Jane the Actuary Retirement

Forbes post, “Chicago Is Considering Issuing A Pension Bond – And Taking A Gamble”

Originally published at Forbes.com on September 15, 2020.

 

How is Chicago going to solve its budget crisis? Two weeks ago, when Mayor Lightfoot gave her budget forecast address, she spoke of dreams for a “world-class” entertainment district surrounding a Chicago casino, the revenues from which are slated to fund the police and fire pension funds. But, it turns out, that may not be the only way in which Chicago attempts to use gambling to resolve its pension woes.

Let’s recap:

Chicago, we learned in late August, is facing a serious budget hole as it ends 2020, and an even worse one in 2021, to the tune of $1.2 billion — or, expressed differently, 25% of the total corporate budget. Mayor Lightfoot hopes to find some coins in the metaphorical couch cushions by refinancing debt, and is urgently pleading for more federal money. Why, their plight is so serious that they may even go to extremes such as reducing payrolls or cutting pay. (Yes, that’s right – unlike elsewhere, neither Illinois nor Chicago has yet made any spending reductions.)

Now we learn, according to Bloomberg (via Yahoo on September 3), that the city is once again considering a pension obligation bond.

“Chicago is looking to the $3.9 trillion municipal-bond market for options to close its ballooning budget deficits, Chief Financial Officer Jennie Huang Bennett said.

“Options on the table include selling pension obligation bonds, as well as refinancing general obligation and sales tax-backed bonds, Bennett said in a telephone interview on Wednesday. . . .

“The city has mulled pension obligation bonds in past years. Former Mayor Rahm Emanuel had considered issuing $10 billion of them to cover rising costs of public employee retirement funds. The pension costs have weighed on the city’s credit rating for years with Moody’s Investors Service giving the city a junk rating in 2015. The city’s unfunded retirement liability stands at about $30 billion. ‘Everything is on the table,’ Bennett said. ‘We’ve spent time analyzing a pension obligation bond, what the pros and cons are and have had a number of conversations about what that could mean for the city.’ . . .

“Any discussion about pension obligation bonds should be paired with potential reforms for how the city pays for retirement costs and what benefits are provided, she said. Bennett declined to comment on specifics for reforms but said the city plans to have conversations with groups including beneficiaries.”

Of course, without a amendment to the state constitution, the hands of state and local government are so tightly tied that it’s hard to imagine what sort of reforms Bennett has in mind.

But there’s a bigger question with the prospect of issuing pension bonds: how would they actually solve the city’s short-term cash flow problems?

After all, neither of the two usual pitches for pension obligation bonds promises short-term cash flow savings.

The first is a straight arbitrage game: giving the pension fund an immediate boost in funded status and, in the long run, earning more in investment returns than you pay in interest on the bonds. It’s as if you used a cash-out home mortgage refinance to invest in the stock market — and, indeed, there have been experts (or people with a credential or a title sufficient to gain publication) calling for cities and states to do exactly this, considering the time to be exactly right for a stock market rebound and low rates on the bonds themselves. (It’s important to know that interest rates are not spectacularly low for these types of bonds because, unlike public works, these are just as taxable as corporate bonds.)

This is, in part, what the state of Illinois did in its historic-at-the-time $10 billion bond sale in 2003, under Gov. Rod Blagojevich, selling bonds, plowing the money into the state’s retirement systems, and using some of the money otherwise earmarked for retirement contributions to cover the debt service payments on those bonds.

But the second pitch is a claim that, like refinancing, this is a way to save on interest expense. Rather than “paying” interest at the rate of 7% in the actuarial valuation, you pay the bond’s interest rate, at, say, 4 or 5%. But I’ve put the word “pay” in quotes in connection to the actuarial valuation, because that savings is illusory. The valuation interest rate is an artificial rate determined for the purpose of measuring the liabilities, at a specific point in time, of all the benefits to be paid out in the future. The real cost is those future benefits and using a Pension Obligation Bond is simply a means of making the current year’s financial reporting look better. (See here for more details and examples.)

This means that, if your concern is the city’s debt, and financial reporting, and bond ratings, then a pension obligation bond might be an attractive way to try to game the system.

But the math around the city’s required annual contributions to the pension fund is a different matter.

Here’s a refresher:

The city of Chicago has four different pension systems (parks, teachers, transit, etc., workers are separate entities). The Municipal plan (23% funded, $16.8 billion in liability) and the Laborer’s plan (41% funded, $2.7 billion in liability) are both on a “ramp” of fixed contributions up until 2023, and thereafter contribute a level percentage of payroll to achieve 90% funding in 2058. The Police (22%, $13.3) and Fire (17%, $6.3) each are in the last year of fixed contributions now, then have a funding target of 90% in 2055.

