Did CalPERS Use Accounting “Gimmicks” to Enable Financially Unsustainable Pensions?
Gimmick – a concealed, usually devious aspect or feature of something, as a plan or deal.
In the past week, from Millbrae’s city hall to the inner sanctum of the CalPERS leviathan in Sacramento, defenders of pensions have been active. In particular, they have criticized the recent analysis, published by the California Policy Center, “How Much More Will Cities and Counties Pay CalPERS?” It would advance the ongoing debate over pensions to summarize the points of the CPC analysis, how CalPERS and their allies attacked those points, and how those attacks might be challenged.
On January 19th, in a report published online by Chief Investment Officer magazine entitled “CalPERS: Ring’s Flippant Claim of ‘Tricky Accounting Gimmicks’ Is False,” author Christine Giordano interviewed CalPERS spokesperson Amy Morgan. Tellingly, they did not discuss the substance of the CPC analysis, which specified, using CalPERS’ own data, how much more cities and counties are going to have to pay CalPERS. They focused instead on specific criticisms of CalPERS that followed those payment calculations.
As noted by the title of the report, CalPERS spokesperson Amy Morgan seemed to suggest the characterization of their accounting practices as employing “gimmicks” is not backed up by evidence. Morgan is invited to review the following evidence, after which she may join our readers in deciding whether or not “gimmicks” were employed.
GIMMICK #1 – THE CORRUPTION OF “ASSET SMOOTHING”
Asset smoothing is a practice whereby pension funds do not overestimate their assets after years of good returns, nor underestimate their assets after years of poor returns. It is a good way to avoid overreacting to market volatility. But in 2001, when the Dow Jones stock index had already been correcting for over a year and the Nasdaq was collapsing, CalPERS abdicated their responsibility to set the rules on smoothing.
When participating agencies in the CalPERS system were contemplating whether or not to follow the lead of the California Highway Patrol (SB 400, 1999) and retroactively increase pension benefits, CalPERS sent projections to these agencies in which a CalPERS actuary presented to elected officials three distinct values for the assets they had invested with CalPERS. Remarkably, that document gave these agency officials the liberty to choose which one they’d like to use – the higher the value they chose for their existing assets, the lower the cost from CalPERS to pay for the benefit enhancements they were contemplating. The usual disclaimers were present, but the mere fact that city officials were given three scenarios is suspect. Obviously these officials would be under pressure to pick the scenario that provided the biggest benefit enhancement for the lowest cost. Read “Did CalPERS Fail to Disclose Costs of Historic Bump in Pension Benefits?” for more details including several source documents.
One of the most revealing documents is exemplified by the “Contract Amendment Cost Analysis,” sent to Pacific Grove by CalPERS in July, 2001. Here is an excerpt from that document, showing the choices CalPERS offered Pacific Grove:
The available rate choices are offered under three different Alternatives:
Alternative 1 – No increase in Actuarial Value of Assets
Alternative 2 – Actuarial Value of Assets increased by twice the increase in the Present Value of Benefits due to the amendment, limited to 100% of Market Value of Assets
Alternative 3 – Actuarial Value of Assets increased by twice the increase in the Present Value of Benefits due to the amendment, limited to 110% of Market Value of Assets
To reiterate: CalPERS provided abundant disclaimers. They suggested that given recent “market volatility,” city officials “are strongly encouraged to have in-depth discussions with your CalPERS actuary about the financial consequences of any amendment.”
Now let’s get real: Further on in this same letter, CalPERS provides a breakdown of how much pension benefit enhancements will cost in terms of annual contributions as a percent of payroll under each of these three scenarios:
Alternative 1 – The actuarial value of the assets is not tampered with, the normal cost goes from 4.6% to 25.0%.
Alternative 2 – The actuarial value of the assets is lifted up to market value, the normal cost goes from 4.6% to 19.9%.
Alternative 3 – The actuarial value of assets goes up to 110% of the market value, the normal cost – to implement a massive, retroactive enhancement to pension benefits – goes from 4.6% to 6.2%.
What option would you choose, if you were a city manager whose own pension would be enhanced, or a city council member who has to answer to powerful unions whose members want more generous pension formulas?
The reason CalPERS was able to cram this through, in July 2001 as the market was cratering, was based on their decision to present various asset “smoothing” options to members. Why? Because the smoothing options they’d been using were understating the value of their assets because stock values had exploded in the final years of the 1990s. One can only speculate as to why they did this as late as July 2001 when it was obvious the internet stock bubble had popped. It’s possible CalPERS officials knew several agencies had already lobbied for pension benefit enhancements and the officials were under pressure to leave no agency behind. But to offer local bureaucrats and elected officials a choice of various asset smoothing methods was passing the buck.
