The Impact of Moody’s Proposed Changes in Analyzing Government Pension Data

January 8, 2013

By John G. Dickerson

About the Author:  John Dickerson is a financial professional living in Mendocino County who is involved in public sector pension analysis and reform. Since 2008 Dickerson has been the publisher of the influential website and newsletter www.YourPublicMoney.com. He focuses on the impact of unfunded pension debt on the 21 California counties that operate their own independent Pension Funds. He is a financial and organizational planner and analyst with 30 years of experience. He earned his MBA in Strategic Planning from the University of Texas at Austin. The original version of this study was published on January 3, 2012 by Dickerson and can be downloaded from his website. This version, incorporating minor changes to the original, is presented here in html or as a PDF document to download and print by the California Policy Center.

ABSTRACT

On July 2, 2012 Moody’s Investors Services published a “Request for Comment” titled Adjustments to US State and Local Government Reported Pension Data. Moody’s is one of the nation’s major credit-rating agencies for state and local governments. As such, while Moody’s doesn’t have authority to make governments change their financial statements or fund pensions differently, they do control how they analyze and report government credit-worthiness. Moody’s has concluded that published government employee pension financial data greatly understates the credit risks created by unfunded pensions. Moody’s proposed adjustments are likely to lead to reductions in credit ratings for many governments which could cause them to pay more interest expense and/or reduce access to credit and loans.

The main importance of Moody’s proposals to concerned citizens is they strongly support the view that unfunded pensions put state and local government finances at great risk, much more than is reported to the people. They help explain how unfunded pensions produce much greater risk and by implication what to do about it.

Moody’s adjustments would have two major impacts on most governments. First, Moody’s states these adjustments would triple the amount of unfunded government pension debt across the US they would use to set credit rates. Second, Moody’s analysis will conclude that most governments are paying far less to their Pension Funds than they should.

Moody’s would make four adjustments – two are very significant. First, pension debt would be adjusted using a high-grade long term corporate bond rate (5.5% for 2010-2011) instead of a Pension Fund’s target rate of return (7.75% more or less). Second, government payments to Pension Funds would be adjusted to reflect the lower discount rate, the need to fully fund pensions by the time employees retire, and a 17 year level-dollar amortization of unfunded pensions.

I developed a financial model to project how Moody’s adjustments would restate published government pension data. I applied the model to the 6 Bay Area – North Coast California counties that do not participate in CalPERS; they have independent County Pension Funds (Alameda, Contra Costa, Marin, Mendocino, San Mateo, and Sonoma). The model uses data from recent Actuarial Valuations. It produces four core restated values – total pension debt, unfunded pension debt, government normal yearly contributions, and government unfunded pension amortization payments.

These County Pension Funds reported total unfunded pension obligations were a little over $4 billion. Moody’s adjustments would increase these unfunded pension obligations by about another $6 billion which would reduce average reported pension funding ratios from 78% to 58%.

Under today’s accounting rules governments don’t report unfunded pensions as debt directly in their financial statements. (New government accounting rules will make them to do so in two years.) Moody’s would include the restated $10.2 billion unfunded pension debt in its credit rating process. In addition these six counties reported Pension Obligation Bond (POB) debt as of June, 2011 of $1.7 billion. They borrowed that money to pay down earlier large unfunded pension deficits. Therefore, including Pension Obligation Bonds, the total unfunded pension-created debt using Moody’s adjustments would be close to $12 billion. All other reported debt (not including unfunded retiree healthcare and other post-employment benefits) was $2.8 billion. Thus Moody’s adjustments would increase the total debt for these counties used in their credit rating analysis from the reported $4.5 billion to $14.6 billion – slightly more than triple what is currently reported.

These counties pay about $640 million each year to their Pension Funds. These adjustments would increase this annual payment to $1.4 billion – from 29% of payroll to 63%. To put this in perspective, payments to Pension Funds and Pension Bonds today consume about half of these counties independent property tax income. These adjustments show they should consume all county property tax income.

Moody’s stated they would recalculate total debt for both state and local governments but would calculate what “prudent” government payments to pension funds should be only for states. I strongly urged Moody’s to apply their “prudent payment” adjustments to local governments as well. Moody’s has not yet announced their final decision.

OUTLINE

    1. Introduction
    1. Moody’s Four Adjustments of Government Credit Analysis.
    1. First Major Impact – Unfunded Pension Debt.
        1. Pension Fund Asset Value.
            1. How Actuaries Calculate Pension Fund Asset Value.
            1. Moody’s Proposed Adjustment of Pension Fund Asset Values.
        1. Total Pension Liability.
            1. First – Estimate Future Pension Payments That Have Already Been Earned.
            1. Calculate Net Present Value of Future Payments Already Earned.
        1. Net Pension Liability – or Asset.
        1. Two Further Considerations.
            1. The Myth that “80% Funding is OK”.
            1. Pension Obligation Bonds Must Be Included in Total Unfunded Pension Debt.
    1. Second Major Impact – Government Payments to Pension Funds.
        1. Two Main Types of Government Payments to Pension Funds.
            1. Normal Yearly Contributions.
            1. Unfunded Pension Amortization Payments.
            1. Other Payments.
        1. Why Moody’s Should Restate Payments by Local Governments – Not Just State Payments.
            1. Many Local Governments are Larger than Many States.
            1. Availability of Data.
            1. The Threat to Owners of California Local Government Pension Obligation Bonds.
        1. Actuarially Calculated Payments to Pension Funds.
        1. Moody’s Adjustments to Government Payments.
            1. Calculate Normal Annual Cost Contribution Payments at 5.5% Rate of Return
            1. Increase Pension Liability Amortization Payments to be “Fully Funded Over a Reasonable Time Horizon”.
            1. Use “Level Dollar” method to Calculate Unfunded Pension Liability Amortization Payment
            1. How Much Will Moody’s Adjustments Increase Catch-Up Payments to Restore Pensions to 100% Funding?
            1. Summary: Impact of Moody Adjustments on Normal and Catch-Up Pension Fund Payments:
    1. Conclusion – What this Should Mean to Concerned Citizens.
    1. Attachment – Data Sources and Footnotes

*   *   *

I.  INTRODUCTION

Moody’s Investors Services and Standard and Poor’s are the most powerful credit rating agencies in the US. They, along with Fitch, are considered the “Big Three Credit Rating Agencies”. [1] On July 2, 2012 Moody’s published a “Request for Comment” titled Adjustments to US State and Local Government Reported Pension Data (referred to herein as “Moody’s Paper”). [2] Moody’s believes government reports about the finances of state and local government pension finances often significantly understate the financial risk of unfunded pension debt. They intend to modify government-reported pension financial data in analyzing credit-worthiness and setting credit ratings for state and local governments in the US.