This means that, in total, the plans had a dramatic jump in contributions of 29% in 2020, and will have further future jumps of 7% in 2021 and 25% in 2022. What’s more, there are even more dramatic jumps in the impact on the city’s budget, because they pay the marginal additional contribution after certain portions are already covered through other dedicated taxes. In the city’s current budget and its projections, contributions increase from

  • $335.5 million in 2020, to
  • $426.9 million in 2021, to
  • $685.3 in 2022, and
  • $859.1 in 2023.

(There are also differences in timing between plan years and when contributions are actually made.)

But what happens if a pension obligation bond is added? Would the city reduce its actual cash flow rather than financial reporting?

That, it turns out, depends on the city’s decision-making.

Here’s one example of a way that the city could manage pension bonds to — if you jump past all the warnings not to issue bonds in the first place — improve its funded status immediately and its contributions over the long term: issue bonds of $5 or $10 billion for the Municipal plan, and $5 or $7.5 billion for the Police plan, pay the bonds of in an amortized (level payments) fashion, recalculate the annual contributions according to the same formula but the new-and-improved funded status, and enjoy a growing level of savings over time. In the case of the Municipal Employees’ plan, a $5 billion bond would boost the funded status to nearly 50%, a $10 billion bond to a bit over 75%. For the Police plan, $5 billion would be a boost to almost 60%, and $7.5 billion, almost 75%.

And the effect on contributions over time would look something like the following charts, if, hypothetically, the assets returned exactly the plans’ forecasted investment returns each year:

(I’m basing all my calculations on the 2018 Municipal Employees’ actuarial report issued in 2019 and the same year for the Police; an updated version exists for the Muni and Police but had not when I first worked through similar calculations in March of this year, and plan actuaries would generate somewhat different numbers as I’m without access to the particulars of timing and calculation methods and am taking certain shortcuts in my math. And I’ve just taken these two as examples because they’re the largest.)

For the Municipal Employees’ plan, the city’s contributions right now are still on the “ramp” and artificially low; for illustration purposes the hypothetical “with bond” contributions assume that the ramp is gone in a new contribution (because, honestly, the math is messier otherwise).

Why doesn’t the state see immediate savings? Two reasons: a 30 year bond is a shorter amortization period than the current time ‘til 90% funding, 39 years for the Municipal plan and 35 years for the Police plan. And the current contributions increase each year at the rate at which pensionable pay increases; in my example, as with a mortgage you or I might have, the payments are level.

But that doesn’t seem to be what the city has in mind.

Regrettably, there is precedent for the city simply issuing bonds for the sole purpose of making its required contributions and, once again, putting future taxpayers in debt to pay current obligations — that is, in 2003, only $7.3 of the state’s $10 billion in bonds was used to boost the pension funding levels. The remainder was simply used to fund part of the 2003 and all of the 2004 statutory pension contributions. Again, in 2010 and 2011, under Gov. Pat Quinn, the state issued bonds solely to cover its statutory contributions.

And a report issued by Chicago’s City Council Office of Financial Analysis in February of this year (that is, before covid) provided more insight on the method by which Chicago might issue bonds, which aims to be more responsible than simply covering annual contributions but places the city at risk, instead.

And, yes, if you’ve read this far, this is where the city moves from questionable decision-making to gambling, or, in more neutral actuary-speak, unacceptable risk-taking.

Chicago, after all, has a credit rating that’s literally junk grade. Only formerly-bankrupt Detroit is worse. In order to manage to borrow money at rates that aren’t so high as to wholly defeat the purpose of a pension bond, the city would have to provide collateral — specifically, by promising the city’s portion of the state sales tax, because doing so is believed to ensure that those revenues are thus guaranteed to the bond buyers even in the event of bankruptcy (though it hasn’t actually been tested yet). To provide further assurance that the funds would be used for pensions, the city would create an entity, the Dedicated Tax Securitization Company, which would be responsible for managing the entire process: it would issue 30-year bonds with interest-only “coupon” payments, invest the money, and, each year for 30 years, earn 6.3% in investment income (according to the report’s assumptions, pay 4.62% in interest to bondholders (again, according to assumptions), and use the leftover investment income to reduce the amount of the city’s annual contribution, keeping the principal untouched to pay back in 30 years.

A $10 billion bond could generate $168 million in free money per year.

Right?

Of course not.

