Sharf: Colorado’s public pension problems persist

The legislative Pension Review Subcommittee on which I serve as a citizen expert (appointed by House Republican leadership) has finished its three statutory meetings for the year, and the news for Colorado’s Public Employee Retirement Association (PERA) is not great.  Not only is the annual actuarial Signal Light report less optimistic than last year, an independent audit of that report using a more rigorous and realistic methodology gives even more cause for concern.

Every few years, the state conducts an audit of the actuary’s report.  This usually involves making sure that the actuary has faithfully implemented its methodology, accurately applying its assumptions and correctly calculating the results.  It almost always results in a clean bill of health because it never questions the validity of the actuary’s approach.

A second pair of eyes

This year, the subcommittee decided to take a different approach, engaging the Swiss auditing firm PNYX to examine PERA’s health and compare its results to those of the current actuary, Segal.  The results were enlightening.

PNYX’s methods differed from Segal’s in several ways, but the most important is that it dynamically models both the liabilities over time and the returns.  It does so by statistically varying the underlying economic assumptions, such as inflation and economic growth, rather than keeping them constant as the actuary does.

Finally, it models returns more accurately during those statistical simulations.  Instead of a simple bell curve favored by most analysts for its simplicity, PNYX assumes three things: 1) extreme events are more likely than in a bell curve; 2) extreme losses are more likely than extreme gains; 3) when extreme losses occur, all your assets decline together, meaning that portfolio diversification doesn’t save you.

Putting together those pieces, PNYX estimates PERA’s long-term returns at 6.71%, more than half a percent lower than its 7.25% long-term assumption.  This is an enormous difference, and results in a nearly 16% higher liability, and a nearly 6.5% higher unfunded liability all on its own.

PNYX estimates that there is roughly a 50-50 chance of the fund being fully-funded by 2048, which at first glance doesn’t seem so bad, even if the amortization period for the state employees fund and school employees funds is slightly longer than Segal calculates.

Even the more optimistic Signal Light report gives the state fund less than a 50% chance of reaching the goal, and the school fund only slightly better than a 33% chance.

But PNYX also focuses on the worst-case scenarios, because that’s where the risk to the fund, the state’s finances, and ultimately the taxpayers lies.  Yes, there may be a 10% chance that you end up 200% funded, but that’s not where the risk is, and in any event, political realities won’t allow that to happen.  Benefits would be increased and/or contributions reduced long before funded levels got to that point.

Unlike the actuaries, PNYX also statistically models economic growth and inflation over the next 25 years.  The baseline scenario gives PERA a 10% chance of being 36% funded or worse.  But in a high-inflation environment, that number drops to 26%. And the median funded ratio falls from nearly full-funded to 70% funded.  If economic policies succeed in suppressing growth, the results are even more dire:

Source: PNYX/Ortec Finance

Even if we don’t worry about the 10% worst-case outcomes, the likelihoods of unfavorable outcomes are sobering.

PNYX’s estimate, using a more sophisticated and more accurate model, gives PERA an 18% chance, better than 1-in-6, of dropping to below 50% funding – a point at which recovery becomes almost impossible.  Further, it gives the plan a 1-in-4 chance of end up at 60% funding in 2048, the same as tossing two heads in two coin tosses.  That is, there’s a 25% chance that all this spending, all these reductions, all these additional payments, will leave us roughly where we were before it all started.

This puts the focus on the worst-case scenarios in the Signal Light report.  The actuaries give the state fund a 33% chance of never reaching full funding, and the school fund a 44% chance of never getting to full funding.

PNYX puts the focus on the worst-case scenario, the likelihood of going bankrupt or treading water at low funded levels that preclude getting out of limbo.  But we should remember that these numbers are only valid as long as assumptions remains valid and investment strategies remain static.

Of the higher unfunded liability, PNYX estimates that $1.5 billion is a result of higher-than-expected salary increases for existing employees.  School boards and even state government departments increased employee salaries well in excess of the actuarial expectations in response to the wave of inflation we experienced over the last several years.

Benefits are based on the average of the highest five years of salary.  If an existing employee sees his salary boosted in excess of the actuarial expectation, then neither he nor the employer will have made the necessary contributions to cover the benefits that result from that increase.

Understating the problem

Unfortunately, the subcommittee’s response to inflation continues to be throwing money at the problem rather than to trying to understand it.  Last year, it voted down a proposal to study the likely long-term effects of inflation on the plan’s funded status.  This year, it voted down another proposal to require PERA to calculate the effects of new contracts that increase salaries beyond the actuarial expectations.

Instead, the subcommittee decided to recommend to the legislature that the annual $225 million contribution from the state be indexed to inflation, thus having taxpayers reward districts for excessively boosting salaries.  Worse, it betrays the original intent of Senate Bill 18-200, which was that inflation and natural growth would make that $225 million annual contribution shrink in importance as the plan’s funded level improved.

Sometime in the fall, probably in October, the subcommittee will meet again to draft its letter to the citizens of Colorado describing PERA’s condition.  While we have an obligation not to panic people, we also have an obligation to be straight with them.  In the past few years, we’ve stated that PERA is on track to on-time full funding assuming it hits its annual 7.25% expected rate of return.

This is, I believe, a material misrepresentation of the situation.  As discussed above, PERA will not hit an annual 7.25% target every year.  Its likely outcome is much better described by the statistical simulations that both Segal and PNYX run, and the subcommittee has a moral obligation to include those likelihoods in its letter.  If the subcommittee fails to do so because it is afraid of alarming people, it will instead misinform and ultimately surprise them when the bill comes due.

According to PNYX’s head, Dr. Elisabeth Bourqui, she has put a number of pensions back onto solid footing by taking small incremental steps to reduce what’s called the conditional value-at-risk (CvaR), which has the incidental and important effect of increasing the chances of reaching full funding.  Many of these steps can be taken without overhauling things like contribution rates.

One opportunity PERA has in that department is its investment strategy.  Certainly diversification across asset classes is important, but PERA’s investments in real estate have both lower return and a higher risk than the overall portfolio (see figure 1).  Similarly, Private equity’s expected higher return doesn’t seem worth the extra 5.5% volatility, especially given how opaque the valuations are and how difficult they can be to unload.  Getting out of those and putting them into stocks and bonds and even carefully-chosen alternatives would boost return while lowering risk.

As always, the proper answer to PERA’s problems is to close off the unstable and unworkable defined benefit plan to new entrants, putting them instead in the defined contribution plan.  New employees since 2018 have been covering the normal cost of their pensions with their contributions and the base employer contributions. The amortization equalization disbursement (AED) and the supplemental amortization equalization disbursement (SAED) are employer (taxpayer) contributions intended to help pay down the unfunded liability.

There should be no cost to putting them and new employees into the defined contribution plan.   And the employers could continue to pay the AED and SAED supplemental contributions until PERA is fully funded.

Joshua Sharf is a senior fellow in fiscal policy at the Independence Institute, a free-market think in Denver. He has been the state House Republican citizen expert appointee to the legislative Pension Review Subcommittee since its inception in 2019.