State Pension System Prediction Markets
What, Me Worry?
On March 15, 2026 – three days ago, I wrote a short essay about prediction markets and concluded they are: “a marvelous tool for teaching students about the efficiency of the price mechanism and elegance of free markets.” That is because the serious students of prediction markets are more well informed about the uncertainties of market performance than government employee pension fund managers.
On the following day, on March 16th, reason.com published their 2025 Pension Solvency and Performance Report and revealed:
Public pension systems in the U.S. saw a decrease in unfunded liabilities since the previous report, dropping from $1.62 trillion to $1.48 trillion, a 9% decrease. This was largely driven by fiscal year 2024’s higher-than-expected investment returns.
Let’s do the math, er, arithmetic. The unfunded liabilities of state and local government pension funds decreased only 9% in year which earned almost 18% US stock market (iShares S&P 500 ETF) and 8% intermediate treasury bonds (iShares 7-10 year ETF) market returns. Both of those are sharply higher than long-term averages and a 50/50 portfolio earned about 12.5%
Bear in mind, the pension fund trustees have access to the most highly credentialed finance experts in America, and they are doing worse than average bettors in the prediction markets. But it’s worse than that. They started 2025 with $1.62 trillion in underfunded liabilities - after a year (2024) that delivered equity returns of 25%. According to the Reason Foundation report, that’s trillion with a “T,” but is Reason’s reasoning reasonable? No.
Below is a summary of the five categories used by the Reason Foundation to deliver “a comprehensive overview of the current and future status of state and local public pension funds.” The relevant paragraphs are excerpted below, and my summary of the evaluation premises will follow.
Funded Status
To calculate the funded status, you simply divide the market value of a plan’s assets by its actuarially accrued liabilities. In other words, this metric measures the extent to which a pension plan has accumulated savings relative to the estimated total liability of retirement benefits it has promised to its members.
For full disclosure, I am not an actuary or statistician, and I have great respect for the science and discipline. My focus will be on the capital market assumptions (CMAs) that are chosen because they may have a significant impact on the calculations. In turn, those will have a significant impact on taxpayers and plan beneficiaries.
With regard to the denominator of this ratio – which is a big deal, it was challenging to find a decent definition for Actuarially Accrued Liabilities, but according to an American Academy of Actuaries report dated June 2004, I discovered:
Actuaries must consider the difference between the actuarial liability, which is the value of benefits already earned, and the assets.
Economic assumptions dealing with current interest rates, salary increases, inflation and investment markets. How will market forces affect the cost of the plan?
Interest Rate – For pension funding, this assumption is used to discount future benefits to determine plan liabilities and it should be a reasonable expectation of the future rate of return on the pension plan’s assets.
Expected Long-term Rate of Return on Assets – It is used to determine the expected return on assets during the year. This assumption reflects the average rate of earnings expected on current and future investments to pay benefits. It is a long-term assumption that is reviewed regularly but generally changes when the long-term view of the market changes or with shifts in the plan’s investment policy.
Demographic assumptions about the participant group make-up and expected behavior and life expectancy. How will participant behavior affect the cost of the plan?
But what if the discount rate used to calculate the present value of future benefits does not conform to the historical evidence? What if it has been fudged to make the plan seem less underfunded? What if political pressure is applied in favor of an arbitrary prediction?
What if the predicted return on risk assets like stocks is a fixed rate for this calculation? And what is the expected return on the plan’s “alternative” investments like hedge funds, private equity, commodities, or cryptocurrencies? None of this is the least bit reliable.
Investment Performance
What really matters is how it compares to the plan’s assumed rate of return (ARR), which is the long-term investment target that pension boards set based on their expectations for future market performance. This assumption is important because it is used in actuarial calculations to determine the present value of all future benefit payments . . .
The good news is that pension trustees are matching their plan liabilities with assets and future contributions. That is fundamental to The Moneyball Method. The bad news is their “reasonable expectation of the future rate of return.” The only reasonable expectation is the historical range of returns.
