Discussion: (0 comments)
There are no comments available.
View related content: Economics
In cities and states around the country, elected officials, public employees and taxpayers are concerned about the funding of public sector pensions, which are placing increasing pressure on government budgets. But in reality, none of these stakeholders have a clear idea how well funded their plans are, due to accounting rules and disclosures that confuse true funding—that is, money you’ve actually got—with pseudo-funding based upon money the plan expects to earn in the future.
Most public employees participate in traditional defined benefit plans, which promise participants a fixed monthly benefit in retirement. The plans hold a variety of investments, including stocks, bonds and “alternatives,” such as hedge funds and private equity. But unlike a Section 401(k)-type plan, the risk of these investments is borne by the plan sponsor—meaning, the government and ultimately the taxpayer. In most places, public pension benefits are protected by law and can’t be cut even if the plan’s funding runs short. It’s understandable that government officials and taxpayers would be interested in knowing how well funded their plans may be.
`True’ versus `future’ funding
But when it comes to pension funding, there’s funding and then there’s “funding.” For instance, take a plan that claims to be 80 percent funded. Part of that 80 percent funding is based on the value of the assets held by the plan today. Those assets could be sold off and used to pay participants the benefits they’ve been guaranteed, with certainty. Let’s call this amount “true funding.”
But a second part of the plan’s current funding health is what we might call “future funding”—that is, the assumption that the plan’s investments will earn some given rate of return (usually around 8 percent) in the future. Plans work future funding into the equation by “discounting” their future liabilities at the 8 percent return they assume they’ll earn in future years. For instance, if a plan owed a $1 million lump-sum payment in 15 years, it would discount this figure at 8 percent to generate a “present value” of around $315,000. If the plan held assets today of $315,000, it would call itself “fully funded.”
Now, these are two different things: part of a plan’s funding is based on things that have already happened—the investments the plan has built up, which could be liquidated today if needed. The other part is based on things that may or may not happen in the future, meaning the expectation of earning an 8 percent return on risky assets. But the person who reads that their state’s pension plan is, say, 80 percent funded wouldn’t be able to tell the difference.
Many plan stakeholders assume away this problem by presuming that future investment returns are essentially guaranteed. That’s what pension accounting disclosures do, and if you talk to plan administrators, they act as if achieving their projected returns is all but certain. After all, if plans generated 8 percent returns over the past 30 years, why won’t they do so over the next 30?
But, as they say, past performance does not guarantee future results.
For instance, one possibility is that pensions got a good luck of the draw over the past 30 years and might not do so well in the future. “Monte Carlo” analysis, which uses computers to simulate the yearly ups and downs of investment returns, shows that most plans have a less than 50 percent probability of achieving their projected returns. There’s a significant chance they could do much worse.
Similarly, we hear that stocks never fail to outperform bonds over long periods. That was true—until it wasn’t. More recently, bonds have beaten stocks even over 30 years. Alternately, one might point out that 30 years ago, the yield on safe U.S. Treasury securities was high, so a pension needn’t take much investment risk to generate an 8 percent return. Today, returns on safe investments are far lower, so a pension investment strategy aiming for an 8 percent return will be a lot more volatile than in the past. That increases the chance of a budget-busting investment loss.
Finally, many professional investment consultants who work for public plans project that future investment returns won’t come near 8 percent. For instance, in a recent survey, eight investment consults projected that a typical public plan portfolio would return only around 6 percent over the next 15 years. Not a single consultant projected average annual returns in excess of 6.9 percent.1
Stocks must be treated as risky investments
The point isn’t to be gloomy. Stocks have been a good investment in the past and (we hope) they’ll continue to be in the future. It’s simply to say that these are risky investment returns and they should be treated as such.
There’s a lot riding on this. The problem is, most plan stakeholders have no idea how much.
But there’s an easy way to find out: Calculate the plan’s funding status on the assumption that it invested only in safe assets designed to match the low risk of the plan’s benefit liabilities. Technically speaking, the plan should discount its liabilities using an interest rate on guaranteed investments, since the plan’s own benefits are guaranteed. This is how most academic economists believe that pension accounting should work, and this “fair market valuation” has been endorsed by federal government agencies including the Congressional Budget Office, the Federal Reserve and the Bureau of Economic Analysis.2
For instance, take a plan that is 80 percent funded on the assumption of earning 8 percent returns on risky assets. If the return on long-term Treasuries is around 3.5 percent, then the plan’s true funding status is around 40 percent. That is, the plan has enough money to guarantee payment of around 40 percent of the benefits that it has itself guaranteed to pay. The other 40 percentage points of the plan’s “funding” comes from the assumption that the plan’s risky investments will pay off. Maybe they will, maybe they won’t—but the benefits must be paid regardless.
Fair market valuation
The argument isn’t that plans should invest only in riskless assets, or that the investments they do hold will necessarily generate the same returns as safe investments such as U.S. Treasury securities. It’s only to acknowledge that there is a chance that risky investments will underperform guaranteed investments, even over the long term. Market prices for riskless investments incorporate these factors.
Fair market valuation also tells us something about investment policy. Under current accounting rules, a plan that takes more risk with its investments instantly becomes “better funded,” because it can discount its liabilities using a higher discount rate. The market approach makes clear what’s going on: The value of the plan’s current assets hasn’t changed, nor have the future benefits the plan is obligated to pay. The only thing that’s changed is the risk profile of the plan’s investments. A riskier investment portfolio doesn’t make the plan any better or worse funded—after all, a dollar of stocks isn’t worth more than a dollar of bonds—it’s just a different funding strategy.