Moody’s doesn’t expect significant changes in state credit ratings but the weakest state pension funding positions would be identified. Although they expect to reduce credit ratings for local governments whose adjusted debt is deemed excessive Moody’s was still evaluating the extent of likely downgrades. [3] I applied these proposed adjustments to the six counties in the San Francisco Bay Area – California North Coast region that have their own independent County Pension Funds. The results are reported in this paper. The results of applying these adjustments to these six counties suggests that if Moody’s takes its methods seriously there would be a significant number of local government downgrades that would reduce access to debt financing and/or increase the cost of borrowing.

Concerned citizens should understand Moody’s reasoning in making these changes and take the dire warning inherent in what they propose seriously. Many state and local governments are putting their finances at great risk through deeply flawed financial management of their pensions, and their financial reports don’t convey that essential fact.

*   *   *

II.  MOODY’S FOUR ADJUSTMENTS OF GOVERNMENT CREDIT ANALYSIS

These are Moody’s four proposed adjustments [4]:

1. Multiple-employer cost-sharing plan liabilities will be allocated to specific government employers based on proportionate shares of total plan contributions (Note – we don’t examine this aspect in this paper; it’s pretty simple.)

2. Accrued actuarial liabilities will be adjusted based on a high-grade long-term corporate bond index discount rate (5.5% for 2010 and 2011)

3. Asset smoothing will be replaced with reported market or fair value as of the actuarial reporting date

4. Annual pension contributions will be adjusted to reflect the foregoing changes as well as a common amortization period

A significant aspect of Moody’s proposals is to attempt to make government pension financial data far more comparable. Governments and their Pension Funds are free to use assumptions that vary significantly from each other. Moody’s believes these hugely divergent pension funding assumptions creates an analytical “Tower of Babel” that seriously compromises the ability to compare the financial impact of pensions on government finances. Moody’s will eliminate un-standardized adjustments made by actuaries to the value of Pension Fund assets, attempt to standardize assumed investment rates of return, establish a common period in which pensions must be fully funded for current employees, and a common amortization period for unfunded pension debt.

The two major financial impacts of Moody’s proposed adjustments would be [5]:

    • Increased Unfunded Pension Debt:
      Moody’s projects these adjustments would nearly triple reported unfunded actuarial accrued liability (“UAAL”) for the 50 states and local governments in their database to $2.2 trillion from $766 billion divided almost equally between state and local governments.
    • Recalculated “Reasonable” Employer Pension Fund Contributions:
      Moody’s adjustments increase 2010 state pension contributions to $128.8 billion, compared to the $36.6 billion states actually contributed. Moody’s doesn’t intend to restate local government payments.

*   *   *

III.  FIRST MAJOR IMPACT – INCREASED UNFUNDED PENSION DEBT

The calculation of unfunded pension debt is a three-step process described below:
A    Pension Fund Asset Value, less
B     Total Pension Liability, equals
C     Overfunded or Unfunded Pension Liability

A.  Pension Fund Asset Value

Actuaries produce financial analyses of Pension Funds in reports called “Actuarial Valuations” [6] . Moody’s will not use the Pension Fund asset values used by Actuaries.

1.  How Actuaries Calculate Pension Fund Asset Value:

Actuaries almost always make one simple adjustment to the value of a Pension Fund’s assets and often make a second adjustment as well. Actuaries start with the Market Value of Pension Fund Assets.

    • Smoothing: Actuaries use a “smoothed” value of assets. “Smoothing” is a type of “moving average” that “slows down” changes in asset values to prevent chaotic one-year surges in government payments to Pension Funds caused by rapid decline in stock markets. The “smoothed” value of Pension Fund assets is called the “Actuarial Value of Assets”, or “AVA”. Smoothing is usually constrained by a “Corridor Limit” that prevents the smoothed value of assets from being more than a set percentage different from the actual market value. However, the difference rarely is greater than the corridor limit so it is not often used. There is considerable variation in how actuaries apply smoothing and the corridor limit.
    • “Actuarial Value of Assets” (AVA) v. “Valuation Value of Assets” (VVA): Most Valuations use the AVA, but some make a second adjustment. Pension Funds set aside “reserves” for various purposes some of which are not available to pay pensions. These are always only a very small part of the Pension Fund’s assets. Some Actuaries deduct these “non-pension” reserves from the smoothed value of Pension Fund assets – the “AVA” – to produce the “Valuation Value of Assets”, or “VVA”. There isn’t much difference between them.

2.  Moody’s Proposed Adjustment of Pension Fund Asset Values:

It’s simple – Moody’s will use the Market Value of Pension Fund Assets, not the Actuarial or Valuation Value. This will result in greater volatility in Pension Fund asset values. It will also eliminate the considerable “artificial” variation in asset values that results from the wide range of allowable smoothing, corridor limit, and reserve options used by actuaries. The table below shows the change in values for the six county Pension Funds. Pension Fund assets increased for three of these counties and decreased for three. The difference is driven by smoothing methods and timing of Actuarial Valuations. The average pension asset value declined by a little less than one percent.

Dickenson-Moody-table-1

B.  Total Pension Liability

There are two steps. The first wouldn’t be changed by Moody’s – the second would be profoundly changed.