Each year that the investment returns exceed the interest rate, the city has money to use to reduce the annual pension contributions.

Each year that the investment returns are lower than the interest rate, the city has to pay more to make up the difference.

Yes, in the long term, on average, returns will exceed the interest rate. But can the city really afford to have its costs bounce up and down each year in this way?

The report acknowledges this risk and proposes that in years of positive net returns, the city reserve some for a rainy day, but acknowledges:

“This would greatly reduce the City’s exposure to short-term downturns, but might not fully protect it if a downturn occurred in the early years of the POB, before an adequate investment capital cushion had grown.”

Now recall that this report was issued on February 7, and former Mayor Emanuel first discussed bonds a year before then, and try to imagine Chicago’s budget shortfall if, in addition to all the other financial woes, the city had to find the cash to top up its pension bond debt service.

So here’s the bottom line: the city really has no particularly good options. But the city risks making a bad situation even worse, if it looks for solutions that promise a free lunch rather than making hard choices.

 

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.

Jane the Actuary Retirement

Forbes post, “Forget Pension Obligation Bonds. Two Cities Are – No Joke – Leasing Their Streets To Fund Pensions.”

Originally published at Forbes.com on September 2, 2020.

 

It sounds preposterous, and the headline of a recent article here at Forbes by Marilyn Cohen is certainly eye-catching: “The Lunacy Of Using City Streets To Collateralize New Municipal Bond Deals.” And these aren’t just any municipal bond deals — two cities in California are issuing bonds with their own city streets as collateral to pay down their unfunded pension liabilities.

In West Covina, the city council voted to do so on July 7, as reported at the San Gabriel Valley Tribune. The city, a suburb of Los Angeles with a population of 100,000, a median household income of $71,200, and nearly $200 million in pension liabilities, is using the proceeds of $205 million in debt to pay off its own debt to CalPERS.

Likewise, according to the East Bay Times, the city of Torrance, also in suburban Los Angeles, population 150,000, median household income $80,900, pension debt $500 million, will issue $350 million in bonds. (See the formal report of the recommendation and the minutes of the July 28 city council meeting.)

Now, it turns out, they’re not turning their streets into toll roads, or giving bond-buyers the ability to “foreclose” or take control either now or in the future.

They’re using a bond-issuing mechanism called “lease revenue bonds.” We’re all used to cities paying for public works, stadiums, and the like by issuing bonds which are paid off by a dedicated revenue source — sewer bills, hotel taxes, etc. But lease revenue bonds are different. Here’s the layperson’s description at Charles Schwab:

“Lease revenue bonds are a unique structure in the muni market. Instead of issuing long-term debt, like general obligation bonds do, to finance improvements on a public facility, the municipality may enter into an arrangement that uses lease revenue bonds. Often a trust, not the municipality, issues bonds and generates revenues to pay the bonds back by leasing the facility to the municipality. The municipality will generally appropriate money during each budget session to meet the lease payment.

“Bonds backed by structures with lower essentiality and limited protections for appropriating funds will usually be lower-rated and have higher yields. Our opinion is to be cautious of bonds backed by lease revenues, as these bonds should be viewed more like general government bonds, not revenue bonds.”

This means that the city of San Francisco used lease revenue bonds to buy items ranging from hospital beds to a witness protection van. And Torrance and West Covina are each using these bonds to, in principle, lease their city streets to a special Financing Authority, which will pay the city their up-front money, and “rent” the streets back to the city for the 25 year term of the agreement, in order to pay off the bonds.

Despite the fact that the streets are nominally being “leased,” the bondholders will not have any particular rights to lay claim to the streets; despite their status as “collateral,” the bondholders can’t take them over and charge tolls if either city defaults on their “rent” payments. The city will simply pay the “rent” based on their ordinary tax revenue rather than any special purpose taxes. The “lease” component then becomes little more than a gimmick, a loophole, a way to use the existing “menu” of bond choices available to them in the most advantageous way possible, especially since, at least in California, “general obligation bonds” require voter approval.

(Lease revenue bonds exist at the state level, too; and a group opposing the construction of prisons has a helpful explainer on these due to their use for that purpose.)

What, then, is the purpose of a lease revenue bond in this case? The Bond Buyer explains that these are functionally pension obligation bonds, but can be implemented more quickly, citing Mike Meyer of NHA Advisors: “Depending on the legal structure, there may be added flexibility for use of proceeds to CalPERS or more strategic timing of investing in the market. . . . These things aren’t possible under a traditional POB structure.” At the same time, there’s a trade-off, as rating agencies rate pension obligation bonds more highly than lease revenue bonds. Brian Whitworth, director of Hilltop Securities, which underwrote the West Covina bonds, further claimed, “This is the fastest form which the city would be able to use and issue bonds.”