What about the other two components that define each asset class – standard deviation and correlation coefficients? All three are needed to define the nature of each asset class category and to anticipate their behavior. But even worse is “changes when the long-term view of the market changes.” Whose view? On what basis?
That kind of hubris is a major contributing factor to today’s $1.48 trillion underfunded plan status of America’s state and local government pensions. Even that number is unreliable considering the information below.
Contribution Adequacy
Contribution rate adequacy measures whether the annual payments made into a pension system are sufficient to cover the cost of newly earned benefits while also paying down existing pension debt . . . “This assumption is important because it is used in actuarial calculations to determine the present value of all future benefit payments” and, consequently, how much money must be contributed each year.
This category introduces another important variable – pension plan debt not included in the unfunded liabilities. Also known as Pension Obligation Bonds, a reasonable person would ask why a pension plan owes anything to anyone other than the plan’s beneficiaries. And a reasonable person asked an even more reasonable question to the Managing Director of Government Finance at Reason Foundation:
Did you take into account Pension Obligation Bonds used to pay down the UALs, since those debts are directly related to the defined benefit pensions? . . . “We didn’t. POBs are generally recognized in the issuing state or local government’s financial disclosures (ACFR), while this report is based on the pension systems’ own financial disclosures.”
From this exchange, Mark Moses, author of The Municipal Financial Crisis, was able to shed light on the fact that pension systems are not required to report their Pension Obligation Bond liabilities - meaning that the debt obligations for malfeasance were not included in today’s $1.48 trillion deficit.
Asset Allocation Risk
These typically include traditional asset classes like publicly traded stocks (equities) and government or corporate bonds (fixed income), as well as a growing category of “alternative investments.” This alternative category includes assets such as private equity, hedge funds, real estate, and private credit. The fundamental goal of asset allocation is to strike a balance between risk and return that aligns with the fund’s long-term objectives.
Yesterday, on March 17th, Yahoo Finance jumped on this year’s bandwagon of warnings about private equity and private credit write downs:
Analysts point to aggressive lending and complex deal structures.. Late last year, the bankruptcies of First Brands Group and Tricolor Holdings, both tied to private credit, forced several banks to disclose large write-offs and raised fears that losses could ripple through financial markets.
To objective investors, there is nothing new or interesting about this. So-called alternative investments do not have long-term price histories from which to predict their behavior under stress, and those that are not liquid will never have reliable data for this kind of risk management. They are fine for savvy, high net worth investors, but for fiduciary accounts and the tax qualified retirement plans of middle-class beneficiaries, they may do more harm than good.
Probability of Hitting Assumed Return
This metric uses forward-looking capital market modeling to forecast the likelihood that a state’s pension portfolio will achieve its assumed rate of return (ARR) over a 20-year period. It provides a risk-based assessment of how realistic a plan’s investment assumptions are, given its specific asset allocation.
A higher rank indicates a higher probability of success . . . and placing a smaller potential burden on future taxpayers.
To be blunt, the probability of hitting an assumed return is identical to Lloyd Christmas dating Mary Swanson in Dumb and Dumber: “I’d say more like one in a million.” It never happens except for a fleeting moment as market values fluctuate – sometime wildly.
But let’s assume a “pension portfolio will achieve its assumed rate of return (ARR) over a 20-year period.” What was the volatility of returns during that 20-year span? Were there 12-month periods that were much higher and lower than the ARR? Was there one or two twelve-month periods that were so much higher that they compensated for eight or ten years of underperformance? That matters - a lot.
The volatility of the returns, especially when integrated with the cash flow needs of the plan, may have a significant impact on the pension’s funding status. And contributing to the problem is the concept of “probability of success.” Another dumb prediction. Is an overfunded plan a success? Is taking unnecessary risk a success? Are ever-increasing charges to taxpayers for plans with less than 100% funding a success?
No participant in prediction markets who would go near any of this - and they know the house has the advantage. In fact, the house has skin in the game, too. But in the government employee pension system, the house is the state government, pension plan trustees and fund managers. They have no skin in the game, but they have suckers known as taxpayers. Particularly in Kentucky, Illinois and New Jersey.