1. Estimate Future Pension Payments That Have Already Been Earned:

The Pension Fund’s Actuary estimates the part of each future year’s pension payments that have already been earned by employees in the past. This is by far the most complicated process performed by Actuaries in Valuations. Thankfully – the result of this step is not the direct focus of Moody’s proposed adjustments. These estimates don’t include the part of future pension payments employees will earn in the future. A government’s total pension liability is entirely created by work performed by its employees in the past. It is part of the cost of providing services in the past – not in the future. [7]

2. Calculate Net Present Value of Future Payments Already Earned:

How much needs to be in the Pension Fund today so that future pension payments that have already been earned can be paid if all assumptions come true? In financial terms this is the total “Net Present Value” of each of those estimated payments in future years.  The most important assumption is the expected annual rate of investment profits the Pension Fund will earn until those future payments are made. Actuaries assume one rate – Moody’s will assume a lower rate.

a) Current Actuarial Calculation – Target Rate of Return is Discount Rate:

Actuaries assume the Pension Fund will earn a “target investment rate of return” (often called an “interest rate”). This shows the assumed rate of investment return (profits) for the six counties in the San Francisco Bay Area – California North Coast that have their own County Pension Funds. These are fairly typical for government Pension Funds.

Dickenson-Moody-table-2

Actuaries use the assumed rate of return to estimate how much should be in the Pension Fund as of their Valuation. This is the “Actuarially Accrued Liability”, or “AAL”. It’s their version of the Total Pension Liability. Moody’s disagrees.

b) Moody’s Adjustment – Target Rate of Return is High-Grade Corporate Bond Rate [8]:

Moody’s proposes to replace the Pension Fund’s target rate of return with a “high-grade corporate bond index” which would have been 5.5% for 2010. They explain their reasoning:

Pension liabilities are widely acknowledged to be understated, and critics are particularly focused on the discount rate as the primary reason for the understatement.(See, for example, Alicia Munnell et al, “Valuing Liabilities in State and Local Plans,” Center for Retirement Research at Boston College, June 2010; Joe Nation, “Pension Math: How California’s Retirement Spending is Squeezing the State Budget,” Stanford Institute for Economic Policy Research, December, 2011; and Robert Novy-Marx and Joshua Rauh, “Policy Options for State Pension Systems and Their Impact on Plan Liabilities,” National Bureau of Economic Research, October 2010.)

In public pension plans, the assumed rate of return on invested pension plan assets is identical to the discount rate that measures the present value of benefits accrued by current employees and retirees. Because plans (often guided by state legislation) develop their own investment rate-of-return assumptions, the discount rate accordingly varies across plans and often among plans within a state. Most public plans currently use discount rates—and assumed rates of return—in the range of 7.5% to 8.25%, which reflects some reductions made in recent years.

We propose replacing the differing discount rates with a common rate based on a high-grade bond index because:

    • Investment return assumptions in use by public plans today are inconsistent with actual return experience over the past decade (when total returns on the S&P 500 index grew at about 4.1% annually) and today’s low fixed income yield environment. According to Wilshire Associates, public plans in the aggregate allocate roughly one-third of assets to fixed income …
    • A high-grade bond index is a reasonable proxy for government’s cost of financing portions of its pension liability with additional bonded debt …

For adjustments to 2010 and 2011 pension data, the proposed discount rate is 5.5%, which is based on Citibank’s Pension Discount Curve. Based on high-quality (Aa or better) corporate bonds, this curve is duration-weighted by Citibank for purposes of creating a discount rate for a typical pension plan in the private sector. The 5.5% rate is a rounded average of the rates published for May, June, and July of 2010 and 2011. This proposed approach to the discount rate is similar to that used in the private sector, where Financial Accounting Standards Board (FASB) regulations require pension systems to discount assets at a rate consistent with the yield on high-quality corporate bonds. We propose to revisit the discount rate annually.

Moody’s doesn’t have the complex projections made by Actuaries for each Pension Fund – they only have the final value reported as the “(Total) Actuarially Accrued Liability”. Moody’s therefore can’t perform a recalculation based on the specific data – so they must come up with a “simplifying calculation” to estimate what the recalculated Net Present Value of the Total Pension Liability would be. They will take the reported “Actuarially Accrued Liability,” project them forward for 13 years using the Pension Fund’s target rate of return, then “discount” the resulting value back 13 years using the high-quality corporate bond rate which was 5.5% for 2010. Moody’s states “a plan with a $10 billion reported AAL based on a discount rate of 8% would have an adjusted AAL of $13.56 billion, or 35.6% greater than reported.”

This table shows the effect of this adjustment on the six Bay Area – North Coast counties:

Dickenson-Moody-table-3C.  Net Pension Liability – or Asset

The third step to determine unfunded pension debt is very simple. Total Pension Liability is subtracted from the value of Pension Fund Assets. This graph shows the proportional change in Net Pension Liability for these six counties.

Dickenson-Moody-figure-1

The table below shows the actuarial calculation and Moody’s proposed adjustment. Moody’s calculates the total value of State and Local Government Pension Liabilities in the nation will roughly triple as a result of this adjustment. The results for these six counties are somewhat less than triple [9].

Dickenson-Moody-table-4

D.  Two Further Considerations

These are two other issues citizens need to consider related to those specifically discussed in the Moody’s paper.

1. The Myth that “80% Funding is OK”:

It’s often said that so-called “experts” say a Pension Funding ratio of 80% is OK. That is a financially absurd assertion [10]. The long-term goal of Pension Funds should be to be 100% funded on average. Pension Funds will never be precisely 100% funded. At the top of stock market cycles the Fund should be around 125% funded. At the bottom – and only at the bottom – the Fund would be about 80% funded. If the long-term average is 80% then a very significant portion of pension expenses incurred by governments to provide services to the people in the past (increased by interest expense) is transferred to future generations. There is no other possible conclusion.

The table below shows the impact of Moody’s adjustments on the “Funding Ratio” of the six Bay Area – North Coast County Pension Funds (Assets/Total Liabilities).

Dickenson-Moody-table-5

 2. Pension Obligation Bonds Must Be Included in Total Unfunded Pension Debt:

Net Pension Liabilities (or UAAL’s) are not the only kind of unfunded pension-created debt. Many governments borrowed money to eliminate Pension deficits by selling “Pension Obligation Bonds” (POB). The Pension Fund got the money – the people through their government kept the debt. Pension Bonds are simply unfunded pensions restructured in the hopes of incurring a lower interest expense. But their source is exactly the same – unfunded pensions.