And why are the cities in such a hurry to issue these bonds? In one respect, it’s the same rationalization as appears every time pension obligation bonds pop up, the notion that they are “refinancing” a debt at a lower interest rate, because of the difference in rates between the bond rate, and the interest being accrued on the books, at the higher actuarial valuation rate — so, for example, a 7% rate appears to be dropped to a 4% rate due to the “savings” of “refinancing.” (See my explainer from 2019, when this was a hot topic in Chicago.) This is a mirage, though — since it’s all just a matter of how liabilities are accounted for; their true cost is the payment of pension benefits in the future, regardless of what the plan account is now. And the nature of a pension obligation bond, the hope to get a higher asset return for the money you’ve borrowed at a low bond rate, remains the same.

Now, to be sure, there is a further wrinkle in California. The state agency CalPERS manages their pensions, and prescribes a required annual contribution. This makes it all the more difficult to perceive that pension bonds’ “savings” come solely from the hope of higher asset returns than bond interest rates (which are, incidentally, fully-taxable rather than offering the investors the benefit being of tax-free).

And what are those annual contributions? The most up-to-date reports on the CalPERS website are from July 2019, based on June 30, 2018 and calculating the required contributions for the 2020 – 2021 plan year. The city of West Covina pension plan is 71% funded, but to pay down its underfunding and fund new accruals, must pay 44% of payroll. The West Covina public safety plan is 62% funded and requires a contribution of 74% of payroll to fund new accruals and pay down underfunding. The Torrance city pension is 79% funded with 24%-of-payroll contributions; the Torrance fire pension, 65% funded, 68%-of-payroll contributions; and the Torrance police pension, 62% funded, 78% of payroll contributions. What’s also important to know is that these high contributions are not the result of having to make up underfunding in an unreasonably-short period of time; the underfunding level as of 2008 was set at a 30 year amortization, and gains and losses since then are likewise given 30 years to be paid off. This means that the high contributions are simply a reflection of the high cost of the pensions themselves, and the tremendous impact of even marginally-poor funding levels.

 

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.

Jane the Actuary Retirement

Forbes post, “No, Pension Obligation Bonds Aren’t A Form Of ‘Refinancing'”

Originally published at Forbes.com on February 1, 2019.

 

That’s the claim, for instance, made by outgoing mayor Rahm Emanuel about the benefits of creating Pension Obligation Bonds, in his December City Council speech, as reported at the Sun-Times:

“We can refinance a portion of that debt at lower rates, locking in savings of as much as 2.5 percent over 40 years. Now, that works out to between $6 billion and $7 billion in savings for Chicago taxpayers,” Emanuel said.

Chicago Alderman Patrick O’Connor, the new Finance Committee chair, repeated that phrasing more recently in an interview promoting the need to prepare for a future bond offering:

“So, in order to avail yourself of the opportunity to refinance, you would need to have this thing in place to do it when the new council comes in.”

Are you scratching your head wondering what this means?  Yes, it’s time for more actuary-splaining, but I really think it’s important to understand what’s going on here — and that’s true for decisionmakers and for voters alike.  (Not sure what I’m talking about?  Get up to speed with my prior article on the topic.)

Certainly, promoters of Pension Obligation Bonds are hoping to capitalize on the perception of “refinancing” as an ordinary household money-saving tool.  Consumers with credit card debt can reduce the amount they spend on interest by taking out a home equity loan with a lower interest rate, then paying off their debt.  Likewise, if your mortgage has an 8% interest rate and you have the chance for a 4% interest rate, you can reduce your spending on interest by paying off the 8% mortgage with the proceeds from a new 4% mortgage.

The city owes pension liabilities.  Are they going to pay these off with money borrowed at a lower interest rate?

In principle, they could indeed “pay off” the benefits, if they purchased annuities for retirees, or if they offered lump sum buyouts to plan participants.  And corporations are doing this increasingly often, have determined that they can save money this way, or, for the same cost, reduce their risk.  They do the math and compare their pension funding projections, which includes the direct plan liabilities as well as administration expenses and the premiums they pay to the PBGC (the government agency that protects pensions), to the money they’d spend to buy annuities or on participant buy-outs, and if the latter is a better deal they implement a program.  (Incidentally, when employers offer buyouts, they have to offer an actuarially-fair lump sum, and are not permitted to make lowball offers.)   This is as close as a plan sponsor can come to “refinancing” because they truly “pay off” their pension debt, whether they use cash-on-hand or even issue bonds to do so.

