Funding Public Pensions - Mississippi Workshop/2017...Funding Public Pensions •...

POLICY BRIEF

Funding Public PensionsIs full pension funding a misguided goal?BY TOM SGOUROS

http://haasinstitute.berkeley.edu

PUBL

ISHE

D FE

BRUA

RY 2

017

PUBLISHED BY:

The Haas Institute for a Fair and Inclusive Society at UC Berkeley brings together researchers, community stakeholders, policymakers, and communicators to identify and challenge the barriers to an inclusive, just, and sustainable society and create transformative change. The Haas Institute advances research and policy related to marginalized people while essentially touching all who benefit from a truly diverse, fair, and inclusive society.

CONTACT460 Stephens Hall Berkeley, CA 94720-2330 510-642-3326 haasinstitute.berkeley.edu

@haasinstitut e

This project was developed by the Just Public Finance program of the Haas Institute for a Fair and Inclusive Society.

AUTHORTom Sgouros has worked as a policy consultant and data scientist for 30 years, providing policy advice to a wide range of elected officials and candidates on public finance, budgeting, and tax policy. He formerly served as the Senior Policy Advisor to the Rhode Island General Treasurer, he currently works at Brown University’s Center for Computation and Visualization.

GRAPHICS & DATA VISUALIZATIONSamir Gambhir

EDITINGWendy Ake Stephen Menendian

COPYEDITINGEbonye Gussine Wilkins

DESIGN/LAYOUTEbonye Gussine Wilkins

ACKNOWLEDGMENTS This report has benefitted from comment and review of experts Cathy O’Neil and Jonathan Reiss. Their time and insights are greatly appreciated.

REPORT CITATIONSgouros, Tom. “Funding Public Pensions: Is full pension funding a misguided goal?” Berkeley, CA: Haas Institute for a Fair and Inclusive Society, University of California, Berkeley, 2017.

The full report can be accessed at haasinstitute.berkeley.edu/ justpublicfinance

Cover Photo: “Money Plant” photo from TaxCredits.net, via Flickr, Creative Commons Attribution 2.0

SPECIAL THANKS Thank you to the Ford Foundation for supporting this research.

TABLE OF CONTENTSAbstract .............................................................................................................4

The problem: underfunded pension systems....................................................5

Why are there accounting rules? .....................................................................7

Problems with accounting rules .......................................................................9

Legal: Governments will not be liquidated ............................................................. 10

Chronological: Present value masks the level of urgency ...................................... 11

Actuarial: Full funding is not required to pay all pension debts ............................ 12

Mathematical: Financial reports deserve more precision ....................................... 14

Financial: Rate of return is for the longest term. ................................................... 14

Economical: Comparisons to pension debt should be chosen properly .................. 16

Political: Overfunding is a risk, too ........................................................................ 16

Philosophical: A pension plan is a mutual, not individual, arrangement ............... 18

Other post-employment benefits crisis ..........................................................19

Common “solutions” that seldom solve anything ..........................................20

Pension obligation bonds: Investing on the margin ............................................... 20

Closing a plan: Riskier than you might think ......................................................... 21

Defined contribution pension plans: Not really a pension, or a solution ................. 22

Potential solutions: different accounting for different questions ........................... 24

Arguments against ..........................................................................................25

Facing facts: On not kicking the can down the road ............................................. 25

Budgeting: Full funding reduces volatility ............................................................ 26

Accounting clarity: What questions can you answer? ........................................... 27

Conclusion.......................................................................................................28

Endnotes..........................................................................................................30

Appendix ..........................................................................................................35

Funding Public Pensions • haasinstitute.berkeley.edu • 4

ABSTRACTPublic pension systems across the United States are, and have been, in crisis. But, to a larger extent than is widely acknowledged, the crisis is the result of the accounting rules governing both these plans and the governments that sponsor them. These rules are designed to insure against risks that public pensions systems do not face, while simultaneously failing to insure against the risks they do face. The rules also encourage “reforms” that frequently do not improve the finan-cial situation of a given pension system. This is not just deplorable, but a recipe for making a bad situation worse—precisely what we’ve seen over the past few decades. A hybrid accounting system could provide a more accurate picture of a system’s financial health while reducing the waste of overfunding. It could relieve unnecessary financial pressures on thousands of governments across the nation while still preserving the integrity of their pension systems.


THE PROBLEM: UNDERFUNDED PENSION SYSTEMSAcross the nation, public pensions are in crisis, and have been so for a long time. Funding pen-sion costs is a political issue in cities, counties, and states from California, to Illinois, to Rhode Island. The rising expense of public employee pensions has become a political hot button justify-ing cuts to education and other necessary government investments, causing acrimonious debate, court cases, protest marches, and more. All the recent incidents of municipal bankruptcies have been blamed, at least in part, on pension obligations. Most famously, this was the case in Detroit, Michigan, but has also been true in the cities of Stockton and Vallejo in California, Prichard, Ala-bama, and Central Falls, Rhode Island.

The city of Chicago is currently feeling some of the warning tremors. According to its own estimates, the city’s various pension funds have only half the funds in hand needed to pay its pensions. This leaves a $28.6 billion difference between the assets and the present value of the debt to all the current and future retirees in the system.1 This difference, known as the “unfunded liability,” was cited as the primary reason that Moody’s, the bond-rating firm, downgraded Chica-go’s bond rating to “junk” status in May of 2015.2

The other common measure of a pension system’s health is the ratio between the assets and the future liabilities, known as the “funding ratio.” Chicago’s funding ratio hovers around 50 percent, but the condition of the pension funds managed by the state of Illinois is even worse, showing a 39 percent funding ratio, with $111 billion worth of unfunded liability.3

Funding ratio calculation for CalPERS, 2014. “Total assets” is the value of the pension fund today, and “Estimated total liability” is the estimate of the future liability of the current employees whom are owed a pension. This is an estimate over several decades, so there are a lot of assumptions built in, and a great deal of uncertainty.

These are just the cases that make the headlines. In thousands of other governments across the country, pension contribution increases are a constant source of fiscal stress, resulting in cuts to schools, infrastructure, and increases in taxation. Despite the stress of added payments, the prob-lem is not going away. America’s public pension systems are, on average, only 74 percent funded as of 2014, with only $3.6 trillion in assets on hand to pay $4.8 trillion in liabilities, an unfunded liability of $1.2 trillion.4 These governments have only a fraction of the assets on hand to make all the pension payments they have promised to their members. Retirement benefits, state and municipal budgets, and taxpayers are jeopardized. It is a crisis all around.

And yet, is it really true? A close look at the Detroit bankruptcy shows that it really had far more to do with the politics of Michigan’s suburbs and the Governor Rick Snyder’s feelings about the city than it did with the mathematical reality of the city finances.5 The narrative of runaway pen-sion obligations sinking an ailing city’s finances is simply not supported by the facts, which had much more to do with a sudden loss of state support and ill-advised interest-rate swaps.6 Long-

Total Assets $301 billion$394 billion

Funding Ratio 76.3%Estimated Total Liability

= =

CalPERS Annual Report 2015, https://www.calpers.ca.gov/docs/forms-publications/cafr-2015.pdf

PENSION FUNDING RATIOCalPERS, June 2014Pension Funding Ratio CalPERS 2014


term debt due decades in the future cannot cause insolvency today even if it is a sign of trouble to come. Insolvency is the result of being unable to pay current obligations; long-term debt is just a threat.

Detroit’s long-term debt of $18 billion was the headline number for the bankruptcy proceedings, but the pension system accounted for less than 20 percent of that, and that was only using very conservative assumptions about the discount rate and demographics. The actual cash-flow issue that triggered the bankruptcy was a $198 million shortfall in fiscal year 2014, a number easily explained by a $194 million decline in revenue—the largest component of which was changes in state policy that cut revenue sharing by $56.5 million—and $547 million in termination fees from swaps deals. The cost of running the city’s pension systems had actually declined in the previous two years. Detroit’s pension contribution in 2013 was $78.3 million, a slight drop from the 2012 contribution of $86.1 million, though still above the $65 million average payment of the previous five years.7 Obviously the city’s willingness to enter into the swaps was a symptom of financial pressure, and it is certainly true that the pension system was among the sources of that pressure. But as we will see, the pressure was applied by the pension accounting rules in place, much more than the mathematical reality of the payments to be made.

Cases of other cities used to illustrate the pension crisis provide equally misleading stories con-tradicted by a closer look. Stockton, California, seems to have been sunk not by pension costs, but by the foreclosure crisis, by some expensive city investments like a sports arena and hotel that did not pay off, and by an ill-advised gamble.8 The purpose of the gamble was to reduce pension liability, but it was this gamble that went wrong, not the pensions.9

Central Falls saw its crisis precipitated by devastating cuts in state aid in 2009 and 2010, and a balloon payment worth 40 percent of the city’s annual budget due in 2010 from a 1990 bond.10 Prichard, a poor town near Mobile, Alabama, could hardly have been sunk by its pension costs, since they had not been paid in years. During its first round of bankruptcy in 1999, Prichard offi-cials “admitted that it had not made payments into its employees’ pension fund for years and had withheld taxes from employees’ pay checks, but had not submitted the withholdings to the state and federal governments.”11

Obviously, these cities were all stressed fiscally, but how and why did it come to be that public employee pensions were argued to be primary causes, when this was not the case? Part of the reason is that the pension funds in these cities were known to be underfunded by the accounting standards used to evaluate them, and those standards use normative language and measures to describe the situation.

A pension plan is “underfunded” and the government deemed not “fiscally sound” if it does not have the assets currently on hand to pay all of the future liabilities, clearly implying big problems ahead.12 The unfunded liability provides a convenient measure of the degree of the problem, and since the liabilities of any pension system are typically large compared to the size of the budget, the unfunded part of that liability often seems immense. For example, the Illinois annual budget is in the $55–60 billion range, an uncomfortable comparison with its pension funds’ liability of al-most twice as much. Nationally, pension liabilities are in the trillions, even if the precise number of trillions is heavily contingent on analyst assumptions.13

Unfortunately, the widely-used measures of pension assets and liabilities are more complicated, less complete, and less reliable than they are typically presented to be. Where the measurements are accurate, they are commonly misinterpreted. They ignore important sources of system strength and create perverse incentives to system managers. They serve not only to exaggerate the problems facing pension funds, but also provide a poor guide to addressing those problems. Certainly the current funding situation of most pension plans could be improved and certainly there exist pension systems that really are in danger of collapse. However, might there be need-less damage done by constantly predicting impending collapse for so many others?


WHY ARE THERE ACCOUNTING RULES?It is worth reviewing the recent history of the accounting rules themselves. This will help to understand how the accounting rules and the choices that created those rules affect the pension debates.

The accounting rules for public pension systems are established by the Governmental Account-ing Standards Board (GASB). This is the private body of accountants that defines the Generally Accepted Accounting Principles (GAAP) in use by local and state governments in the United States.14 (See figure on page 8.)

Old-age pensions were among the reforms advocated by the socialists, communists, and the Pro-gressive movement in the United States during the 19th century. In the latter years of the century, private employers around the country adopted the idea, some prompted by labor unrest, others prompted by the desire to avoid it.