Overnight, the CalPERS practice of asset smoothing went from being a prudent accounting guideline to a clever rationalization for disastrous policy decisions. If that’s not a gimmick, I don’t know what is.
GIMMICK #2 – CREATIVE AMORTIZATION OF UNFUNDED LIABILITY
When you talk about “tricky accounting gimmicks,” it’s hard to find one worse than the methods the participating agencies chose to amortize their unfunded liability. To be fair, final responsibility for these decisions usually rests with the cities and counties. But CalPERS should have tried to crack down on these practices a long time ago, and indeed, has recently become more aggressive in doing just that. The basic choice facing agencies with huge unfunded liabilities is whether they want to pay them off aggressively, or come up with creative accounting techniques that push the tough repayments into the future. For example, instead of using a “level payment” repayment calculation, many of them use a “percent of payroll” scheme which allows for graduated payments.
In practice, this means calculating a stream of payments that will pay off the liability in 30 years, but varying the payments so that as projected payroll increases, the payment increases. This allows agencies to make low payments in the early years of the amortization term, which frequently means the unfunded liability isn’t even being reduced in the early years of the amortization term. Then when the payments become burdensome, they refinance the new, larger unfunded liability, to get that unfunded payment down again, in a new tranche, again using the same “level percent of payroll method.”
Anyone who lost their home because a “negative amortization” loan conned them into buying something they couldn’t afford would likely call that type of loan a “gimmick.” Similarly, negative amortization payment schedules on unfunded pension liabilities are also gimmicks. To their credit, CalPERS is now recommending 20-year straight line amortization. Which begs the question, why didn’t they do this all along?
GIMMICK #3 – OVERESTIMATING LONG-TERM RATE-OF-RETURN ASSUMPTIONS
CalPERS spokesperson Morgan correctly claims that CalPERS returns have averaged an 8.4% return over the past 30 years. But Morgan conveniently selects the 30 year timeframe to capture all of the pre-1999 run-up in stocks that began in the Reagan years as interest rates were reduced from inflation-fighting highs of 16% (30 year T-bill in the early 1980s) and American consumers began piling on debt. The 20-year return for CalPERS investments through June 30, 2017 is 6.58%. And these last 20 years of returns are far more relevant, because not quite 20 years ago is when CalPERS began to offer pension benefit enhancements that were sold as affordable when they clearly are not.
But if CalPERS is exceeding its projected rates of return over the past 30 years, why is it only 68% funded (ref. CalPERS 2016-17 CAFR, page 4, “Funding”)? At the end of a prolonged bull market, pension systems should be overfunded. Being 68% funded would not be terribly alarming if we were at the end of a prolonged bear market, but we’re in the opposite place. How can CalPERS possibly claim their actuaries are doing a competent job, if the system is this underfunded at this point in the market cycle? For more on this, read “If You Think the Bull Market Rescued Pensions, Think Again.”
It is important to emphasize that even if CalPERS can get a 7.0% return on investment – and there is some chance that they can – why did the agency wait until it was 68% funded to announce the drop in its projected returns from 7.5% to 7.0%? The United States economy is in the terminal phases of a more than 60 year long-term credit cycle, and one might argue there is a stronger case to be made that even 7.0% is highly optimistic. But we like optimism, so never mind that for now. Why wait until 2018 to phase in that half-point drop? The actuaries at CalPERS are well aware how sensitive their payment schedules are to even half-point drops in long-term rate-of-return assumptions. Overstating returns understates true cost. Is this an accounting gimmick? Only if you can prove intent. But read on.
GIMMICK #4 – QUIETLY ALLOWING THE UNFUNDED PAYMENT TO DWARF THE “NORMAL” PAYMENT
Every year, each active worker who gets CalPERS benefits vests another year of service. This means that in the future, during their retirement years, they will have an incrementally greater pension benefit in recognition of one more year of work. To pay for that incrementally greater pension benefit in the future, additional money must be invested today. That amount of money is called the “normal” contribution. But when the “normal” contribution isn’t enough, and it hasn’t been for years, the so-called unfunded liability grows. This unfunded liability represents the amount by which invested pension assets need to increase in order to earn enough to eventually pay for all the future pensions that have been promised.