But the city can’t do that.  Here’s why: public pensions are, with few exceptions, valued using the expected return on assets as their valuation interest rate, rather than, as with corporate pensions, a bond rate.  As long as they invest in risky/return-seeking assets, this means that, for the same set of obligations in terms of future benefit payments owed, a city’s reported liability will be lower than that of a private-sector company.  When a city does the same math, the cost of buying annuities is considerably higher than the (artificially) lower liability valuation.  And workers would be foolish to accept a lump sum at lower value.

So why does POB supporters use the label “refinancing”?

If you squint hard, it makes sense.  What’s going on is this:  they are reducing the annual expense reported in their financial statements.  Here’s my best attempt at explaining this for non-experts.

Let’s start with the basic financial data for the largest Chicago plan, the Municipal Employees’ Annuity and Benefit Fund, that I’ve been digging into in previous articles, based, again, on the latest actuarial report as well as my own calculations.

MEABF Core financial reporting data

 own work

The key balance sheet figures are the liability, assets, and the funded status.  But the city also discloses its cost for the year, using the Service Cost — the employer’s share of new benefit accrual, after subtracting out employee contributions — and the net interest cost, that is, one year’s interest accrual on the unfunded amount.  (If you read my prior actuary-splainer, you’ll recall that this is a key driver in the increasing unfunded amount.)  In addition, there are additions or subtractions based on gains and losses in investments, plan experience, changes in assumptions, benefit increases, but that doesn’t really factor in to this question.

But what happens if the city issues a $10 billion bond and directs the proceeds into the pension fund?  For simplicity, let’s assume that the total amount goes into this one pension plan.  We’ll assume the city has to pay interest of 5% on this bond.  What happens?

Here are those same numbers, with a pension bond added in:

POB effect

 own work

Here’s where the cost savings comes in:  because assets are boosted and the funded status drops, the magnitude of the interest cost (at 7%) drops by a relatively greater degree than the additional interest from the pension bond at 5%.  The city records on its books for the year an amount that’s $200 million less than it otherwise would have been.  It looks like a huge win, before even taking into account the hoped-for asset return in excess of the interest paid out to bondholders.

But this is all contingent on using a liability valuation interest rate that’s higher than the bond rate the city hopes to pay.

Here’s the same calculation repeated using a 4% interest rate, which is closer to what a private-sector business would be using to disclose its pension liabilities.  (This calculation was based on sensitivities disclosed in the actuarial valuation and a pension math-specific extrapolation.)

Municipal Employees at 4%

 own work

At this rate, liabilities are 50% higher than what’s in the city’s financial reporting, and the funded status drops from 28% to 19%.  The Service Cost increases, too (and the amount of increase is actually greater than shown in this estimate, due to its approximations).  But the interest cost actually drops, because it is based on the interest rate being used — though in this case, the lower amount is not enough to offset the higher Service Cost.

Finally, one more table:  what would happen to the Pension Obligation Bond savings if the city used a more conservative valuation rate, similar to private-sector accounting?

uncaptioned

The answer, of course, is that it would vanish.

What does all this mean?

To back up for a minute:  the city’s debt for this plan does not consist of $16 billion in liability at 7% or $24 billion at 4%, or $12 billion or $19 billion after subtracting out assets.

That’s just a set of numbers used to report those liabilities in a standardized, transparent way.

The real liability of this plan is the 975 million anticipated to be paid out in benefit payments in 2019, the billion in 2020, the 1.1 billion in 2021 and so on.

Whether a 4% rate or a 7% rate or another rate is chosen to determine the present value of those benefits for financial reporting purposes will not change the future benefit payments.  And issuing a bond and reducing the expense reported, in total, because of a lower interest rate, does not impact the ultimate cost of those benefit payments.

The liabilities are still out there.  The bonds have to be paid back.  The taxpayers gain only if the actual return on the money invested with those bonds exceeds the interest rate the city pays to bondholders.  If the expected return is overstated, then the city will record asset losses in future years.  On the other hand, if the city reported at the lower 4% rate, and actually saw returns on its investments greater than that rate, the city would be reporting gains instead.

So the next time a politician says that Pension Obligation Bonds are a way to refinance to save money, be very wary of their claims.

 

December 2024 Author’s note: the terms of my affiliation with Forbes enable me to republish materials on other sites, so I am updating my personal website by duplicating a selected portion of my Forbes writing here.