pension formula), as well as projected automatic postemployment benefit changes (those written into the benefit terms), including automatic cost-of-living-adjustments (COLAs). For the first time, projections also will include ad hoc postemployment benefit changes (those not written into the benefit terms), including ad hoc COLAs, if they are considered to be substantively automatic.
- Discount Rate. The rate used to discount projected benefit payments to their present value will be based on a single rate that reflects (a) the long-term expected rate of return on plan investments as long as the plan net position is projected under specific conditions to be sufficient to pay pensions of current employees and retirees and the pension plan assets are expected to be invested using a strategy to achieve that return; and (b) a yield or index rate on tax-exempt 20-year, AA-or-higher rated municipal bonds to the extent that the conditions for use of the long-term expected rate of return are not met.
Let's get some reaction via the NYT:
“This is still a flawed accounting system,” said Joshua D. Rauh, an associate professor of finance at Northwestern University. He says states should measure their pension promises much as an insurance company sets the price of an annuity — using the market rates for so-called risk-free bonds that are considered very safe. Using that standard, many states appear to have promised far more than they can reliably pay.
Robert H. Attmore, chairman of the Governmental Accounting Standards Board, said the organization had considered using such an approach. “We decided that it was not appropriate in the government environment,” he said. “I understand why it’s favored in the private sector, but governments are different.”
Yes, they are different. They go bankrupt in different ways. And there's no PBGC for public plans (thank goodness).
All logical fixes that remove three of the most common tools of deception from the public-plan actuaries’ kit and which will raise the value of unfunded liabilities by about 20% but that’s a number that will appear in a footnote. If a $53.9 billion number didn’t scare away prospective New Jersey bond-buyers then will a $64 billion number? After all it’s not as if New Jersey will be forced to pay off that liability, as we see from the next reason.
3) These rules are for reporting not for funding. At the last Enrolled Actuaries meeting the public pension sessions were all on what actuaries would do now that GASB was abnegating responsibility for setting funding recommendations. There won’t be any more ARC required for funding as GASB has washed its hands of setting funding standards for recalcitrant governing bodies that were free to thumb their noses at those standards and who, from here on, won’t even need to raise that finger.
I am not terribly bothered by the lack of an ARC (actuarially required contribution). Because so many governmental bodies have been essentially ignoring the ARC. Those who were being responsible before can continue being responsible; state laws can be updated to reflect what they consider to be the ARC under new accounting standards.
But I agree with Bury that this makes little difference in the bond market except for the more unsophisticated buyers. Oh wait — a lot of muni buyers are rather unsophisticated. About 70% of munis are held by individuals, many of whom are the stereotypical little-old-ladies trying to eke out some retirement income. (To be sure, you also have people like Teresa Heinz Kerry holding these assets for the tax benefits). I actually do see the possibility of some shakeout in the muni market once these new valuations come online.
There is a math problem with that discount rate "compromise". It gives even more incentive to boost the assumed return on assets. The higher that assumed return, the less of the liability one has to discount at a rate that's probably about 400 basis points less.
Back to the NYT:
Last year, the Center for Retirement Research at Boston College tried to determinehow much bigger, using its public pension database. As of 2010, those plans reported having, in aggregate, 76 cents on hand for every dollar they must pay retirees in the future. The center’s director, Alicia H. Munnell, estimated that if the new accounting rules had been in effect, the funds would have reported just 57 cents on the dollar.
Ms. Munnell’s research suggested the biggest loser would be the Illinois teachers’ pension fund, which would have been able to claim just 18 cents on the dollar if the new rules had been in place in 2010. Experts say when a plan is that depleted, it is almost impossible to revive. Top officials of the Illinois teachers’ plan have acknowledged that its problems are severe.
Ya think? That sounds awfully close to that pension death spiral I posted about before. Yes, a 16% fundedness rate sounds a lot worse than 43% funded (or whatever number they had then). But it wasn't just the discount rate that got massaged to get that optimistic fundedness ratio.
Here's one more trick that's getting damped down (again from the NYT):
The board will also do away with the commonplace practice of “smoothing” the value of pension investments, or spreading the recognition of market gains and losses over several years. The standard actuarial practice has been used in many places as a tool for engineering pension numbers.
Illinois, for example, adopted a new smoothing technique in 2009, when its funded ratio had plunged to a frightening 33 percent; the change pumped the ratio up to 43 percent, at least on paper.
In any case, I'm not particularly sanguine about the new standards bringing any better behavior than currently exists, and if the sight of the panhandling Providence mayor didn't wake one up to the magnitude of the public pension problem, I don't know that these new numbers would.
As many have to write about numbers, I know most people's eyes glaze over. Most politicos use numbers only as weapons, and I doubt many pay attention to that. But they do pay attention to stuff like cities declaring bankruptcy, and slashing pensions of current retirees.
So, good for GASB in updating their standards, but it's not going to stop public pensions failing across the country. Have fun counting those deck chairs.