This “unfunded liability” may seem theoretical when a pension system has hundreds of billions in assets. But it has to get paid down, because when there aren’t enough assets in the pension system earning interest, higher contributions are inevitably required from the participating agencies. If the unfunded liability isn’t reduced via catch-up payments, it will grow even if the normal contributions are adequate to cover newly earned benefits.
This reality is corroborated using CalPERS’ own data, which announces that payments required, as a percent of payroll, are set to increase by 50% (in some cases much more) over the next six years in nearly every agency it serves. And where are these projected increases most pronounced? In the unfunded contribution – that payment to reduce the unfunded liability.
And why does the unfunded liability grow in the first place? Because the normal contribution is too low. Why is the normal contribution too low? Could it be because public employees are only required to assist (via payroll withholding) to pay the normal contribution? Could that be the reason that lifespans were underestimated and returns were overestimated? The actuaries obviously got something wrong, because CalPERS is only around 68% funded. You can download the spreadsheet that shows the impact of this on California’s cities and counties here – CalPERS-Actuarial-Report-Data-Cities-and-Counties.xlsx.
In the original CPC report, along with the term “gimmick,” the term “outrageous” was used. If you don’t think sparing the beneficiaries of these pensions any responsibility to share in the costs to pay down the unfunded liability isn’t outrageous, you aren’t paying attention. For example, by 2024, using CalPERS own data, the City of Millbrae will be paying CalPERS a normal contribution of $1.0 million, and an unfunded, or “catch-up” contribution of $5.8 million – nearly six times as much! Is Millbrae just an isolated example? Not really.
Again, using CalPERS’ own data, in 2017-18, their 426 participating cities will contribute $3.1 billion to CalPERS, an amount equal to 32% of their cumulative payroll. In 2024-25, just six years from now, they are estimated to contribute 5.8 billion, 48% of payroll. And the normal vs unfunded contributions? This year in the cities in the CalPERS system, 13% of payroll constitutes the normal contribution and 19% of payroll constitutes the unfunded contribution – for which current employees and retirees have no responsibility to help pay down. In 2024-25? The normal contribution is estimated to increase to 16% of payroll, and the unfunded contribution, rising to $4.0 billion, is estimated to increase to 33% of payroll.
Put another way, today the unfunded “catch-up” pension contribution for California’s cities, cumulatively, is 140% of the normal contribution. By 2024-25, that “catch-up” contribution is going to be 210% of the normal contribution, more than twice as much! And participating individual employees and retirees have zero obligation to help pay it down, even though that payment is now twice as much as the normal payment.
But it’s not the fault of the individual beneficiaries. The responsibility lies with CalPERS and the politicians they reassured for all these years, using gimmicks.
Let’s review these practices: (1) Letting the agencies decide which type of asset smoothing they’d like to employ, (2) permitting the agencies to make minimal payments on the unfunded liability so the liability would actually increase despite the payments, (3) making overly optimistic actuarial assumptions, (4) not taking action sooner so the unfunded payment wouldn’t end up being more than twice as much as the normal payment.
“Gimmicks”? You decide.
THE CASE OF MILLBRAE
On January 22, the San Mateo Daily Journal published an article entitled “Millbrae officials question, criticize pension cost report.”
The paper’s Austin Walsh reports that Millbrae officials told him that using staffing projections to calculate Millbrae’s future pension burden won’t work because Millbrae has fewer employees than most municipalities. Here’s how Millbrae’s Finance Director DeAnna Hilbrants put it: To limit pension costs, Millbrae contracts for positions in police, fire and public works departments. Quote: “Most notably, Hillbrants pointed to Millbrae joining the Central County Fire Department with Burlingame and Hillsborough and contracting with the San Mateo County Sheriff’s Office for law enforcement.”
What Millbrae officials are saying is that because they contract out much if not most of their personnel costs, their pension contribution is a small percent of their total budget. What they neglect to acknowledge is the fact that the Central County Fire Department and the San Mateo Sheriff’s Office themselves have pension costs, which are passed on to Millbrae to the extent Millbrae uses their services. Millbrae may have made a financially beneficial decision to outsource its public safety requirements. But they did not escape the pension albatross.
CALPERS IS NOT UNIQUE
What has been described here does not just apply to CalPERS. It is the rule, not the exception, for every one of California’s pension systems to engage in the same gimmickry. The consequences for California’s cities, counties, agencies, and system of public education are just beginning to be felt.
* * *