Government pensions have a longer history. Pensions for disabled and retired military personnel were common decades earlier. George Washington prevented a mutiny in the Continental Army by intervening personally in a dispute over pensions during the Revolution. The granting of pen-sions to civilian government employees, however, marched in step with similar advances among private industry, starting with New York City’s establishment of a disability pension system for its police in 1857 and advancing through several of the nation’s large cities. By 1917, 85 percent of the nation’s cities with populations over 400,000 had some form of police pension. 15

The states were a bit slower than the big cities. Massachusetts was the first state to establish a pension system, in 1911, and by 1929, there were only 1,003 retired individuals receiving pension benefits from only five states. The pace quickened substantially with the onset of the Great De-pression, and plans were established quickly enough that by 1935, when Social Security was in-troduced, there were over 400,000 pensioners across the country receiving benefits in 32 states.16

By the end of the 1930s, many government employers across the country were offering pensions to their employees. Many offered them on a pay-as-you-go basis, with payments drawn from tax revenue, without an associated pension fund, or with a relatively small fund whose purpose was only to manage cash flow. Some of these were funded by special dedications of tax revenue, such as fines, permit fees, or, in at least one city, dancing school licenses.17 Over time, out of fear that these pension commitments would balloon in decades to come, many plans moved to an actuarial system of funding pensions, with a pension fund whose income would pay much of the pensions, and by the 1980s, this transition had been made for most plans.18

Because of the widely varying nature of the accounting systems used, it is difficult to make blanket statements about the state of pension funds at the time. Some consistent data is available from the Census Bureau, which began a survey of state and local government employee retire-ment systems in the 1940s. Early data from that survey show that investment earnings were sub-stantially overshadowed by plan contributions from employers and employees until the 1980s, implying that most plans were not funded on an actuarial basis until then.

In the new era, a wide variety of accounting standards were used to assess the health of these pension funds. There were different ways to estimate the long-term liabilities, the marginal cost of a new employee, the value of the assets, and even the lifespan of the employees. Comparing one fund to another was challenging where it was possible at all. In 1994, GASB sought to address this problem with their Statements 25 and 27. These established, for pension plans and their spon-soring governments respectively, that actuarial data be included in fund annual statements, and strongly discouraged managing plans on a pay-as-you-go basis. They dictated that a plan calcu-late an “Actuarially Required Contribution” (ARC) according to their formula, and specified how a government should report whether or not it had contributed that amount to the plan fund.


The effect of GASB 25 and 27 suggested that the nation was undergoing a slow-boiling pension funding crisis. Many governments were not making the ARC demanded by their pension fund actuaries, and future liabilities seemed colossal.20 Debates raged, and continue to rage, about the proper discount rate to use to calculate future liabilities of systems, and whether pensions were viable at all. Some plans were closed, others scrapped in favor of simple savings plans now called “defined contribution” (DC) plans,21 while others were funded with bonded debt.22

After these reforms, despite the changes in incentives and standards, the unfunded liabilities continued to mount. Faced with what seemed to be a rising tide of red ink, GASB acted again in 2012, with Statements 67 and 68, again separated to apply to the fund and their sponsoring gov-ernment, respectively. By adding specificity to the rules, these statements created a greater degree of uniformity across pension plans. However, they also create a much more demanding set of rules for predicting future liabilities. For example, GASB 68 removed a government’s discretion to choose among the methods used by actuaries to allocate the marginal cost to the fund of an employee’s time on the job and put strict controls on the discount rate that a pension system must use to estimate its long term liabilities.

GASB 67 also codifies what was more or less standard practice for actuarial accounting, complete-ly excluding from consideration future contributions to the system from future employees or future employers. In essence, an unfunded liability calculated under these conditions, asks how much the sponsoring government will owe if the system is closed tomorrow and all current pen-sion debts paid off over the ensuing decades. However, most systems will not be closed tomorrow and will continue to receive contributions from both the employees and employers for decades to come. This makes an unfunded liability potentially useful as a planning value, but not a good prediction of actual payments to be made, as we will see.

1940 1950 1960 1970 1975 1985 1990

10%

20%

30%

40%

50%

60%

% t

otal

pay

men

ts

Year

11.51% 12.44%13.21%

23.75%

28.31%36.18%

42.76%

29.26%

38.20%

48.24%46.71%45.87%45.72%

58.30%

48.59%

28.01%

24.98%

17.95%

US Census Bureau State and Local Government Employee Retirement System Survey, as complied by Jun Peng. State and Local Pension Fund Management. Boca Raton, FL: CRC Press, 2009

Employee contributions

Employer contributions

Investment earnings

PENSION FUND EARNINGS BY REVENUE SOURCEPercent of total revenuePension Fund Earnings by Revenue Source

Percent of total revenue

Earnings from state and local pension funds. (Dollar values in millions.) Until the 1980s, investment earnings from pension funds did not play the largest role in funding pensions, many of which were pay-as-you-go systems with small funds used for managing cash flow rather than for generating investment income.19 You can see the importance of investment income growing substantially over time, but really picking up momentum in the late 1970s.


It is important to recognize the source of the pension funding crisis. Obviously, our nation—like all the others since the beginning of time—suffers from improvident politicians. There have been many skipped or shorted payments into the nation’s public pension systems. Financial market crises have also done important work to increase the pressure on pension funds. The financial market turmoil of 2000–2001 was especially severe, and the losses of 2007–2008 have not been won back by many funds. For example, CalSTRS, the fund for California teachers, was fully fund-ed as recently as 1998, before the popping of the tech bubble in 2000–2001.23

Skipped payments and disappointing investment returns are only part of the story. For many funds, these only exacerbated an original funding shortfall due to the transition between the pay-as-you-go model and the actuarial model. For governments making that transition, pension commitments already existed from the pay-as-you-go period. These commitments were made to employees who had not made contributions to the pension fund, not through any fault of their own, but because that was not how the system worked during their careers. Thus, many funds were established with a significant unfunded liability at the outset. In an accounting sense, this might be considered the original sin.24

But what is the source of the feeling of crisis that so dominates the discussion of pensions today? As with Detroit, debt due in the distant future is not a crisis today, even if it is a cause for concern. To a large extent, the source of the crisis is the accounting rules themselves and their misapplica-tion by policy makers and ratings agencies. The GASB statements themselves are bland and even seem thoughtful, but the uses to which they have been put somewhat less so.25

PROBLEMS WITH ACCOUNTING RULES“The pension fund could run dry,” is a common enough talking point that it could be made about almost any pension fund in the country. A 2016 Google search for “pension fund run dry” pro-vides ten hits, eight of which name Pennsylvania, New Jersey (twice), California, Chicago (twice), Texas, and Alabama with the remaining two hits leading to articles about the national pension funding crisis. Some of these plans named in these articles may be in real trouble, but reading them suggests that most of the writers are under the mistaken impression that a system funded at anything less than 100 percent is necessarily in danger of running out of money at some time in the near future.

“The state’s pension goliath, the California Public Employees’ Retirement System, had $281 billion to cover the benefits promised to 1.3 million workers and retirees in 2013. Yet it needed an additional $57 billion to meet future obligations.”26

“Before the crash, retirement systems were underfinanced (they did not have sufficient funds to pay promised benefits), but the day of reckoning was distant.”27

“ ‘Their benefits are in question,’ said Gary Wagner, a professor of economics at Old Dominion University.”28

Some experts will see these as alarmist statements. GASB members themselves might say these writers are misunderstanding the rules, even if they agree with the conclusions. The GASB State-ments 67 and 68 specifically say they are only about reporting and do not dictate funding. And yet, stories like these appear across the country on a near-daily basis. The clear conclusion is that the effect of GASB rules is not just on the construction of a balance sheet, but on the interpreta-tion of the numbers found there and the actions of the parties who make those interpretations: policy makers, citizens, or bond-rating agencies.


If GASB itself is not the enforcer of misinterpretations, if the enforcer is city council members preening about their soi-disant fiscal responsibility, or analysts at Moody’s determined to justify a downgrade, or newspaper columnists looking for a good hook, they are doing so with tools supplied by GASB. It is disingenuous to erect a framework of tough rules and disavow their con-sequences. In this case, the consequences are a drive to full funding, whatever the cost.

It is crucial to understand, now and during the discussion to come, that the goal of the GASB rules is not to bankrupt governments, or to eliminate pension systems. At their root, the GASB rules are meant to create an accounting framework through which the cost of government and the cost of any individual employee are made clear. Unfortunately, the framework erected has created more problems than it has solved. The problem is not merely that the rules are commonly misinterpreted. GASB accountants have acted to insulate public plans from risks they do not face, while simultaneously failing to insure them against risks they face every day. Furthermore, by try-ing to bring clarity to the accounting, the rules undermine the benefits of aggregation—the whole rationale for a defined-benefit pension plan.

Again, this is hardly meant to say that there have not been improvident politicians and unwise bureaucrats, but the pension “crisis” currently affects responsible and irresponsible governments alike. Two decades of disastrous experience with the GASB pension accounting demand that we examine the rules themselves. We categorize the problems as legal, chronological, actuarial, mathematical, financial, economical, political, and philosophical. We examine them in this order.

Legal: Governments will not be liquidatedBeginning in 1994, with Statements 25 and 27, the GASB rule changes about public pensions were made to mimic rules in the private sector. However, when applied in the public sector the rule changes make pensions more expensive than necessary.

Consider the issue of full funding. A fully-funded pension system can, at least in theory, pay off all its current debts with no further contributions from the sponsoring employer. This is vital in the private sector because at any time, a private corporation can go out of business, be liquidated and disappear. Full funding and custody by a third party is the only way to make sure a pension granted by such a business will be paid. A pension system in the private sector must be ful-ly-funded in order for the promise of the pension to mean anything at all.

By contrast, a government will not disappear in the same way. To claim so is only to agree with, among others, GASB itself. In a 2006 paper called, “Why Governmental Accounting and Finan-cial Reporting Is—And Should Be—Different,” they defend the difference between governmental accounting standards and those appropriate for the private sector. Early on, the authors point out that the lack of a threat of liquidation is among the primary differences:

“[M]ost governments do not operate in a competitive marketplace, face virtual-ly no threat of liquidation, and do not have equity owners.”29

A hundred years from now there will still be a New York City, even if its area has been reduced by rising sea level. It may have suffered a bankruptcy—maybe two or three—but bankruptcy is not liqui-dation. It may have been split into its five boroughs, or it may have been overtaken and merged with a rapidly growing Yonkers, or maybe even taken over by the state. Unlike the liquidation of a private company, each of those possible transitions, however absurd or unlikely, leaves a successor to assume the responsibilities of the previous government. Insurance against the city’s disappearance is therefore a waste of money.30


Chronological: Present value masks the level of urgencyPension accounting relies on a presentation of assets and liabilities in their “present value,” the value today of a sum of money tomorrow. At a 5 percent annual discount rate, a present value of $100 today corresponds to a future value of $105 a year from now. The appropriate discount rate depends on your estimates of inflation and potential investment returns, so there is a degree of subjectivity in any present value calculation. Nonetheless, there is a time dependency of the val-ue of money, so it makes no sense to compare 2016 assets with 2046 liabilities, except by comput-ing the present value of the liabilities to compare to the current value of the assets.

Unfortunately, though the concept of present value is a way to translate future funding into the present day, the calculations made to describe the present value of a pension debt elide important issues surrounding a payment schedule, even beyond the issues of subjec-tivity. By definition, present value captures the monetary value of a future series of payments, but it fails to capture the urgency of those payments. By making all calculations in terms of present value, pen-sion accounting rules imply an equal urgency to all debts, something that is obviously not the case. For a pension liability, the relevant payment schedule not only extends out decades into the future, but it also ex-tends decades longer than the 30-year amortization periods required by GASB 27. The last payment owed by any pension system will not be made until the youngest current employee dies. If the youngest employee is in their 20s, this could be more than 60 or 70 years in the future. For systems that offer survivor benefits, it could be longer than that.

Consideration of the payment schedule is important to the stress of a debt because two debts with the same present value can require dra-matically different plans for payment. Imagine a debtor with assets of $600 and a debt of $1000, due tomorrow, and compare him to another debtor with the same $600 in assets, but who owes

REVENUE SOURCE 1 $11.76/week from savings

REVENUE SOURCE 2$7.96/week from paychecks

Only 60% funded

REVENUE SOURCE 1 $600 in savings Only 60% funded

Funding crisis.

I owe you $1,000.12 months from now.

Funding is going to be okay.

I owe you $1,000.Tomorrow.