Put another way, current accounting rules act as if a plan’s funding health improves when a shift into risky assets is made. In reality, funding health improves only once those risky assets have paid off. But as long as pension accounting rules literally hold that “more risk = better funding,” we’re likely to see dysfunction in how public plans are managed.
If fair market valuation were implemented, public plans could and very likely would continue to invest in stocks. After all, private sector pensions, along with public employee plans in other countries, must use accounting rules similar to fair market valuation, yet these plans continue to hold stocks. But these other plans don’t hold as much stocks as U.S. public plans, who have significantly riskier investment portfolios than private pensions or public plans abroad.3 Fair market valuation wouldn’t mean the end of U.S. public plans. But, because a plan couldn’t call itself “fully funded” merely by taking more investment risk, it would mean less risk-taking and higher contributions up front.
One seeming reasonable objection to the fair market approach is that, by encouraging higher pension contributions, it might force current generations to “overpay.” That is, if plan sponsors increased pension contributions but maintained the same investment portfolio, there is a good possibility that plans will end up generating surpluses. Future generations could use these surpluses to reduce their own contributions. This violates the rule of “intergenerational equity,” under which each generation should fully fund the benefits owed to public employees in that generation, thereby ensuring that no generation overpays or underpays for the services they receive.
Let’s say that a plan follows current practice and “fully funds” its future benefits under the assumption of earning an 8 percent return. This, at best, implies a 50 percent chance that today’s contributions will be sufficient to pay the benefits we’ve promised to public employees today. In reality, due to the way stock returns are distributed, there’s actually only a 40 percent to 45 percent chance of a “fully funded” plan being able to meet its obligations.4 Even more importantly, those future shortfalls are likely to happen at precisely the wrong times. Because stock prices are correlated with the rest of the economy, the plan will fall short—and future taxpayers will be asked for greater contributions—when the economy is weak, unemployment high and tax revenues low. Surpluses, by contrast, will happen in good times, when money is easy to come by. It’s wrong to treat deficits and surpluses as offsetting, since people don’t value them the same.
Quantifying with options
It’s possible to quantify these effects using prices paid for financial instruments known as options. A put option gives the holder the right to sell some underlying asset for a specified minimum price at a specified future time. Thus, a pension plan could purchase put options as an insurance policy against their investments falling short, thereby protecting future generations against overpaying. Likewise, the plan could sell a call option, which gives the holder the right to purchase a given asset for a specified price. By selling off any future surpluses in this way, the plan could lower costs to current taxpayers and assure that they’re not being overcharged.
The key insight is that the cost of a put option protecting against deficits is far greater than can be gained by selling off surpluses using a call option. In fact, if you add up all the costs—the initial pension contribution as calculated under Governmental Accounting Standards Board rules, plus the cost of a put option protecting against deficits, minus the cost of a call option selling off surpluses—the total cost of truly fully funding a pension’s liabilities is exactly the same as if you simply valued those liabilities using a riskless interest rate. This result, which is due to a financial principal called “put-call parity,” shows that fair market valuation isn’t an academic abstraction. It’s not based on the assumption that pensions can only invest in bonds or that their investment returns will be the same as bonds. It’s based on the reality that pension investments are risky and that pension stakeholders—meaning, the taxpayers who must fund pension plans, or the citizens who lose other government programs as they are squeezed out by rising pension costs—value different risks in different ways.
Ideally, “full funding” would mean that a plan that guarantees a given benefit can actually guarantee to pay that benefit. But we can’t know what full funding means as long as we confuse funding based on the assets a pension plan actually holds with pseudo-funding based on the expectation of earning high, but risky investment returns. The fair market valuation approach provides that information.
1 See James J. Rizzo and Piotr Krekora (of Gabriel Roeder Smith & Company). “The Goldilocks Principle and Investment Return Assumptions” (presentation, Florida Government Finance Officers Association 2013 Annual Conference, Boca Raton, Fla., June 25, 2013). p. 40, http://www.fgfoa.org/Assets/Files/Jim_Rizzo_Presentation_PDF.pdf.
2 See Congressional Budget Office. “The Underfunding of State and Local Pension Plans.” May, 2011.; Kohn, Donald L. “Statement at the National Conference on Public Employee Retirement Systems Annual Conference.” New Orleans, La., May 20, 2008. Wilcox, David. Testimony before the Public Interest Committee Forum sponsored by the American Academy of Actuaries, Sept. 4, 2008. Reinsdorf , Marshall B. and David G. Lenze. “Defined Benefit Pensions and Household Income and Wealth.” Bureau of Economic Analysis. Research Spotlight. August 2009; Lenze., David G. “Accrual Measures of Pension-Related Compensation and Wealth of State and Local Government Workers.” Bureau of Economic Analysis, April 2009.
3 Andonov, Aleksandar, Bauer, Rob and Cremers, Martijn, Pension Fund Asset Allocation and Liability Discount Rates: Camouflage and Reckless Risk Taking by U.S. Public Plans? (May 1, 2013). Available at SSRN: http://dx.doi.org/10.2139/ssrn.2070054.
4 A pension portfolio has a less than 50 percent chance of meeting its assumed return because stock returns are skewed: the average, or arithmetic mean, is made up of a small number of very high returns and a larger number of smaller returns. Thus, the median or typical return will be lower than the average return. Pensions could easily address this problem by basing discount rates on their investment portfolio’s geometric mean, or compound, return. This would produce the outcome that the plan has a 50 percent chance of meeting its assumed return. However, the geometric mean is slightly lower than the arithmetic mean, which would imply lower funding ratios and higher contributions.
There are no comments available.
1150 17th Street, N.W. Washington, D.C. 20036
© 2015 American Enterprise Institute for Public Policy Research