All too often government and retirement officials only report the Pension Fund’s ratio as reported in Actuarial Valuations. Not only does that ignore the reasons given by Moody’s as to why reported pension funding ratios significantly overstate the real funding position of most government Pension Funds, but it ignores unfunded pension debt in the form of Pension Bonds. It is one of the most obvious examples of how all too many government and retirement officials do not tell the important financial truths about unfunded pension debt to the people.

The outstanding balance of Pension Bonds must be added to Net Pension Liabilities to quantify the total debt created by unfunded pensions. And payments for Pension Bonds are part of total debt payments created by unfunded pensions.

This shows the balance owed on Pension Bonds for the six counties added to Moody’s adjusted Net Pension Debt. The Total percentages are averages of the six percentages for the counties – not based on total dollars. Dickenson-Moody-table-6

The impact of adding the balance of Pension Bonds to Moody’s adjusted Net Pension Liability is profound. Sonoma County’s Actuary reported the pension funding ratio was 84%. Moody’s adjustments reduce that ratio to 59%. But the addition of Pension Bonds drops the ratio falls to 40%. The people of Sonoma County still owe 60% of what should be in the Pension Fund assuming Moody’s is correct. Even if we use the actuarially calculated Pension Funding ratio and deduct the outstanding balance of Pension Bonds Sonoma County still owes more than 1/3 of what should be in the Fund.

It’s extremely important to understand that Actuarial Valuations are in fact government pension funding plans (see “Normal Yearly Contributions”). Simply put – there should never be significant unfunded pension obligations (or at least never more than 20% of total obligations and then only at the bottom of stock market cycles) or Pension Bond debt. The existence of significant balances in either form of unfunded pension-created debt is on its face proof of the failure to achieve their self-proclaimed pension funding goals. On average, given Moody’s adjustments, these six counties and their Pension Funds achieved only half their self-proclaimed pension funding goals.

*   *   *

IV.  SECOND MAJOR IMPACT – GOVERNMENT PAYMENTS TO PENSION FUNDS

Debt wouldn’t be so bad if it weren’t for the annoying habit of creditors to expect to be paid. The creditors in this case are retirees who expect to get their pensions. But governments don’t pay pensions directly to retirees. They make payments to independent Pension Funds. The Pension Funds are “fiduciaries” responsible to make sure that retirees get their pensions. Therefore the immediate creditor is the Pension Fund. Moody’s proposes to make very significant adjustments to restate what governments should be paying to their Pension Funds. The more unfunded pension debt that develops the more a government’s payments to eliminate that debt will rise. That in turn reduces government services, which erodes governments’ core duty – to be government of the people, by the people, and for the people.

A. Two Main Types of Government Payments to Pension Funds

In general there are two types of payments governments make to their Pension Funds.

1. Normal Yearly Contributions:

In general governments and their employees pay what’s called the “Normal Annual Cost Contribution” to their Pension Fund each year. This is the amount the Pension Fund Actuary calculates is necessary so that if all their assumptions and projections for the next 60 years or so come true there will be enough money in the Pension Fund in the future to pay the part of future pension payments that will be earned that year. It’s extremely important to realize that Pension Fund Actuarial Valuations are in fact pension funding plans based on the fundamental assumption that the only money that should ever have to be paid to a Pension Fund is the annual Normal Contribution.

2. Unfunded Pension Amortization Payments:

If a significant Pension Fund deficit develops usually only the government must make additional payments to eliminate that deficit. Further, if such a significant deficit develops it is proof on its face that the government’s basic pension funding plan has failed.

3. Other Payments:

a) Pension Obligation Bonds are Unfunded Pension Debt Payments:

Payments on Pension Bonds must be added to a government’s payments to Pension Funds to understand the total impact on current and projected government spending of their pension benefits.

b) Other Types of Payments:

Governments can make several other types of payments to Pension Funds such as reimbursement of administrative expenses and payments towards other types of benefits. But we won’t consider those payments in this paper.

B. Why Moody’s Should Restate Payments by Local Governments – Not Just State Payments

Moody’s states “current disclosures allow us to propose making the adjustment only for states at this time.” Moody’s will do what it wants to do – but I disagree for 3 main reasons.

1. Many Local Governments are Larger than Many States:

Los Angeles County’s Total Pension Liability is $50 billion – larger than 30 states in the US. Many US local governments have larger pension liabilities than many states.

2. Availability of Data:

The data Moody’s needs are readily available for many local governments. Users of Moody’s ratings will be better served by receiving more complete creditworthiness information about these entities than to have it not provided because the necessary data is unavailable for some other local governments.

3. The Threat to Owners of California Local Government Pension Obligation Bonds:

The purchasers and insurers of California local government Pension Obligation Bonds are facing an extreme threat. The cities of San Bernardino and Stockton have filed for federal bankruptcy and propose to not pay most of the remaining balance of their Pension Bonds. If they sustain this threat other local governments are very likely to follow.

The municipal finance industry led investors in these bonds into an extremely dangerous political situation. The people of California have repeatedly voted to require their approval before governments enter into the levels of debt represented by Pension Bonds. But the municipal finance industry developed methods of issuing Pension Bonds to avoid the people’s expressed demand to require a vote [11]. In this situation “moral hazard” properly refers to the industry’s purposeful design of methods to avoid what they knew was the express will of the people to control their governments’ debt. The industry placed the purchasers of these Bonds at far greater political risk than was disclosed. Even many conservatives, when they find out about how the industry purposefully side-stepped their repeated demand to submit debt of this kind to a vote, feel less morally obligated to pay these bonds.

Moody’s did not reflect this significant political risk in its credit ratings when these bonds were originally issued. They should do so now and in the future. Moody’s modifications of what government payments to their Pension Funds should be provide important additional information to properly evaluate investors’ credit risks.

C.  How Actuaries Calculate Payments to Pension Funds

This is a general description of how Actuaries calculate the two main payments governments make to Pension Funds.

a)  Payment #1, Calculating the Normal Yearly Contribution

Actuaries calculate the “Normal Yearly Contribution” (aka “Normal Cost”) to Pension Funds exactly as they do “Total Pension Liability” as described above with only one difference.