THE FUNDING RATIO OF DEBTWhat else matters?The Funding Ratio of Debt

What else matters?

The same present value and the same funding ratio can be a crisis or not, depending on variables that the present value of a debt does not capture. One of these persons is in serious trouble, and the other only mildly concerned, but both have debts with the same present value.


$19.72 per week for a year. Assuming a discount rate of 5 percent, these debts have precisely the same present value, and therefore precisely the same $400 unfunded liability. Yet one debtor is in much better shape than the other. The first debt could justifiably stimulate panic; the second is certainly worth more than a yawn, but much less than panic. Action is required, but the debtor has until week 32, over seven months away, to mull over action or to raise additional funds. The Illinois pension debt is $111 billion, but this need not be paid tomorrow, even if GASB 68 will have it appear on the same balance sheet as debts that are due tomorrow, or even past due. Ac-countants might claim these two debts are equivalent, but is this really the case?

As an incidental point, one can note that the traditional 30-year amortization schedule is only that: traditional. A pension system that dutifully follows such an amortization schedule will see its debts prepaid decades before they are actually due. Paying them in advance is not necessary to making payments, and yet it is the accepted wisdom. One searches in vain for any author present-ing a reasoned justification for equating 30-year debt with immediate debt, or using a 30-year term to pay off a 60-year debt. These rules represent little more than blind acceptance of prece-dent.

Actuarial: Full funding is not required to pay all pension debtsThe drive to full funding cannot be justified actuarially, either. Though the details depend on actuarial characteristics of the employee and retiree population, many, if not most, defined-ben-efit pension systems can operate forever at far less than full funding. A retired teacher in Chicago who passed away in 2014 after a long and happy retirement had every penny of her pension paid by a system far below full funding, and yet all her pension checks cleared. A system at 70 per-cent funding can likely pay all its obligations in a given year, and if at the end of that year it is at 70.1 percent, who is to say this cannot be repeated the following year if the actuarial facts on the ground do not change significantly? Social Security operated at what amounted to a few percent-age points of full funding in its trust fund for two generations and only a very few pension plans are funded at levels so low.31

To put it more rigorously, the normal cost to a pension plan accumulated within a calendar year is the present value of the additional benefits accrued by all the employees in that year. If the contributions to the fund (employer and employee contributions, as well as investment income) are adequate to offset the normal cost and inflation, then the unfunded liability of a plan will not change from one year to the next.32 If the unfunded liability does not change one year to the next, the fund can operate indefinitely with that same unfunded liability.

A pension fund must pay 100 percent of its debts. But it need not pay them a moment before they are actually due, and since a pension plan is constantly receiving new contributions, the fund itself need not be the only source of payments. As a result, even if all the debts are paid, at any one time, the fund itself may be at some level well below 100 percent funding.

Recall the example above. Perhaps I took a $1,000 loan from you, promising in return to pay you $19.72 per week for a year. If I have only $600 in the bank, then I have an unfunded liability of $400. If I also have some source of income of just $7.96 per week, I will be able to pay 100 percent of this debt, down to the penny, out of the combination of my income and my savings. Every step of the way, my funding ratio—the ratio of my assets to the present value of my remaining debt—will be 60 percent or less. (See figure on page 13.)

This is a toy example and tracks the debt to only a single party. A pension system might have debts owed to tens of thousands of members or more, all owed on their own schedule. The debt estimates are also subject to considerable uncertainty, since the demographic mix of employ-ees and retirees changes over time, too. The principle, however, is the same. So long as there is another source of income, the ratio between the fund and the present value of the debt has little


to do with how much of that debt is ultimately repaid. An active pension system has three sources of income: (1) the contributions from the employer, (2) from the employees, and (3) from the returns on investment. Belaboring a point of arithmetic like this seems a waste of time, but the implication that the funding ratio has some relevance to the full repayment of the pension debt is not only a staple of public pension criticism, 33 but is enshrined in official policy statements from bond-rating agencies, actuaries, and the National Association of State Retirement Administra-tors.34

A 2008 report issued by the Congressional Government Accountability Office (GAO) agreed with the assessment offered here:

“Most public pension plans report having sufficient assets to pay for retiree benefits over the next several decades. Many experts and officials to whom we spoke consider a funded ratio of 80 percent to be sufficient for public plans for a couple of reasons. First, it is unlikely that public entities will go out of business or cease operations as can happen with private sector employers, and state and local governments can spread the costs of unfunded liabilities over a period of up to 30 years under current GASB standards.”35

Paying a Debt Over a Longer TimePaying a debt with a limited funding rate

5 10 15 20 25 30 35 40 451 2 3 4 6 7 8 9 50

Funding Ratio

Remaining Debt

Savings

60.0

0%59.95%

59.90%

59.84%

59.7

9%59.73%

59.68%

59.62%

59.56%

59.5

0%

59.1

7%

58.7

9%

58.3

2%

57.7

1%

56.8

5%

55.4

3%

52.4

1%

39.8

3%29

.85%

$100

0.00

$981

.24

$962.46

$943.66

$924.84

$906

.00

$887.15

$868.28

$849.39

$830.48

$811

.55

$716

.65

$621

.29

$525

.47

$429

.18 $3

32.4

4

$235

.22

$137

.54

$39.

39

$600

.00

$588.24

$576.47

$564.71

$552

.94

$541.18

$529.41

$517.65

$505.88

$494

.12

$435

.29

$376

.47

$317

.65

$258

.82

$200

.00

$141

.17

$82.

35

$23.

53

Dol

lar A

mou

nt($

)

Time(weeks)

Funding Ratio

(%)

PAYING A DEBT OVER A LONGER TIMEPaying a debt with a limited funding ratio

Savings

Funding Ratio

Remaining Debt

5 10 15 20 25 30 35 40 451 2 3 4 6 7 8 9 50

Funding Ratio

Remaining Debt

Savings

60.0

0%59.95%

59.90%

59.84%

59.7

9%59.73%

59.68%

59.62%

59.56%

59.5

0%

59.1

7%

58.7

9%

58.3

2%

57.7

1%

56.8

5%

55.4

3%

52.4

1%

39.8

3%29

.85%

$100

0.00

$981

.24

$962.46

$943.66

$924.84

$906

.00

$887.15

$868.28

$849.39

$830.48

$811

.55

$716

.65

$621

.29

$525

.47

$429

.18 $3

32.4

4

$235

.22

$137

.54

$39.

39

$600

.00

$588.24

$576.47

$564.71

$552

.94

$541.18

$529.41

$517.65

$505.88

$494

.12

$435

.29

$376

.47

$317

.65

$258

.82

$200

.00

$141

.17

$82.

35

$23.

53

Dol

lar A

mou

nt($

)

Time(weeks)

Funding Ratio

(%)


Savings

Funding Ratio

Remaining Debt

5 10 15 20 25 30 35 40 451 2 3 4 6 7 8 9 50

Funding Ratio

Remaining Debt

Savings

60.0

0%59.95%

59.90%

59.84%

59.7

9%59.73%

59.68%

59.62%

59.56%

59.5

0%

59.1

7%

58.7

9%

58.3

2%

57.7

1%

56.8

5%

55.4

3%

52.4

1%

39.8

3%29

.85%

$100

0.00

$981

.24

$962.46

$943.66

$924.84

$906

.00

$887.15

$868.28

$849.39

$830.48

$811

.55

$716

.65

$621

.29

$525

.47

$429

.18 $3

32.4

4

$235

.22

$137

.54

$39.

39

$600

.00

$588.24

$576.47

$564.71

$552

.94

$541.18

$529.41

$517.65

$505.88

$494

.12

$435

.29

$376

.47

$317

.65

$258

.82

$200

.00

$141

.17

$82.

35

$23.

53

Dol

lar A

mou

nt($

)

Time(weeks)

Funding Ratio

(%)


Savings

Funding Ratio

Remaining Debt

5 10 15 20 25 30 35 40 451 2 3 4 6 7 8 9 50

Funding Ratio

Remaining Debt

Savings

60.0

0%59.95%

59.90%

59.84%

59.7

9%59.73%

59.68%

59.62%

59.56%

59.5

0%

59.1

7%

58.7

9%

58.3

2%

57.7

1%

56.8

5%

55.4

3%

52.4

1%

39.8

3%29

.85%

$100

0.00

$981

.24

$962.46

$943.66

$924.84

$906

.00

$887.15

$868.28

$849.39

$830.48

$811

.55

$716

.65

$621

.29

$525

.47

$429

.18 $3

32.4

4

$235

.22

$137

.54

$39.

39

$600

.00

$588.24

$576.47

$564.71

$552

.94

$541.18

$529.41

$517.65

$505.88

$494

.12

$435

.29

$376

.47

$317

.65

$258

.82

$200

.00

$141

.17

$82.

35

$23.

53

Dol

lar A

mou

nt($

)

Time(weeks)

Funding Ratio

(%)


Savings

Funding Ratio

Remaining Debt

Amortizing a debt with both income and savings. The lower straight line is the savings balance at the beginning of the week, and the line above that is the value of the debt remaining. Both lines decline gently to zero as the savings are exhausted and the debt retired. Above them is the funding ratio, and you can see that at no time does the funding ratio go above 60 percent—in fact it declines each week—and yet 100% of the debt is paid. The table assumes an annual discount rate of 5%.


Mathematical: Financial reports deserve more precisionAnother objection to GASB 68 has to do with precision. A pension plan’s unfunded actuarial liability is a planning value, based on dozens of assumptions about market performance, pop-ulation mortality, and the life choices of hundreds or thousands of employees. Comparisons of identically-calculated planning values like this are tremendously useful for identifying trends in funding progress or lack thereof. However, the accuracy of these numbers is contained in bounds much larger than is normal for other numbers found on financial statements, such as accounts payable or bonded indebtedness. Pension planning values are a guess about the future, useful primarily to compare to each other. Financial statements are a record of the past. Despite the statistical methods used to develop these estimates of future liabilities, actuarial and accounting precedent do not demand the explicit reporting of the error bounds, as one sees in other statis-tical estimates, such as polling data. Rendering these inherently inaccurate numbers as part of a government’s statement of net assets, as demanded by GASB 68, dramatically reduces the accura-cy of that statement’s bottom line.

The issue is not the mere uncertainty of the numbers. Accountants use uncertain numbers in many of their calculations. The problem is that uncertainty is infectious. Adding a high-precision number to a low-precision number results in a low-precision number, so the result is low-preci-sion financial reports.

Given the size of pension debts, adding them into the total can make the actual value of a government’s net assets vary significantly from the stated value. According to its 2014 financial statements, the unfunded liability for Illinois pension funds is estimated at $111 billion, but the potential error in that estimate is as large as the absolute value of the state’s revenue shortfall of $45 million. Depending on the specific fund, the assumed rate of return varies from 3 to 6.5 per-cent. These are conservative assumptions compared to many peer systems. A difference of only a single percentage point between these guesses and the reality of the next few decades will change the state’s bottom line by over $20 billion.36 It is not typical to think of a statement of net assets as potentially uncertain by more than 50 percent of the bottom line, but that is the new standard of accounting according to GASB 68.

Financial: Rate of return is for the longest term.GASB 68 specifies that a pension system below a 100 percent funding ratio must use a “risk-free” rate of return as the discount rate for estimating its future liability. This liability is the number that must appear on a government’s financial statements, alongside the more traditional com-ponents of its liabilities, like bonded indebtedness, and accounts payable.37 The risk-free rate of return is what one can count on earning on investments with no risk. More or less, this would be a portfolio that is entirely invested in US Treasury bonds or the equivalent: no stocks, no municipal bonds, no private equity funds, no commercial paper, no hedge funds.

Statement 68 does not insist that a fund actually be invested solely in Treasury bonds, only that it use that rate to predict its future liabilities. A fund can continue to use whatever funding strategy its managers see fit to use. But the liabilities will be calculated using this lower rate of return, thus will appear much larger than if calculated with a higher rate. This is important because GASB 68 requires the entire unfunded debt to appear each year in the government’s balance sheet. Before this statement, under the requirements of the earlier Statement 27, a government only had to acknowledge whether or not it had made the ARC, the appropriate annual payment, to the fund.

There is an important point that is often brought up here, that the discount rates commonly in use for pension funds around the country are too high, and the modern world of low interest rates and low investment returns is here to stay. If it is indeed impossible for a well-managed portfolio to average 7.5 percent returns over the next few decades, then 7.5 percent should not be used.


The debate about what will happen in investment markets over the next 50 years often seem unnecessarily heated. The debate is between people who correctly point out that these numbers have been achievable for decades and people who claim that things are different now. While it is true that many pension systems have been able to meet their marks over the long term, it is also true that we have suffered a decade of reduced investment returns, and the prospects for im-provement are not obvious.

In truth, neither side of this debate has any better claim than the other about the future. Both sides are defensible, and a determination about who is right can only await the coming decades. It is certainly true that a system that assumes a low but achievable rate is more conservative than a system that relies on a standard that might or might not be reached. But these discussions fre-quently elide an important point: few such debates are about how to design a brand new pension system. Rather, debates about such issues are debates about how to manage the systems we have already. Were one to consider establishing a new pension system, certainly choosing a low rate of return will be a good idea. It will make the system more expensive to run, but it will be more secure, too.

But what will the effect be on an existing system to dramatically lower the discount rate? This is not a small step to consider, as it will increase that overall liability considerably. A $10 billion future liability over 30 years at a discount rate of 7.5 percent will see that liability balloon to well over $15 billion with a 5 percent rate. Upon the adoption of GASB 68, the average funding ratio will decline around ten or fifteen percentage points as hundreds of billions of dollars of “new” liability is recognized under the new rules.38

Every cent of that new liability will appear on the governments’ statement of net assets. For many leaders and critics of their governments, showing a huge debt on the bottom line will be too much red ink to contemplate with equanimity and political pressure to do away with these obligations will build further—with potentially destructive consequences. A rate too high risks un-derfunding which may lead to higher taxes in the future, but a rate too low risks political pressure which may lead to reduced or eliminated benefits in the future. Readers will differ about which is the more salient risk.

In other words, GASB 68 creates large disincentives for governments to use a lower rate at the same time it requires them to do so. This is a recipe for unnecessary political crises across the country. It is vital that government financial reports be clear about what a government’s debts ac-tually are, but it is also vital that these statements should not be misconstrued. For all the reasons outlined here, there is a substantial downside risk to acknowledging reductions in the assumed rate of return. If changes like these are to be made, they must be made slowly, as financial and political circumstances permit. These are systems meant to be run in perpetuity; most can afford to be patient.

One can see a cautionary tale in the woes of the US Postal Service and its retirement systems. The USPS is technically not bound by GASB accounting rules, but in late 2006, President Bush and the Republican Congress passed a law to force the system to estimate its retiree and health-care liabilities 75 years in advance. The requirements insist, that is, that the system account for the retirement expenses not only of postal workers not yet hired, but of those not yet born. Increas-ing the funding horizon so dramatically and suddenly is essentially the same kind of shock to the system that an abrupt increase in the discount rate would create. In the ensuing decade, the Postal Service has largely managed to fulfill its funding mandate. As of 2015, the USPS had over $335 billion saved, covering over 83 percent of the liabilities anticipated over 75 years, under very conservative assumptions about the discount rate and value of current assets.39 But the price has been high, and the service incurred $51.7 billion in operating losses between 2007 and 2014 as a result. These losses have shorted new capital investment and service expansions and left the service open to persistent charges that it is an obsolete money-loser at the same time it was forced to put aside a breathtaking sum of money.


In a 2014 interview, the deputy director of the California State Teachers Retirement System (CalSTRS) pointed out his fund had averaged better than 7.5 percent returns for decades. Under the new rules, he said they must use a rate of about 4.5 percent, increasing the present value of their liabilities by more than 50 percent, but that the portfolio would go along, earning its 7.5 per-cent as before. In what way, he asked, do the new rules provide an accurate picture of that fund’s condition?40

Economical: Comparisons to pension debt should be chosen properlyConsider again the Illinois pension plans’ unfunded liability of $111 billion in 2015. This is a vast sum of money, especially for a state with an annual budget of only a bit more than half that amount. However, this debt took decades to accumulate and it will be decades before it must be paid off completely. Most, but not all, of the debt will be paid out over the next 50 years. Over that time period, using a very conservative 2 percent inflation estimate, the state’s income tax collections alone will be in the neighborhood of $1.1 trillion and the total state budget will in-volve spending well over $4 trillion. Therefore, this supposedly colossal debt in reality constitutes about 2.5 percent of the state budget. Personnel costs are around a quarter of the Illinois budget, so this is roughly 10 percent of the payroll costs over that time period.41

Stepping a bit further back, the Illinois economy is much larger than the state budget, and over those same 50 years, can be expected to produce around $64 trillion.42 That is, another way to look at this debt is that it is 0.17 percent of the state’s gross product over the term during which it will be paid, a much less frightening number. These are not spurious comparisons; the state’s economy and the revenue it receives are precisely the resources the state will use to pay this debt.

A roughly equivalent way to state this objection is that the GASB rules do not acknowledge as an asset the strength of the local economy and the ability of its taxpayers to pay in the future. Ensuring that the government can pay its obligations in the future is virtually the entire goal of the GASB accounting rules, but the rules are narrowly drawn so that only qualifying funds kept in trust are counted as a strength. Is it necessary to take such a narrow view of the ability to pay? Another vital difference between a government and a corporation is that a government has a claim on the potential future income of its citizens that no private corporation can make. Indeed, this works in both directions, since money withheld from the economy in the near term, say by raising taxes or cutting schools to make inflated payments to a pension fund, can reduce econom-ic growth and thereby make pension payments more onerous in the future by reducing the size of that future economy.43

Political: Overfunding is a risk, tooA serious risk that does not get much attention at present is the overfunding of a pension plan. Such a risk may seem almost laughable given current circumstances, but it is a serious risk, worth serious concern, not least because it is the goal. The primary objective of pension funding policy choices from GASB to city hall is full funding and full funding is one good investment year away from overfunding. Indeed, full funding is arguably a synonym for overfunding.

The primary concern is that overfunding is a tangible waste of resources, money unnecessarily diverted from other priorities, but there are other potentially disturbing consequences. A pension system—especially a well-funded one—is not insured against the depredations of politicians, for whom the near-term cost of a pension gift or skipped payment is quite low, if not zero. The GAO report cited above puts it this way:


[S]everal [experts] commented that it can be politically unwise for a plan to be overfunded; that is, to have a funded ratio over 100 percent. The contributions made to funds with “excess” assets can become a target for lawmakers with other priorities or for those wishing to increase retiree benefits.44

In other words, it is virtually a law of nature that an overfunded pension plan—or any plan over, say, 90 percent funded—will see retiree benefits increase or budgeted contributions decrease.45 In the context of full funding, these changes will have little or no current cost. And as predictably as the sun rises, after the next investment downturn, it will be an underfunded pension plan again, but now with a government budgeting for lower payments and retirees accustomed to higher benefits. This will unavoidably cause a delay in raising payments to catch up.46 This is precisely how events worked out for CalSTRS, the giant pension fund for California teachers. Fully-funded in 1998, its investment returns plummeted when the tech bubble popped in 2000–2001, but not before the state cut its payments into the system and increased some classes of benefits.47 The system currently has a $73 billion unfunded liability and a 68 percent funding ratio, quite a fall from 1998.

CalSTRS was hardly alone in its experience. The Chicago Teachers Pension Fund was fully-funded in 1995, which was used to justify a ten-year “holiday,” dramatically reducing payments into the fund. Even by 1999, the system was still full-funded. Along with this holiday came increased ben-efits and management expenses and those, combined with the two colossal incidents of financial market turmoil since then. As of 2014, the plan is 51.5 percent funded.48 The pattern is reflected in national averages. Census Bureau data shows that 1997 was a high-water mark for employer contributions to public plans, many of whom cut contributions in the warm glow of full funding, but had to restore them by 2003.49

Under the GASB rules, a mayor who chooses to skip a pension payment or settle a labor dispute with an ill-considered increase in pension benefits might incur the displeasure of a rating agency in future years. That, in turn, might raise the cost of borrowing down the road: a problem for a future mayor. Furthermore, many cities currently have what are essentially junk bond ratings. For these cities, the ratings risk only barely rises to the level of a material concern, and these conse-quences are only potential, not certain. In other words, the cities most likely to skimp on their pension contributions are the least likely to be harmed by the consequences in the near term.

For CalSTRS, the Chicago schools, and so many others, reduction of the dollars flowing into the pension systems harmed the health of the funds, but in each of these cases the effect on the cur-rent budget of the sponsoring government was zero. In fact, to the extent that a skipped payment or a new police department contract relieves financial pressure on the city, the effects can be posi-tive, in the short term. A good accounting system is supposed to provide an accurate picture of an organization’s financial health and a useful guide to action. In these cases, the accounting system guides its users to destructive and inappropriate action.

We have already seen how the GASB rules insure against the liquidation of a city or state, a risk that does not exist. Here we see the rules simultaneously failing to insure against risks that public systems do face. In some cases, the rules even encourage those risks. The divergence between the risks insured and the risks actually faced is not merely ironic; it is a recipe for failure. The insured risks never materialize while the others always do, given enough time. In this case, failure means further increases in the cost of employee pensions, further ire directed at teachers, police officers, and other public employees, and more stress on the tottering finances of state and local govern-ments.


Philosophical: A pension plan is a mutual, not individual, arrangementThe last, philosophical, objection to the GASB framework requires a look into the fundamental principles behind a pension plan.

It is typical to speak of pension plans as belonging to one of two varieties, defined benefit (DB) and defined contribution (DC). A “DB” plan is a modern name for a traditional pension plan while a DC plan is just another name for an employer-sponsored savings plan. These two catego-ries of plan are frequently portrayed as distinguished by the level of risk incurred by the employ-ee and the employer. The employee can be said to bear much more of the risk in a DC plan than for a traditional plan and vice versa. This is all too true, but portraying this as a question of which of two parties assumes the risk is inaccurate, because with a traditional plan there is essentially a third party to the equation: the body of plan members as a whole.

A traditional pension plan works because not all the people paying into the system will have a long and happy retirement. Baldly put, some of the members will die before enjoying all the benefits to which they are entitled. Those who do not will see their retirement financed by those unused contributions. The consequence of this reality is that it is very difficult to separate the value of one member’s contribution from another. The value to each member is that they are all in the fund together, insuring each other.

By contrast, the principle behind the GASB accounting reforms is roughly that accountants ought to be able to match the marginal cost of each employee with the marginal contribution from that same employee. A DB plan is a collective entity, but the GASB accounting insists on looking at individuals. Quoting directly from Statement 68:

“For defined benefit pensions, this Statement identifies the methods and as-sumptions that should be used to project benefit payments, discount projected benefit payments to their actuarial present value, and attribute that present value to periods of employee service.”

In other words, the rules provide guidance for determining what fraction of the fund “belongs” to any individual employee, given in exchange for their work in some particular year. This is the root of the GASB insistence that current employee contributions should not be used to pay cur-rent retirees, and the insistence that all plans should be 100 percent funded. Only a fully-funded plan can have all of the employee-years allocated to employee shares of the fund assets. There is no clarity to the accounting otherwise.