In calculating Total Pension Liability – called “Actuarially Accrued (Pension) Liability” – they first project the part of future pension payments employees and retirees have already earned in the past. Second, they “discount” that by the Pension Fund’s assumed target rate of return to come up with the amount that should be in the Pension Fund today.

When calculating the Normal Yearly Contribution instead of projecting the part of future pension payments that have already been earned the Actuary projects the amount that will be earned by employees in the upcoming year. That is “discounted” by the assumed rate of return to project how much money must be contributed to the Pension Fund next year so that pensions being earned next year can be paid. This is expressed as a percentage of each payroll for each class of employee rather than as a dollar amount, although a dollar estimation is included in Valuations. Then the Actuary allocates part of the Normal Cost to the government’s employees and the rest is allocated to the government. Many governments pay a portion of the “Employee Share” – but that is rarely reported. It should be – but it isn’t. This shows the allocation of the Normal Contribution between governments and their employees for the six Bay Area – North Coast counties with their own County Pension Funds as of the most recently available Actuarial Valuations.

Dickenson-Moody-table-7

Except for Contra Costa the portion of employee contributions actually paid by these counties is not shown in the Actuarial Valuations for these County Pension Funds.

b)  Payment #2, Calculating the Unfunded Pension (UAAL) Amortization Payments

The number of years a government plans to take to eliminate an Unfunded Actuarially Accrued (Pension) Liability (UAAL) is called the “amortization period”. California county governments may take as long as 30 years to pay off a UAAL. The interest expense incurred is the same as the Pension Fund’s target rate of return. There are two common methods used to calculate the amount of these UAAL Amortization Payments:

    • Level Dollar – the same dollar amount is paid each pay period.
    • Level Percent of Payroll – the same percent of each payroll is paid each pay period.

The “Level Dollar” method is basically the same as the traditional 30 year fixed interest rate & payment home mortgage except the number of years can be shorter. You borrow money to buy a house and pay it back by making equal payments every month for 30 years. Most of the early payments are interest but the last payments mostly reduce debt.

Under the “Level Percent of Payroll” method the Actuary first estimates how much the government’s total payroll will
be in each year of the amortization period assuming it will grow the same percent each year. Four percent yearly growth
is often assumed. Then the Actuary calculates a fixed percentage of each year’s payroll the government needs to pay to
eliminate the unfunded pension debt by the end of the amortization period. That fixed – or “level” percent of each
future year’s projected payroll is projected to be each future year’s payment.

Let’s compare the results of these two methods. Assume:

    • UAAL = $100 million
    • Amortization period = 30 years
    • Interest rate = average target rate of return for the 6 Bay Area – North Coast County Pension Funds (7.73%)
    • One payment made at the beginning of each year [12]
    • Every assumption and projection over the next 30 years works out perfectly

(1) Level Dollar Amortization:

Payments are $8,882,743 a year for 30 years. This method is simple to calculate and understand.

(2) Level Percent of Payroll Amortization:

Assume Payroll is $75 million and will grow 4% a year in each of the next 30 years. If the government pays exactly 7.62146% of each of these payrolls through the next 30 years the UAAL will be eliminated. Payments start at $5.7 million and grow to $17.8 million 30 years from now. Both payroll and payments grow at a rate of 4% a year.

(3) Comparing Amortization of Unfunded Pensions Using “Level Dollar” vs. “Level Percent of Payroll”:

Those who advocate the “Level Percent of Payroll” method provide two arguments for why this is a good idea. First, although annual payments will grow as total payroll grows, at the same time the government will supposedly be earning more income because of increased taxes, fees and grants. So – this method is thought to even out the financial burden over time. Second, the regular yearly pension contributions are figured as a percentage of regular payroll each pay period. Since this “Level Percent” method is used for regular yearly payments, it’s easy to tack on a set percentage for unfunded pensions. But the Level Percent approach has problems.

a) Percent of Payroll – Much Lower Payments Early – Much Higher Later

Payments that are much less for the first third (or so) of the amortization period using the Level Percent method than it would under the Level Dollar method. As shown in Figure 2, for the first several years, the payments using the Level Percent of Payroll method are much lower than the payments required under the Level Dollar Method. The government officials who make the decision are in office now. It is quite likely they decide to use the Level Percent of Payroll method to make things a whole lot easier for themselves. The way they make it easier on themselves is to shove much larger payments off into the future when they won’t be in office. Someone else will have to deal with the problem.

Dickenson-Moody-figure-2

(b) Percent of Payroll – Negative Amortization Increases Debt

The next  graph, Figure 3, shows the annual payments under the Level Percent of Payroll method and the annual interest expense. For the first 12 years the payments are less than the annual interest expense. This is called “negative amortization” – the unpaid interest actually increases the debt. The pink area is unpaid annual interest – the debt is actually increasing.

Dickenson-Moody-figure-3

The bottom graph in Figure 3 shows the balance of unfunded pensions over the entire 30 year amortization period. The payments for the first 12 years are less than the annual interest expense and so the balance of unfunded pension debt increases for the first 12 years. At that point if we assume payroll really did grow 4% a year and therefore payments also increased 4% a year, finally the payments “catch up” with interest expense. But over those 12 years unfunded pensions actually increased nearly $16.5 million making total debt $116.5 million in year 12 instead of the beginning debt of $100 million. It then takes another 8 years for the increasing payments to pay the accumulated unpaid interest balance over the first 12 years. Therefore the balance of unfunded pensions doesn’t get back to its original $100 million until 20 years in the future.

The entire original balance of $100 million of unfunded pensions will be paid from 21 to 30 years from now. In this realistic example
today’s government officials shoved today’s unfunded pensions off to an entirely new generation of citizens and officials.

Both methods pay off the debt over 30 years. But the Level Percent of Payroll causes nearly $65 million more in interest expense – about 40% more (again – assuming all other assumptions and projections come true).

D.  How Moody’s Adjustments Will Affect Required Government Pension Fund Payments

These are county payments to Pension Funds projected in their Actuarial Valuations. We’ll see how Moody’s
adjustments would change these payments. (These don’t reflect County payments of part of employee contributions.)

Dickenson-Moody-table-8

Moody’s delves into their proposed adjustments to these two government payments to Pension Funds [13].

Ideally, participating government employers make annual contributions to their pension plans that result in those plans becoming fully funded over a reasonable time horizon. We propose to adjust annual contributions to reflect the adjustments we have made to pension liabilities. We believe this adjustment would function as a more accurate indicator of fiscal burden. We would not intend it to be a prescriptive funding strategy…

…We will adjust the ENC (Employer Normal Cost) to reflect our common discount rate, and the amortization payment to reflect our adjusted unfunded liability, a common amortization period, and a level-dollar funding approach.”

Moody’s Adjustment #1. Normal Annual Cost Contribution Payments Calculated at Much Lower Projected Rate of Return

In adjusting the normal yearly government contribution Moody’s would project the normal cost forward for 17 years at the plan’s reported discount rate, and then discount it back at 5.5%, after which employee contributions are deducted to determine the adjusted government yearly contribution. Using this approach, a reported normal cost payment of $100 million based on an 8% discount rate would grow to $149 million based on a 5.5% discount rate. This adjustment is quite simple mathematically. This is only an approximation of what the normal contribution would be if Moody’s were to use the Actuary’s projections of the part of future pension payments estimated to be earned in the upcoming year. That data isn’t available to Moody’s and even if it were it would be a lot of work to process. This shows the effect of this proposed Moody’s adjustment on the six Bay Area – North Coast counties:

Dickenson-Moody-table-9This adjustment increases the amount of the normal annual contribution Moody’s believes these counties should be paying from 66% to about 90% – not quite double. This is a very significant increase – but not nearly as significant for many governments as the next payment adjustment.

Moody’s Adjustment #2. Return Pensions to “Fully Funded Over a Reasonable Time Horizon”

a) Furthers the Goal of Intergenerational (or Inter-period) Equity

Who should pay for government employee pensions – people who receive their services, or the next generation? Employees earn their pensions while they are still working for government – it’s part of their compensation. Retirees never earn their pensions when they receive them. The payment of pensions in the future is the payment of a debt. Assuming they meet all the requirements to receive a pension, the day someone leaves a government’s employment they have 100% earned all those future pension payments.

But there’s a huge problem. We don’t know how much a retiree’s total pension payments will be when they retire. We can’t be sure how much the true economic expense of that employee’s pension was when he or she earned those future pension payments, nor do we know how much should be in the Pension Fund when they retire.

The Governmental Accounting Standards Board (GASB) establishes “Generally Accepted Accounting Principles” (GAAP) for state and federal governments. GASB describes “Inter-generational Equity” this way – “the current generation of citizens should not be able to shift the burden of paying for current-year services to future-year taxpayers. … financial reporting should help users assess whether current-year revenues are sufficient to pay for the services provided that year and whether future taxpayers will be required to assume burdens for services previously provided” [14]. Moody’s proposed adjustments are necessary in large part because current GASB accounting standards have not fulfilled this principle. GASB has allowed governments to report the pension expenses that created today’s massive unfunded pension debt decades after employees earned those pensions. The public didn’t see that “future taxpayers will be required to assume burdens for services previously provided” [15].

b) Moody’s To Use 17 Years as “Reasonable Time Horizon”

Moody’s believes a “reasonable time frame for government payments to Pension Funds” is the remaining number of years current employees will continue to work for a government. Moody’s will assume a standard 17 year remain service life for all governments. At that point the Pension Fund should be fully funded. The failure of governments to achieve this goal of “prudent” financial management imposed hundreds of billions of unfunded pension debt on future generations. They won’t receive one minute of public services or one dime of public infrastructure for those massive payments. If government officials had been required to fully fund pensions by the time employees retired this unfair unfunded pension debt would not have been shoved onto future generations.

Moody’s Adjustment #3. Unfunded Pension Liability Amortization Payments to Use Level Dollar Method

Moody’s proposes to require the Level Dollar method to calculate the catch-up payments necessary to restore pension funds to 100% funding within 17 years, not Level Percent of Payroll method. This adjustment clearly exposes the deeply flawed financial management of many if not most public pension systems.

The Level Percent of Payroll method actually increases a government’s unfunded pension debt 15% to 25% over the first 1/3 of the amortization period. As discussed earlier, it is likely the main reason government officials choose to use this method is to make it easier on themselves while they are in office at the cost of shoving even greater debt onto the next generation. It’s beyond the scope of this analysis to determine the prevalence of this method nationally. However, the Actuarial Valuations of the six counties in the San Francisco Bay Area and California North Coast region clearly indicate all six use this method. This table shows each county’s assume yearly increase in payroll and therefore UAAL amortization payments and the number of years taken to amortize the UAAL.

Dickenson-Moody-table-10
Unfunded pension debt is actually increasing for five of these six counties as a result of this method. It’s decreasing for one – San Mateo – because that county chose to eliminate unfunded pensions over an unusually short number of years – even less than Moody’s standard 17-year amortization period. They appear to be beyond the point at which payments begin to be larger than interest expense. Therefore they are beyond the period of “negative amortization” inherent in Level Percent of Payroll amortization.

This sample of six counties is far too small to generalize about the broad behavior of state and local governments. But I think it’s safe to say that the temptations of the Level Percent of Payroll amortization method are so great that many and probably most governments use that method – thereby actually choosing to increase their unfunded pension debt. Very few citizens realize their elected officials responsible to manage their local government finances have actually chosen methods to amortize unfunded pension debt that in fact significantly increases that debt. My experience is that when citizens realize this fact – regardless of their political ideology – they find it appalling.