A pension plan is a mutual insurance arrangement. The collective financial strength of the body of members is greater than the sum of the parts. In a pension plan, every member has an equiv-alent claim on every dollar coming into the system. There is no sense in which a dollar of contri-bution “belongs” to this or that member, and systems do not define a priority among members. The employer and the fund itself are a source of strength, but so are all the other members: three sources of security. If one is weak, the others are still available. A plan with a 30 percent funding ratio obviously is less secure than a plan at 80 percent, but under the right demographic condi-tions, it can still run indefinitely because of the employer and all the other plan members. By insisting the only important thing is accounting for individual contributions and expenses, the GASB rules seek to erase this source of security from consideration. Worse, the rules lead to de-cisions that undermine the third source of security, by making full or partial plan closures seem like a sensible idea.

When critics complain that “generational equity” demands that one age cohort must not subsi-dize another, they are demanding less security than pension plans were invented to provide.


One age cohort does provide security for another and the same security will be provided to them in turn, as well as those after them. This is how these plans were designed. But this kind of equity is not an absolute good. Presumably if generational equity were valued above all other consid-erations, then teacher salaries would only be financed with debt, so the children who reap the benefit would eventually pay the expense. Perhaps it is best to describe this as a different kind of generational equity, where one generation receives the benefits it enjoyed in the past. The kind-ness we extend to our children is different from the kindness we expect from them, but that does not excuse them from extending the same kindness to their children in turn.

The invention of mutual insurance redefined life among the working and middle classes in the nineteenth and early twentieth centuries. Insurance was not the only such innovation, but was part of a movement toward the democratization of finance that included life insurance, disabil-ity insurance, and unemployment insurance, not to mention savings banks, savings bonds, and mutual funds. The push to old-age insurance was a part of a political movement that began years before 1911, when the pioneering director D. W. Griffith produced a film called “What Shall We Do with Our Old?” The answer the movie proposed was to establish old-age insurance to provide pensions to alleviate poverty among the elderly. It took decades of effort, but eventually politi-cians and industry responded. Private pension systems were established, and between 1914 and 1934, when Social Security was established, twenty-eight states had established experiments with old-age pension plans for their poorer residents.50

The plot of Griffith’s film was a tragedy, meant to illustrate a preventable irony: old-age penury in a wealthy society. The refinement of mutual insurance into “old-age insurance,”—which subse-quently became known as pension plans—has all but erased this kind of poverty in developed countries. It seems an odd claim that “clarity” in accounting should be deemed more important than that.

OTHER POST-EMPLOYMENT BENEFITS CRISISThe issue of “Other Post-Employment Benefits” (OPEB), mainly health care expenses for retirees, is technically somewhat different than the issue of pension plans. However, the two are often spo-ken of together, and so the OPEB crisis deserves at least a brief treatment here. GASB Statements 43 and 45 did for OPEB accounting roughly what Statements 25 and 27 did for pensions, with three important differences.

The first difference is that many fewer governments have established a fund for OPEB benefits at all. When GASB made its Statements 25 and 27, most pension systems were already funded on an actuarial basis. The statements had the effect of creating a welcome degree of uniformity, in addi-tion to the less salutary effects described here. In 2004, however, when the OPEB statements were issued, few governments were pre-funding those expenses, so the effect of the new rules was to demand that such funds be created. The unfunded liability for all government-sponsored OPEB plans in the country is estimated to be over $2 trillion.51

The second difference between OPEB funding and pension funding is that typically there are no contributions to an OPEB fund by current employees. OPEB benefits are usually incurred as an expense by the employer alone, removing a significant contributor to the security of those funds.

The third problem is the most significant, that the rate of health care inflation is—at present—greater than the rate of inflation for pretty much everything else.52 To make OPEB liability calculations, actuaries typically project the current rate of health care cost inflation forward 50 years or more. Projecting even a small percentage growth forward over half a century produces a tremendous sum, and the resulting liabilities are indeed enormous.


However, the current rate of health care inflation is not sustainable over the next 50 years by any component of society, not just pension plans. It is not just city and county pension funds that will suffer if our nation fails to control these costs, the federal government, hospitals, non-profits, schools, corporations, and individuals will all be transformed in appalling ways by these escalat-ing costs.53

Fifty years from now, if health-care inflation is not lowered significantly, all fifty states will have been bankrupted by Medicaid costs. The federal government will only have avoided a similar fate by printing enough money to devastate the value of Treasury bonds, causing worldwide financial instability. American corporations will pay more in health care to their employees than in salary. As the rest of the economy withers, health care costs will grow until they are much more than a quarter of GDP.

The GASB accountants made a valid point about costs when they issued Statements 43 and 45. Indeed this nightmare scenario could happen. The issue is hardly the cost of funding these ben-efits. The issue is the escalating cost of health care for retirees and everyone else, retarded only slightly by the Affordable Care Act. The GASB rules in practice are akin to planning for an aster-oid collision with earth by putting away enough money to pay the electric bill when it happens. The asteroid collision could happen, but the electric bill will hardly be anyone’s first concern. The irony is amusing, but if saving for the electric bill actually interferes with taking protective action for the rest of society, then it is actually destructive.

COMMON “SOLUTIONS” THAT SELDOM SOLVE ANYTHINGThe accounting rules for pensions as described by GASB are not merely poor indicators of a system’s financial health and stability, they are an equally poor guide to remedial action. What appear, under these rules, to be sensible strategies for improving a pension’s funding are, when judged by practical experience, highly risky and speculative ventures that routinely create ca-tastrophe where none was necessary.

Obviously any plan with a funding shortfall can simply modify the contributions to the plan or the benefits paid from it, and these solutions are legion: raise the retirement age, lower the bene-fit, raise the employer contribution, raise the employee contribution. Some plans have opted to create tiers of employees, to economize by providing lower benefits for newer hires. These are all part of the standard austerity playbook, and adequately covered in many texts on the subject.54

But a few other options exist and are widely used, often to the detriment of the policy makers who choose them. These include (1) pension obligation bonds, (2) closing a pension plan, and (3) shifting employees to a defined contribution plan. Each incurs risks at least partly obscured by the accounting rules. A partially funded pension plan that uses one of these strategies takes on this hidden risk and therefore may face a higher risk of funding shortfalls in the future than a similar plan that forgoes them.

Pension obligation bonds: Investing on the marginThe liability of a pension is a debt owed to the future retired members of a pension plan. The fund assets are the wherewithal currently on hand to pay that debt, so the unfunded portion of that liability can be thought of as an unpaid debt compounding at an interest rate equal to the discount rate used to forecast liabilities.


According to this analogy, the sooner it is paid off, the better. If it is possible for a government to borrow at a lower rate than that discount rate—and for most governments it is—an obvious strate-gy is to borrow at the low rate in order to pay back the debt compounding at the higher rate. This is a pension obligation bond (POB).

However, another way to look at it is that a government that issues a POB is borrowing on the margin, and betting the returns achieved will be better than the interest to be paid. This is high-risk investing. Oakland, California, issued the first such bond, in the 1980s, and Congress quickly acted to say that such bonds are not tax-exempt. Pension bonds are thus market rate bonds. Since the governments that issue them tend to be the troubled governments, whose bond ratings are likely not AAA, the margin between the interest rate paid and the pension fund discount rate tends to be small.55

Imagine a pension fund with $2 billion in assets, with an unfunded liability of a billion more, and thus a funding ratio of 67 percent. Assume all goes very well, and the government borrows that extra billion at 5 percent over 30 years, while managing to invest it all at an average of 7.5 percent over the whole term. This sounds good, but another way to look at this is that the fund actually earns 7.5 percent on two-thirds of the fund and only the difference between that number and the POB interest rate, or 2.5 percent, on one-third of the fund. This creates an effective rate of return of 5.83 percent, much less than the assumed rate. In other words, the cost to the government of this pension plan is the same as a fully-funded plan that is not achieving its investment targets. This is hardly a knock-out argument in favor of such a bond.

Furthermore, the bond capital must be repaid, not just the interest. If the POB payments capture the bulk of what might have amortized the unfunded liability, and the government and employ-ee contributions just barely cover the normal costs of plan members, the plan may not be in ap-preciably better shape at the end of the POB term than it was at the outset. That is, after a 30-year POB has been paid back, the funding ratio for our example fund might not improve at all—even if everything goes as well as can reasonably be expected. This will seem especially harsh since, for the previous 30 years, policy makers will have regarded the plan as fully-funded. The temptation simply to roll over the debt by issuing another such bond will be very strong.

Of course it is very seldom that everything goes as well as can be expected. In reality, financial market returns are volatile, and POB margins are small, so the success of such a bond depends heavily on the market timing. In a survey of several thousand POBs, researchers with the Center for State and Government Excellence found that, as of 2014, of the POBs issued in the previous 20 years, most were only barely in positive territory, and a few had lost quite a lot of money. Tim-ing and luck appear to be the determinants, awkward components out of which to build sensible policy.56 The Government Finance Officers of America is fairly blunt in their assessment (part of their collection of “best practice” documents): “The GFOA recommends that state and local governments do not issue POBs…”57

Even more unfortunately, the fixed borrowing costs of any bond market transaction encourage governments planning to take such bets to take large ones. Detroit, Michigan, sold $1.44 billion in POBs in 2005—and lost $2.8 billion on the deal.58 Stockton, California, sold $125 million in POBs in 2007, and had to go into bankruptcy when the gamble failed. The fund was left in worse shape than before the bond was issued, and the city owed the POB capital to its bondholders.59

Closing a plan: Riskier than you might thinkMany political leaders, faced with an apparently colossal pension debt have concluded that clos-ing their government’s pension system is the only feasible alternative. The obvious problem with such a strategy is that it is not an answer to the question of funding the debt. That is, closing the plan because the debt is too large does not excuse a government from paying that debt, even if it does keep the debt from growing further.


Closing a pension plan has important risks that may not rise to the awareness of a policy maker looking at the raw numbers. These risks are again masked by the accounting rules. The first is sim-ply that one of the real strengths of a pension system, essentially unacknowledged by the GASB rules, is the flow of money into the system from its members and their employers. Depending on the financial market conditions, this flow is usually a comparable size to the investment returns or larger. The consistency of this contribution makes it at least as important in providing security to the system as investment earnings, if not more so. Removing this source of strength leaves a pension plan to rely on the financial markets alone, not a well-known source of security.

Beyond these concerns, there are investment constraints imposed by closing a plan. One of the advantages of managing investments in perpetuity is that managers are permitted to take the long view in everything. Earning the necessary 7.5 percent may be feasible when all investments can be long ones. A system without that long horizon has a serious disadvantage when it comes to the array of investments open to it. As a pension plan winds down, more and more of its fund must be kept in relatively liquid short-term investments. These seldom earn the higher returns of the longer-term assets. Therefore, the greater the proportion of the fund that must be kept in short-term investments, the less feasible it is to achieve the typical investment return targets. A 2011 analysis done by the California Public Employees’ Retirement System (CalPERS) showed that closing that system would give up $150–200 billion in investment returns over the course of winding the system down, about half the size of the overall fund.60

Sometimes, closing a plan is done with a POB. This is subject to all the risks of a POB, with the added risk of ending the employee payments into the fund. If a POB is a tightrope act, only successful when interest rate and market conditions are just right, closing a plan with a POB is a tightrope act without the net. The city of Woonsocket, Rhode Island, issued a $90 million POB to close its police and fire employee pension plan in 2002. A dozen years later, as reported in their 2014 financial report, after some poor investment years, the plan has assets of only $46.4 million, still has an unfunded liability of $42.2 million, only a single employee still paying into the system ($5,000 each year, compared with $8 million in expenses), and the city has a $79.4 million left of the bond debt to repay.61

Defined contribution pension plans: Not really a pension, or a solutionAnother common strategy to addressing an unfunded pension liability is to transition to a DC plan, or a hybrid pension plan with DB and DC components to it. The claim here is that a govern-ment can shed some of its risk by moving employees to that kind of pension. Certainly the infla-tion risk and the investment risk are with the employee under a DC plan. But converting from DB to DC is approximately the same thing as partially closing a plan, so the risks of closing a plan are not avoided, merely reduced. Beyond those risks, there are others in this approach.

To begin with, the administrative costs tend to be higher for a DC plan than for a DB plan. In the CalPERS analysis cited above, the authors point out that fees for their DB plan amount to about a quarter-percent per year, while a typical DC plan charges between one and two percent per year.62 Thus the investment returns must be that much higher to achieve the same standard of living for the retiree.

Large plans, the size of a state or big city, can usually negotiate substantially lower fees for a DC plan,63 sometimes only a few basis points higher than a comparable DB plan if the options for the employees are few and simple, but other investment constraints still offer cause for concern. Because the investments are being made on behalf of an individual with a finite lifetime rather than a perpetual group, more must be saved, and the returns will be lower.


Isolated from a group, an individual employee must save for his or her maximum expected life expectancy, not merely for the average, as with a DB plan. Paradoxically, this not only increases the risk of underfunding a retirement, but also increases the risk of overfunding as well. That is, it is harder to reach the appropriate target, so fewer people will succeed, and the appropriate target is too high, so only the luckiest of the successful—the ones who both save enough and live their maximum life expectancy—will not be wasting money. Everyone else who has successfully saved enough for their maximum life expectancy will have had earnings they were unable to enjoy while alive. Their heirs may not consider this a problem, but from the perspective of efficiency, it is a very real waste of resources. And these are only the lucky few; there will be substantial hard-ship among the many who have not saved enough.

The CalPERS analysis points out the other problem with investing for an individual, that where-as a traditional pension plan can be invested for a very long horizon, a DC plan must adjust its investments for the ages of its members. An older member needs a more conservative investment mix than a younger member—and a more liquid mix once the member retires—and this reduces the potential rate of return, increasing the risk that an individual’s savings will be inadequate to finance a comfortable retirement.

In addition, many pension plans incorporate some form of disability or survivor’s benefits. A government that must honor that commitment to its pension plan members will see an addition-al cost to provide the same benefit to its DC members.

Claims that money can be saved by converting from a pension plan to a DC savings plan usually rest on the argument that the costs of running a pension plan are underestimated, rather than on a denial of these transition costs.64 Other common arguments are that the assumed rate of return for a pension plan is unrealistic, or that the lowered investment returns from the shortened time horizon are inconsequential, if not negligible.65 As we have seen, the existing accounting rules tend to exaggerate the costs of a pension plan. It is that very exaggeration on which these arguments rely to make their point about the relative savings of closing a pension plan, or the advantage of DC plans.

Another argument is frequently made, that private industry has “moved away” from the un-sustainable costs of pension plans. This is a tendentious reading of recent history. Private de-fined-benefit pension plans across the country have been closed not because they were unsus-tainable, but because the executives of the corporations sponsoring them wanted to use that capital for different purposes and existing law provided a way for them to do so.66 The 1980s saw a blossoming of the private equity industry, devoted to buying companies and repurposing their financial assets to the benefit of the new owners. The decade also saw repeated raids on the pension funds of the newly purchased companies, raids that continue to this day. More recently, congressional changes to the law governing pensions and the Pension Benefit Guaranty Corpo-ration provided incentives for corporate managers to end these plans and turn the responsibility for their retirees over to the government, a managerial moral hazard. But even as of 2000, when the most recent wave of disassembly began, most remaining corporate pension funds were fully-funded or even overfunded. The stock market losses of 2000–2001 cut into that margin, but not significantly. It was the desire to repurpose that capital, combined with the federal guarantee and new rulings about bankruptcy, that made ending these plans desirable, and thus inevitable.67

A report from the National Institute on Retirement Security listed several reasons for the decline of DB pension plans, including decreasing union presence in the work force, the legal and regu-latory environment, and changes in corporate priorities, and added: “It is interesting to note that each of these reasons has little to do with the underlying economics of maintaining DB pension plans.”68

In truth, given that a DC plan does not envision a level of retirement income at all, it is perhaps arguably not even appropriate to label it a “pension” plan at all. In fact, the phrase was not used


until the 1970s. Until then, it would have been called a savings plan, or a stock purchase plan, if it had a name at all. In many ways, the rebranding of a simple savings plan as a “pension” plan is a triumph of marketing in service of a massive reappropriation of pension plan assets.

POTENTIAL SOLUTIONS: DIFFERENT ACCOUNTING FOR DIFFERENT QUESTIONSIf this article were merely a plea for less demanding pension funding requirements, that would be a difficult argument to sustain in the face of the facts. But though a central argument here is that 100 percent funding is hardly necessary to keep pension checks from bouncing, it is equally troubling that the abstract nature of pension funds also permits governments to ignore already existing funding requirements. The impacts on the current budget are always years away, making it easy to shave a little in the current year, which becomes a little more the following year, more the next, and so on. In a similar fashion, accounting rules that make clear a system’s strengths include the ongoing stream of payments from future employees would be preferable to rules that envision a system that might close tomorrow. Finding a form of funding and accounting prac-tice that encourages adequate funding is a valuable goal—and is precisely the goal sought by the GASB in their 2012 revision—but one where the current rules fall short due to the many problems outlined here.

Alternate accounting rules are possible. Unfunded liabilities and funding ratios are not the only kind of planning values available. For example, instead of estimating a funding ratio, the Social Security trustees predict the year in which that fund will run out of money, given current trends. If, each year, the date advances, the system is in good shape. If it does not, there is cause for concern. This is simply another way to do the accounting for a pension plan, with advantages and disadvantages compared to the traditional method. It is worse at allocating the cost to any indi-vidual plan member, but it is arguably much better at respecting the fundamental philosophy of aggregated security behind a traditional pension plan.69

A pay-as-you-go system is also a different method of accounting. Like the depletion-year method it has positive and negative features. It does a poor job of predicting how much money ought to be put aside for future years, the original reason this practice was largely abandoned. However, a pay-as-you-go system will provide instantaneous budgetary feedback to pension changes. That is, changes enacted to make a system more generous will be immediately reflected in increased costs to the current budget (rather than the pension system budget) and a reduction in payments to the system will immediately be reflected in pension checks bouncing. These are the sorts of effects that, in a practical sense, constrain the actions of politicians, unlike vague promises that a bond rating might be threatened, or that a big tax increase will be necessary a decade or two hence.

A hybrid system, combining the better aspects of pay-as-you-go and the GASB-approved systems would appear to be possible, perhaps by developing a formula that would keep a certain amount of the annual retiree payments for a pension system as part of the sponsoring government’s annual budget. This must be done not simply by recalculating the necessary contribution to the pension fund, but by arranging government finances so the pension checks to retirees will not clear unless such an appropriation is made. The linkage between policy and outcome must be im-mediate and clear in order to have an effect on policy. Under the status quo, the linkage is neither.

For example, a government could seek to “monetize” the stream of payments made by its employ-ees into the pension fund as a way to make clear that these payments are an important asset of the fund. A revenue bond backed by the premiums paid into the fund in future years could be bought annually from the government in exchange for a portion of the benefits paid to retirees in the current year. The bond itself would be an asset of the pension fund, offsetting its liability.


The bond could be rated or have some small piece of it sold to another party to establish its value. Under such an arrangement, the government, rather than the pension fund, would be respon-sible for issuing benefit checks. The fund would pay the government for the next year’s bond, retaining enough to maintain or increase its funding level. The government would use the funds received to pay retirees, along with its employer share premiums. The debt to the retirees is of the government, while the pension plan incurs a debt to that government.

Under such a system, an abrupt change in benefits would have a direct impact on the govern-ment’s current budget. Were a mayor to promise large pension increases to the firefighters, part of the first installment would be paid directly from the city budget the very next year. Similarly, a reduction in premium payments would affect the value of the bonds, with an impact on the balance sheet of both the pension fund and whatever other agency might hold them. If the state owned some of the bonds in its cash pool, a governor who cut pension payments would see the value of the bonds drop, creating a loss to the state’s own balance, as well as that of the pension fund. Unlike the current system, the consequences of these decisions would be immediate.

Such an accounting change could not remove political considerations from pension manage-ment. Whatever forces exist to keep the compensation of government employees from being cut—applied by organized labor or the job market—would remain, and a change in accounting could hardly do away with political pressure to hold down or cut budgets. But the consequences of policy decisions would be clearer and sooner, with much less opportunity for decision makers to shrug away responsibility for highly contingent events years hence.

There is much detail to be added here, and this suggestion is only one possibility. It is merely a sketch of a possible system, intended to demonstrate that there are other ways to configure the relationship between pension fund, retirees, employees, and the government that may create clearer and better feedback and therefore better incentives for the various parties.

ARGUMENTS AGAINSTThe movement of GASB in recent years, through Statements 67 and 68, have been in precisely the opposite direction to the arguments presented here. The arguments against the perspective presented here are thus well known, and worth airing and addressing.

Facing facts: On not kicking the can down the road “The problems that we have right now were caused in the largest part by delay-ing payments, kicking the can down the road,” said Sen. Daniel Biss, D-Evan-ston, a key pension negotiator. “It’s very dangerous. It should only be done if it’s paired with a very specific, clear plan for how all the payments will be made and the pension systems will be brought back to full payment.”70

In debates over pension funding, proponents of austerity, in favor of paying off the unfunded liability as quickly as possible will often accuse their opponents of simply wanting to “kick the can down the road” and avoid facing the hard truths. The Chicago Tribune article quoted above even uses the phrase in a sub-head. The argument made is that these problems must be addressed eventually so therefore a hard-nosed look at the facts will support addressing them now. People who suggest otherwise are irresponsible procrastinators who will eventually be forced to face the facts.


The metaphor itself is interesting, because kicking the can down the road is only a problem if there comes a point at which one can kick that can no further. As we saw above, under the right conditions, a pension system can pay benefits indefinitely at funding levels much lower than “full” funding. In other words, unless the combination of funding level and demographics creates a liquidity crisis, there is always room to “kick the can” further, and the metaphor misleads rather than enlightens. Obviously a system must be managed so as not to create a liquidity crisis, but this need not be done through 100 percent funding, as thousands of technically underfunded pension systems demonstrate each year.

It is true, however, that since the assumed rate of return in most systems is, at least in 2016, higher than the combination of inflation and economic growth, there is an advantage to putting money into the system sooner. One hundred dollars deposited now in an investment earning 7.5 percent interest will very likely be worth more ten years hence than one hundred dollars compounding at 3 percent inflation plus 2 percent economic growth. But the simple fact of this advantage does not then imply that all dollars not invested this way are wasted. Schools must be staffed and roads maintained; a government’s many commitments, opportunities, and responsi-bilities must be weighed against one another. Furthermore, many government expenses, such as infrastructure, public health, and education, can usefully be considered to be investments. Some of them will pay off at rates higher than a pension fund’s assumed rate of return. Even mainte-nance costs for some assets can be considered to be investments that pay off handsomely.71 There are few governments where one expense can be allowed to trump all others.

Budgeting: Full funding reduces volatility Another reason to value full funding is the consistency of employer payments. In a fully-fund-ed system, when investment returns vary, the resulting variation in necessary premiums (from employees or employer) is usually small enough to be unimportant. Because the payments into the system are larger, a partially funded system will see correspondingly larger variation in the required premiums from the same investment variations. Volatility is an important concern in public finance. Because the budgeting process is cumbersome, dramatic changes in expenses tend to become controversial, no matter their relative size.

As true as this is, one way to amplify the volatility of a partially funded system is to amortize the unfunded liability on a fixed schedule. As the fund progresses down the road to amortization, the effects of variation in investment returns become ever more extreme.

A plan with $4 billion of liabilities, funded at only 50 percent, is probably assuming a return of about $150 million in investment income, if it uses a typical 7.5 percent rate. Assume that, in one year, contributions to the system include $175 million devoted to paying down the unfunded liability on year five of a 30-year course. If investment returns fall $150 million short that year, so there is no investment income, there will be little progress on the unfunded liability, as most of the contributions must go to fund the pension checks. As a result, the unfunded liability will remain the same, and there will be one less year in which to amortize the same debt, by raising the required payment for the next year. (See table on page 27.)