4.  How Much Will Moody’s Adjustments Increase Catch-Up Payments to Restore Pensions to 100% Funding?

a) Recalculate the Amount of Underfunding Using the 5.5% Rate of Return:

As shown already on Table 4, by lowering the discount rate, or annual rate of return, on the future pension obligations already earned, to Moody’s much more conservative 5.5%,  the unfunded pension liability for the six counties increases from $4.1 billion to $10.2 billion. This change is summarized per county in Table 11 below:

Dickenson-Moody-table-11

Moody’s adjusted Net Pension Liability averages about 2.5 times larger than the UAAL reported by these six counties. By itself this will significantly increase the amount Moody’s considers a “prudent” unfunded pensions payment.

b) Calculate Catch-Up Payments Using New 17 Year Amortization Schedule and Apply Level Dollar Payment Method:

Table 12 shows the astonishing impact of Moody’s adjustments on the payment calculations. If Moody’s is correct these six counties – on average – should be paying nearly three times as much as they are to eliminate their Net Unfunded Pension Liability. (Again – the total percentage increase is the average increase for the counties – not based on total dollars.)

Dickenson-Moody-table-125. Summary: Impact of Moody Adjustments on Normal and Catch-Up Pension Fund Payments:

a) Government Payments to Pension Funds:

The graph below shows the proportional change in County payments to Pension Funds. The first stack for each county is the amount projected in the Actuarial Valuations. They equal 1.0 (or 100%). The second is the Moody’s adjusted payment. Green is the Normal Cost and pink-red is Unfunded Pension payments.

Other than San Mateo payments more than double using Moody’s adjustments. Sonoma is pushing 3 times greater and Mendocino is 2.5 times greater.

Dickenson-Moody-figure-4

The table below shows the actuarial calculation and Moody’s proposed adjustment. This table summarizes the changes in these six counties’ payments to their Pension Funds described above. It shows payments projected in Actuarial Valuations compared to those produced by Moody’s adjustments. Again – the total percent change is the average percentage for the 6 counties – it isn’t “dollar-weighted” (not based on the total dollars).

Dickenson-Moody-table-13

All together these six counties were projected to pay about $312 million to their County Pension Funds as their Normal Contribution and a little more than that – about $328 million – as UAAL Amortization Payments (catch-up payments to restore 100% funding). Total payments were projected to be about $640 million.

Moody’s adjustments more than doubled those combined payments for all six counties to over $1.4 billion – 125% more than currently defined by the Actuaries hired by these counties. The adjustments increased total annual Normal Contributions 78%, but increased payments to eliminate the Net Pension Liability 172%. San Mateo had the lowest increase. Even so Moody’s adjustments suggest it should be paying nearly double (an 87% increase overall). San Mateo’s comparatively short UAAL Amortization Period of 15 years for each year’s UAAL produced much higher Actuarially-defined UAAL payments than the methods used by the other counties. Moody’s adjustments nearly tripled Sonoma County’s payments – an increase of 182%.

b) Government Payments to Eliminate Additional Unfunded Pension POB Debt:

Pension Obligation Bonds (POB) are simply unfunded pension debt restructured into bonded debt in the hopes of obtaining a lower interest rate. Pension Bond payments must be added to Net Pension Liability amortization payments to see the total yearly “cost” to eliminate unfunded pensions. Once again Moody’s adjustments double these payments.

Dickenson-Moody-table-14

c) Overall Impact on County Budgets:

Most of the money spent by local governments comes from the State and Federal governments. Local governments must “match” a portion of Federal and State funds out of their independent local revenue base which includes property and sales taxes. The more they divert their independent tax base to pay Pension Funds and Pension Bonds the less they have to match and the less they have to spend on locally-funded projects. The more these payments consume their local tax base the more local governments lose control of their budgets.

These are, first, the 6 counties’ property tax income [16] compared to current payments to Pension Funds and Pension Bonds, and second, to the modified payments restated by Moody’s. Total Percentages are averages of the percentages for each county – they aren’t dollar weighted based on Total Dollars.

Dickenson-Moody-table-15

These six counties retained nearly $1.6 billion of the property taxes paid in those counties as their own revenue. Their total actuarially defined payments to their County Pension Funds were nearly $640 million or 37% of this County property tax revenue. Payments on Pension Bonds were about $177 million which was 12% of total property tax income. Therefore total payments projected by Actuaries to County Pension Funds and Pension Bond payments were projected to be slightly less than $820 million, half of their property tax income.

Moody’s adjustments make a huge difference. Total adjusted payments to Pension Funds jumped from 37% of County property tax income to 86% – well more than double. Then when Pension Bond payments are added in all these payments combined they consume almost all these counties’ property tax income.

If Moody’s is correct – that these are the payments necessary to prevent imposing significant unfair debt on the next generation and to prevent further deterioration of these counties’ finances – then these counties should have already lost control over practically all their property tax income. The only reason they still retain control over some of their property tax income is they are not managing their pension benefit finances prudently – they are in effect choosing to force the next generation to pay huge unfunded pension debts.

*   *   *

V.  CONCLUSION – WHAT THIS SHOULD MEAN TO CONCERNED CITIZENS

Moody’s believes current government financial reporting badly understates the financial risk created by government unfunded pension debt. Further – they believe today’s financial reporting by governments and their Pension Funds allow such a wide range of assumptions and accounting treatments that it’s practically impossible to compare governments meaningfully in terms of the finances of their pension benefits.

Although the focus on this paper is on Moody’s proposed adjustments, Moody’s conclusions are highly consistent with those reached by the Governmental Accounting Standards Board when they issued radically new pension financial reporting requirements this past summer (June 2012). Both will greatly increase reported unfunded pension debt. And the conclusion that most governments should be paying considerably more to their Pension Funds will be inescapable.

Moody’s believes the rate of investment profits assumed by practically all local and state government Pension Funds is significantly too optimistic. They believe return assumptions should be more consistent with those used for private sector Pension Funds. Moody’s believes the “true economic” state and local government unfunded pension debt is more like three times greater than the values those governments report today.

The six California counties analyzed for this report would report Net Pension Liabilities nearly three times more than their Pension Fund Valuations report today.

Given Moody’s adjustments and including the balances of Pension Bonds just these six counties together owe a total of over $1.6 billion in unfunded pension-created debt. They’ve only achieved about half their self-proclaimed pension funding requirements. By extension the debt for all local governments in California is massive.