After such a shortfall, in order to stay on schedule to pay off the debt, the following year’s amor-tization payment will have to be $188 million, an increase of 9 percent. This may not seem tragic, but if the government had been on year 20 of the same amortization schedule, suffering the same loss would cause a 12 percent payment increase the next year. Were it on year 25, with only 5 years to go before achieving full funding, the required increase in the amortization payment would be 21 percent. On year 26, the increase would be 26 percent, ramping up dramatically as the remaining time in the schedule declines. This is true for a pension fund that has stayed exact-ly on track on its amortization schedule until then. For a pension fund that has fallen behind at all, the increases at the end will be much larger, as the amortization of losses accumulates. Large increases will frequently be deemed politically infeasible and so many amortizations will not be completed successfully.


In the author’s experience, amortization schedules developed under these rules tend to seem reasonable only at the outset. A few years in, after being exposed to the vagaries of real-world investments, they usually call for utterly unmanageable increases in amortization payments at the very end of the schedule, often just in the last three or four years. This is an unremarkable outcome of the arithmetic of the reductions in the amortization period. The only remarkable part about it is when the government in question decides to “buckle down” and try harder rather than simply relax and restart the amortization schedule.

In other words, full funding is a worthy goal to prevent volatility in payments. But if, in order to prevent a problem one has to endure it—in an amplified form—sensible people might be led to question the therapy.

A better way to amortize shortfalls due to volatility is to amortize them on the same time scale as the system amortization. If a shortfall occurs in year 20 of a 30-year schedule, that shortfall is amortized over 30 years rather than the remaining 10. This will unavoidably make the amortiza-tion much longer, but 30 years is an arbitrary choice for a system being managed in perpetuity. Some pension systems (e.g. the State of Rhode Island employee pension system) have adopted this method of accounting, but it is far from universal.

Accounting clarity: What questions can you answer? Accounting systems are designed to make clearly available the answers to the important ques-tions one might ask about some enterprise. A business, for example, must be able to track wheth-er it is earning a profit. Accrual accounting was invented to make clear whether a business is solvent, or even profitable. It does not, however, give a very good picture of the cash position for a business. This is less important for a company that might make its routine purchases with a line of credit, for example, so accrual accounting is used in most businesses.

An entirely different set of considerations obtain for a typical household, which is usually more interested in simply not experiencing a liquidity crisis. For a household, cash accounting is more useful in a day-to-day sense, even if it does not give a perfect picture of a household’s solvency. The appropriate accounting system depends heavily on the important questions, and the context in which they are asked.

For a pension system, the question of solvency in perpetuity is the important question. It is vital to know whether a system will receive the assets it needs to pay the liabilities it owes, out into the indefinite future. By making the standards uniform, and closing some reporting loopholes, GASB 68 makes answering important questions about the funding situation of a particular plan simpler and easier.

Schedule years remaining

Unfunded Liability

Assumed Income

Actual Income

Amortization Payment

Amortization Next Year

% Increase

25 $1.67 billion $150 million $0 million $175 million $188 million 7.6%

15 $1.00 billion $150 million $0 million $175 million $192 million 9.7%

10 $667 million $150 million $0 million $175 million $197 million 12%



The effect of failing to meet investment targets becomes more significant as a pension fund advances on its amortization schedule.


Unfortunately, the questions this system answers are not the most pertinent. That is, the GASB rules do a good job of answering, “How much money will this plan need to pay off its debts if it is closed tomorrow?” But most plans are not going to be closed tomorrow, so this is usually not very useful information. A more useful question might be along the lines of “How are we doing?” or “How much volatility will we suffer as we go along paying our bills?” The GASB rules provide those answers only in an oblique manner, for those willing to read past the headline numbers, just as a household’s solvency must be ascertained by looking beyond their bank balance.

In other words, the clarity provided by the GASB rules comes at the expense of making the situ-ation seem much more dire than necessary. By ignoring, if not actually undermining, the value of the collective strength of a pension plan, the rules do a deep disservice to those who have contributed loyally to the plans for decades. The new rules will also require governments to add a deficit of billions of dollars to their governments’ bottom line, something few political leaders have the will to ignore.

As of this writing, governments across the country have only just begun to comply with the new rules, and as yet, no one really knows whether the bond-rating agencies will overreact or ignore the colossal new debts that are appearing on municipal bottom lines. Either way, it is not at all certain that clarity has been added to the situation.

CONCLUSIONCurrent trends are not promising for addressing the concerns described in this paper. Detroit’s immolation, for example, has produced only a limited backlash. Citizens are certainly angry about what has happened, but with so many possible targets for ire—the auto makers’ abandonment of the city, the corruption of its leaders, the exploitation by Wall Street—technical debates about accounting rules are unlikely to rise to the top of the list.

Certainly GASB itself sees no problem. Indeed, recent developments there, such as the adop-tion of Statements 67 and 68, have all been in the direction of making the situation somewhat worse. The people who made these changes happen are still at GASB. Their rhetoric forces them to continue to wave this flag and to defend it vigorously. Reform and the concomitant relief will not come from within GASB. Similarly, it is difficult to expect reform to come from Congress, or any other legislature. They are temperamentally unsuited to do anything but delegate technical problems such as this.

In a theoretical sense, it is quite possible for governments simply to ignore the GASB rules. They are, after all, only guidelines to accepted accounting practice. However, in a practical sense they are woven into a straitjacket that government must wear. Bond ratings and practical politics con-spire to force governments to comply with these rules and only the strongest leaders are likely to resist. Unions are somewhat conflicted. Though the threat of bankruptcy is beginning to awaken some to the dangers of the policies they have abetted, if not pursued, their concerns have largely been about the benefit side of the equation.

The accounting rules have been a convenient club to wield against public employees and their unions for those who would do so. The “obvious” poor state of the pension funds makes it easy to claim that the public has been duped into obligations it cannot afford, and this has been very useful in weakening the political position of these unions. In turn, many union leaders have participated in their own weakening by making unnecessary demands for full funding. In the author’s per-sonal experience, labor leaders frequently perceive concern about the cost of a pension plan to be equivalent to arguing to weaken the guarantee of their benefits. A suggestion of ways to make managing pension funds less expensive is therefore an attack on benefits.


Full funding is a valuable goal, but so is reducing poverty, keeping roads paved, educating chil-dren, fighting crime, and so on. Few government goals can be considered to trump all others, and diversion of taxes into pension savings is a choice not to fund something else.

In the end, accounting rules should be a guide to action, not a guide to perfection. It is one thing to know how far from the promised land of full funding a system is, but it is a different and more valuable thing to understand how to get there. When those rules guide us to action that makes things worse, leaving us farther from the target than before, what are we to think of the rules?

Ultimately the best argument against the current rules is that following these rules is not neces-sary to keep the checks from bouncing. “Full” funding of any pension system requires spending more money than necessary to meet the government’s obligations. Is that not the very definition of waste? The remaining question is whether our nation will drive thousands of municipalities into bankruptcy, deprive millions of public employees of pension benefits they have earned, and discredit one of the great financial advances of the last century—all in order to preserve this waste. To date, the answer is not comforting.

Funding Public Pensions • haasinstitute.berkeley.edu • P. 30

ENDNOTES

1. http://www.cityofchicago.org/city/en/depts/mayor/iframe/just_the_facts.html.

2. Fran Spielman. “Moody’s, citing pension crisis, downgrades Chicago’s debt to junk status.” In: Chicago Sun-Times (May 2015). 6/28/15. URL: http://chicago.suntimes.com/politics/7/71/600585/moodys-down-grades-chicagos-debt-junk-level.

3. Andrew Harris and Elizabeth Campbell. “Illinois Bid to Solve $111 Billion Pension Shortfall Is Dead.” In: Bloomberg Business (May 2015). URL: http://www.bloomberg.com/news/articles/2015-05-08/illi-nois-pension-reform-legislation-over-turned-by-state-court-i9fs4sid.

4. Ed Ring. Estimating America’s Total Un-funded State and Local Government Pension Liability. Tech. rep. 6/21/15. Tustin, CA: California Policy Center, Sept. 2014. URL: http://californiapolicycenter.org/estimating- americas-total-unfund-ed-state-and-local-government-pension-lia-bility/. F. John White makes a slightly lower estimate of $833 billion in Addressing the National Pension Crisis: It’s Not a Math Problem. 6/6/15. The PFM Group. Philadelphia, PA, 2013. URL: https://www.pfm.com/financial-advi-sory/retirement/resources/.

5. Saqib Bhatti. “Why Chicago Won’t Go Bankrupt—And Detroit Didn’t Have To.” In: In These Times (June 2015). 6/23/15. URL: http://inthesetimes.com/article/18096/a_scam_in_two_cities.

6. Liz Farmer. “Detroit’s Pension Is Actual-ly Well-Funded, So What’s All the Fuss?” In: Governing (Nov. 2013). 6/21/15. URL: http://www.governing.com/topics/finance/gov-detroits-pension-is-actual-ly-well- funded----so-whats-all-the-fuss.html ; Cate Long. “The real history of public pensions in bankruptcy.” In: Reuters.com, Muniland blog (2013). 6/25/15. URL: http://blogs.reuters.com/muniland/2013/08/08/the- real-histo-ry-of-public-pensions-in-bankruptcy/ .

7. Wallace Turbeville. The Detroit Bank-

ruptcy. Tech. rep. 6/24/15. New York, NY: Demos, Nov. 2013. URL: http://www.demos.org/publication/detroit-bankruptcy.

8. Gosia Wozniacka. “Stockton Bankrupt-cy: Mayor Says Chapter 9 Filing ’Very Likely’.” In: Associated Press (June 2012). 6/25/15. URL: http://www.huffingtonpost.com/2012/06/26/stockton-bankrupt-cy_n_1628575.html.

9. Mary Williams Walsh. “How Plan to Help City Pay Pensions Backfired.” In: New York Times (Sept. 2012). 6/21/15. URL: http://www.nytimes.com/2012/09/04/business/how-a-plan-to-help-stockton-calif-pay-pen-sions-backfired.html?_r=1.

10. Tom Sgouros. “We All Are Central Falls.” In: golocalprov.com (Aug. 2011). 6/25/15. URL: http://www.golocalprov.com/poli-tics/tom-sgouros-we-all-are-central-falls.

11. Michael De Angelis and Xiaowei Tian. “Until Debt Do Us Part: Subnational Debt, Insolvency, and Markets, Lili Liu and Ota-viano Canuto, eds.” In: 6/25/15. The World Bank, Feb. 2013. Chap. United States: Chapter 9 Municipal Bankruptcy— Utilization, Avoidance, and Impact, pp. 311–351. URL: http://elibrary.worldbank.org/doi/abs/10.1596/978-0-8213-9766-4.

12. For example, Daniel Borenstein. “Labor perpetuates pension myth that 80 percent funding goal is OK.” in: Contra Costa Times (Mar. 2015). 6/24/15. URL: http://www.contracostatimes.com/daniel-borenstein/ci_27700825/daniel-borenstein-labor-per-petuates-pension-myth-that-80.