Moody’s adjustments indicate they believe that if the finances of these six counties’ pension benefits were being prudently managed they would be paying over $1.4 billion each year to their Pension Funds instead of the less than half that amount they are in fact paying. When payments of Pension Obligation Bonds are included – even given the lower payments defined by actuaries today these counties are paying half their property tax income to their Pension Funds and to holders of their Pension Bonds. But if Moody’s adjustments are correct they should be paying all their property tax income. Again, if Moody’s is correct, the only reason these counties are retaining any of their property tax income for other purposes is they are pushing hundreds of billions of unfair debt onto the backs of the next generation.

Today’s financial management of government pension benefits is deeply flawed and in terrible need of major reform.

*   *   *

Download Print Version

VI.  DATA SOURCES

Audited financial statements as of 6/30/2011 for these six counties were downloaded and analyzed. They have been aggregated and are all available for downloading on the www.yourpublicmoney.com website. Here are the direct links:

Alameda County’s 6-30-2011 Audited Financial Statements

Contra Costa County’s 6-30-2011 Audited Financial Statements

Marin County’s 6-30-2011 Audited Financial Statements

Mendocino County’s 6-30-2011 Audited Financial Statements

San Mateo County’s 6-30-2011 Audited Financial Statements

Sonoma County’s 6-30-2011 Audited Financial Statements

In addition the most recent available Actuarial Valuations were downloaded in early fall, 2012. They are for these counties “County Employees Retirement Associations” – as in “Alameda County Employees Retirement Association”.

Dickenson-Moody-data-sources

Here are the direct links to download these actuarial valuations for all six counties:

Alameda County Actuarial Valuation

Contra Costa County Actuarial Valuation

Marin County Actuarial Valuation

Mendocino County Actuarial Valuation

San Mateo County Actuarial Valuation

Sonoma County Actuarial Valuation

There are three kinds of Pension Funds that offer “guaranteed pension benefits” to state and local government employees”

Single Employer Funds – The Pension Fund is for only one government employer. Therefore all the financial balances and activity in the Pension Fund impact that one government.

Then there are two types of “Multi-Employer” Pension Funds in which the Pension Fund provides pension benefits to retirees of more than one governments.

Agent Funds – the Pension Fund maintains its books so that the obligations and balances of each employer government can be specifically identified. Therefore one government may have a lower relative level of unfunded pension obligation than another.

Cost-Sharing Funds – Balances are not maintained for individual governments. As a result unfunded pension obligations are shared among all participating governments even though some governments may have paid a higher portion of its obligations than others. Balances are allocated to individual governments based on the portion of that government’s payments to the Pension Fund relative to all other governments’ payments.

FOOTNOTES

[1] Big Three (Credit Rating Agencies), Wikipedia, http://en.wikipedia.org/wiki/Big_Three_(credit_rating_agencies), (downloaded 12/2/12)

[2] Moody’s paper and other information about Moody’s proposed adjustments are available at my website www.YourPublicMoney.com in the Data/Reports/Video section of the site.

[3] Adjustments to US State and Local Government Reported Pension Data – Request for Comment, Moody’s Investors Service, July 2, 2012, page 2

[4] Moody’s, ibid, page 1

[5] Moody’s, ibid, page 2

[6] For a “plain-language” description of pension math that explains many of the terms in this paper (such as “smoothing”, “corridor limit”, “UAAL”, “Pension Obligation Bonds”, etc.) see How Pension Funds Work (click to access) in the “Data/Reports/Video” section of my website.

[7] This is the most important concept people concerned about unfunded pension debt need to understand to know how the huge unfunded pension debt that grips state and local governments was created and what needs to change – regardless of what changes you wind up wanting to see. See How Pension Funds Work.

[8] Moody’s, ibid, pages 5 – 6

[9] There’s an interesting “twist” to the math of Moody’s adjustments. The percentage change in asset values shown in Table 1 – Moody’s Adjustment of Pension Fund Asset Value – Six CA Counties ($Millions) at first glance seems to be less than the percentage change in Table 3 – Moody’s Adjustment of Total Pension Liability – Six CA Counties ($Millions). But the range of the percent change in asset values is much greater than that for Total Pension Liability. It’s this greater range of change in asset values that drives the even greater spread in the percentage change in Net Pension Liability.

[10] For an outstanding explanation of the absurdity of this “myth” see Pension Puffery: Here are 12 half-truths that deserve to be debunked in 2012, Governing Magazine online, Girard Miller, 1/5/12

[11] See Appendix B – “California” in An Introduction to Pension Obligation Bonds and Other Post-Employment Benefits – Third Edition (click to access), Roger L. Davis, Orrick Herrington & Sutcliffe, LLP, 2006. Also see “Pension Obligations Bonds” on page 21 of How Pension Funds Work.

[12] In an email exchange the author of Moody’s paper indicated Moody’s believes the “correct” method of calculating interest expense is “beginning of period” rather than “end of period”. Also they are simplifying the amortization schedules by assuming annual payments. I and several other financial professionals don’t understand the idea that “one payment per year” wouldn’t incur interest expense in the first year. It would have to be made on the first day of the first year in order to have no interest expense – it would all go to reduce debt principal. Payments are generally made to Pension Funds when governments make payrolls or shortly thereafter. Government paydays are sometimes once a month, or twice a month, and sometimes every two weeks or 26 times a year. In any of these “real world” systems there would be at most only one payment for anywhere from a two-week to one-month period that doesn’t accrue interest expense and then only if that payday occurs on the first day of the first fiscal year. However – since this is Moody’s assumption the analysis in this paper uses that assumption – even though we don’t think it’s realistic.

[13] Moody’s, ibid, page 8

[14] Concepts Statement No. 1: Objectives of Financial Reporting, Governmental Account Standards Board, No. 037, May 1987, page i

[15] GASB released new pension financial reporting standards in June 2012 that must be implemented by governments no later than their financial statements for fiscal years that begin after 6/15/14. These new standards will generally correct this glaring error in their current standards. I prepared several papers analyzing these new GASB standards. They are available at www.YourPublicMoney.com in the “Data/Reports/Videos” section.

[16] California counties collect all the property taxes paid within a county and then disburse significant amounts of the proceeds to other property tax supported agencies such as cities, school districts, and so on. The counties retain a set portion of the property taxes for themselves. The property tax values shown in this table are the amounts retained by the counties as their own revenue – not the total paid by property owners in the county.

*   *   *