13. Ed Ring, Estimating America’s Total Unfunded State and Local Government Pension Liability.

14. GASB itself is part of the Financial Accounting Foundation (FAF), as is the Financial Accounting Standards Board (FASB), which governs accounting for pri-vate corporations. The FAF, a private organization, is supported by fees established by the Dodd-Frank legislation, and by the sale of its publications. GASB rules are for state and local governments. Accounting standards for federal government departments and agencies in

http://www.cityofchicago.org/city/en/depts/mayor/iframe/just_the_facts.html

http://www.cityofchicago.org/city/en/depts/mayor/iframe/just_the_facts.html

http://chicago.suntimes.com/politics/7/71/600585/moodys-downgrades-chicagos-debt-junk-level



http://www.bloomberg.com/news/articles/2015-05-08/illinois-pension-reform-legislation-overturned-by-state-court-i9fs4sid




http://californiapolicycenter.org/estimating-%20americas-total-unfunded-state-and-local-government-pension-liability/




https://www.pfm.com/financial-advisory/retirement/resources/

https://www.pfm.com/financial-advisory/retirement/resources/

http://inthesetimes.com/article/18096/a_scam_in_two_cities

http://inthesetimes.com/article/18096/a_scam_in_two_cities

http://www.governing.com/topics/finance/gov-detroits-pension-is-actually-well-%20funded----so-whats-all-the-fuss.html




http://blogs.reuters.com/muniland/2013/08/08/the-%20real-history-of-public-pensions-in-bankruptcy/



http://www.demos.org/publication/detroit-bankruptcy

http://www.demos.org/publication/detroit-bankruptcy

http://www.huffingtonpost.com/2012/06/26/stockton-bankruptcy_n_1628575.html



http://www.nytimes.com/2012/09/04/business/how-a-plan-to-help-stockton-calif-pay-pensions-backfired.html?_r=1




http://www.golocalprov.com/politics/tom-sgouros-we-all-are-central-falls

http://www.golocalprov.com/politics/tom-sgouros-we-all-are-central-falls

http://elibrary.worldbank.org/doi/abs/10.1596/978-0-8213-9766-4

http://elibrary.worldbank.org/doi/abs/10.1596/978-0-8213-9766-4

http://www.contracostatimes.com/daniel-borenstein/ci_27700825/daniel-borenstein-labor-perpetuates-pension-myth-that-80





the United States are under the purview of the Federal Accounting Standards Advisory Board, a wholly different organization.

15. Robert Louis Clark, Lee A. Craig, and Jack W. Wilson. A History of Public Sector Pensions in the United States. p.170. 6/19/15. University of Pennsylvania Press, 2003. URL: http://www.pensionresearch-council.org/publications/0-8122-3714-5.php.

16. Ibid., see p.189.

17. Ibid., see p.175.

18. Though not for all. In Rhode Island, for example, the pension system for judges was only established in 1990 and there are still judges whose pensions are paid directly from the court budget (Dennis Hoyle. State of Rhode Island Employee Retirement System Report. Tech. rep. 6/15/15. State of Rhode Island Auditor General, June 2014. URL: http://www.oag.ri.gov/reports/Retire2014.pdf, p.21). The Indiana State Teachers’ Retirement Fund is funded in a similar fashion, pay-as-you-go for employees hired before 1995, and actuarially funded since (Jun Peng. State and Local Pension Fund Management. Boca Raton, FL: CRC Press, 2009, p.103).

19. Source: US Census Bureau State and Local Government Employee Retirement System Survey, as compiled by Jun Peng. State and Local Pension Fund Management. Boca Raton, FL: CRC Press, 2009.

20. Clark, Craig, and Wilson, A History of Public Sector Pensions in the United States.

21. In opposition to the DC plan, a tradition-al pension plan has come to be known as a “defined benefit” plan, where an employee’s retirement benefits are well-defined and the system’s managers must come up with a strategy to fund those benefits. A DC plan, by contrast, defines only the employee’s contributions to the plan, leaving benefits to the market results of the employee’s investment choices. Because of that, it is arguable whether such a system merits being called a “pension” system at all, see page 22.

22. These are Pension Obligation Bonds, discussed on page 22.

23. California Assembly Legislative Analyst’s Office, Addressing CalSTRS’ Long-Term Funding Needs. Tech. rep. 6/26/15. Sacramento, CA: Mar. 2013. URL: http://edsource.org/wp-content/uploads/CalSTRS-Funding-032013.pdf.

24. Clark, Craig, and Wilson, A History of Public Sector Pensions in the United States. make this same point (p.203), adding that police and fire pensions were often funded in a pay-as-you-go fashion in the same states and cities that were creating teacher and other employee pension systems funded on an actuarial basis.

25. The GASB Statements discussed in this review may be found at http://www.gasb.org. They are the following: 25, Nov. 1994, Financial Reporting for Defined Benefit Pension Plans and Note Disclosures for Defined Contribution Plans; 27, Nov. 1994, Accounting for Pensions by State and Local Governmental Employer; 43. Apr. 2004. Financial Reporting for Postemployment Benefit Plans other than Pension Plans; 45, Jun. 2004, Accounting and Financial Re-porting by Employers for Postemployment Benefit Plans other than Pension Plans; 67, June 2012, Financial Reporting for Pension Plans—An Amendment of GASB Statement No. 25; and 68, Jun. 2012, Accounting and Financial Reporting for Pension Plans—An Amendment of GASB Statement No. 27.

26. Marc Lifsher. “California pension funds are running dry.” In: Los Angeles Times (Nov. 2014). 6/27/15. URL: http://www.latimes.com/business/la-fi-controller-pen-sion-website-20141114-story.html.

27. Roger Lowenstein. “The Next Crisis: Public Pension Funds.” In: New York Times Magazine (June 2010). 6/28/2015. URL: http://www.nytimes.com/2010/06/27/magazine/27fob-wwln-t.html.

28. Eric Boehm. “Pennsylvania pension funds could run dry in as little as 10 years.” In: Daily Local News (Apr. 2015). 6/25/15. URL: http://www.dailylocal.com/gener-al-news/20150419/pennsylvania-pension-funds-could-run-dry-in-as-little-as-10-years.

29. GASB. Why Governmental Accounting And Financial Reporting Is—And Should Be—Different. 6/21/15. Governmental

http://www.pensionresearchcouncil.org/publications/0-8122-3714-5.php



http://www.oag.ri.gov/reports/Retire2014.pdf

http://www.oag.ri.gov/reports/Retire2014.pdf

http://edsource.org/wp-content/uploads/CalSTRS-Funding-032013.pdf

http://edsource.org/wp-content/uploads/CalSTRS-Funding-032013.pdf

http://www.gasb.org

http://www.gasb.org

http://www.latimes.com/business/la-fi-controller-pension-website-20141114-story.html



http://www.nytimes.com/2010/06/27/magazine/27fob-wwln-t.html

http://www.nytimes.com/2010/06/27/magazine/27fob-wwln-t.html

http://www.dailylocal.com/general-news/20150419/pennsylvania-pension-funds-could-run-dry-in-as-little-as-10-years




Accounting Standards Board. Norwalk, Connecticut, Nov. 2006, revised 2013. URL: http://www.gasb.org/jsp/GASB/Page/GASBSectionPage&cid=1176156741271.

30. For more about the difference between private and public entities, and the differences in risk, see Peng, State and Local Pension Fund Management, section 4.2.

31. Author’s calculations from (Social Securi-ty Trustees. Trustee’s Report. 6/23/15. Social Security Administration. 2014. URL: http://www.ssa.gov/oact/STATS/table4a1.html) The system was unable to sustain this low level of funding because of demographic changes. That is, the actuarial facts on the ground did change, as the baby boom worked its way through the system.

32. This is equation 7.5 of (Howard E. Winklevoss. Pension Mathematics with Numerical Illustrations. Second edition. Pension Research Council of the Wharton School of Business / University of Pennsylvania Press, 1993), chapter 7. Winklevoss was writing about private pension plans, so he subsequently points out that plans will seek full funding. The federal law governing those plans requires them to do so. The same restriction is not true of public systems, for the reasons noted here.

33. Borenstein, “Labor perpetuates pension myth that 80 percent funding goal is OK.”

34. Keith Brainard and Paul Zorn. The 80-percent threshold: Its source as a healthy or minimum funding level for public pension plans. Tech. rep. 6/28/15. Lexington, KY: National Association of State Retirement Administators (NASRA), Jan. 2012. URL: http://www.nasra.org/files/Top-ical%20Reports/Funding%20Policies/80_percent_funding_threshold.pdf.

35. Barbara D. Bovbjerg. State and Local Government Pension Plans: Current Structure and Funded Status. Testimony before the Joint Economic Committee. 6/19/15. Washington DC: United States Government Accountability Office (GAO), July 2008. URL: http://www.gao.gov/as-sets/130/120599.pdf.

36. Author calculations from 2014 Illinois financial statements.

37. See GASB 68 paragraphs 27 and 28.

38 .Alicia H. Munnell et al. The Funding of State and Local Pensions: 2012-2016. Tech. rep. SLP 32. 6/17/15. Center for Retirement Research at Boston College, July 2013. URL: http://crr.bc.edu/briefs/the-funding-of- state- and- local- pensions- 2012- 2016/. Note that the new requirements do not ap-ply to systems considered to be fully-funded, and some systems already use a low discount rate. Both causes will tend to lower the effect on the “average” system, even while the effect on any individual system may be cataclysmic.

39. John E. Cihota. Considerations in Struc-turing Estimated Liabilities. Tech. rep. FT-WP-15-003. Washington, DC: Office of the Inspector General, United State Post-al Service, Jan. 2015. URL: https://www.uspsoig.gov/sites/default/files/docu-ment-library-files/2015/ft-wp-15-003_0.pdf.

40. Interview with Ed Derman, deputy CEO of CalSTRS, June 6, 2014.

41. Further calculations from 2014 Illinois financial statements.

42. Author estimates from Bureau of Economic Analysis data.

43. This kind of argument is a staple of anti-tax argumentation, and it is remarkable how seldom it is deployed in this context.

44. Bovbjerg, State and Local Government Pension Plans: Current Structure and Funded Status.

45. The risk of a fully-funded, or overfunded, pension plan is not only the political risk of increased benefits and reduced contributions, but also that policy makers will perceive an opportunity to close the plan entirely. According to the GASB frame-work, this is a rational choice for a fully-funded plan. There may be a political cost to such a decision, but the accounting says there would be no financial cost. In reality, closing a plan substantially increases the risk to the taxpayers, and experience shows that few such decisions have turned out to be good ones. (See page 22.)

http://www.gasb.org/jsp/GASB/Page/GASBSectionPage&cid=1176156741271

http://www.gasb.org/jsp/GASB/Page/GASBSectionPage&cid=1176156741271

http://www.ssa.gov/oact/STATS/table4a1.html

http://www.ssa.gov/oact/STATS/table4a1.html

http://www.nasra.org/files/Topical%20Reports/Funding%20Policies/80_percent_funding_threshold.pdf



http://www.gao.gov/assets/130/120599.pdf

http://www.gao.gov/assets/130/120599.pdf

http://crr.bc.edu/briefs/the-funding-of- state- and- local- pensions- 2012- 2016/

http://crr.bc.edu/briefs/the-funding-of- state- and- local- pensions- 2012- 2016/

https://www.uspsoig.gov/sites/default/files/document-library-files/2015/ft-wp-15-003_0.pdf





46. Peng, State and Local Pension Fund Management, chapter 6.

47. California Assembly Legislative Analyst’s Office, Addressing CalSTRS’ Long-Term Funding Needs.

48. Ted Dabrowski, John Klingner, and Tait Jensen. CPS pensions: From retirement security to political slush fund. Tech. rep. Chicago and Springfield, IL: Illinois Policy Institute, Aug. 2015. URL: https://www.illinoispolicy.org/reports/cps-pensions-from-retirement-securi-ty-to-political-slush-fund/.

49. Peng, State and Local Pension Fund Management, p.22.

50. Dora L. Costa. The Evolution of Retire-ment: An American Economic History, 1880-1990. 6/26/15. University of Chicago Press, 1998. URL: http://www.nber.org/books/cost98-1.

51. Girard Miller. “Strategies To Consider As OPEB Costs Escalate.” In: Government Finance Review (Feb. 2011). 6/28/15, pp. 28–35. URL: http://www.gfoa.org/sites/de-fault/files/GFR_FEB_11_28.pdf.

52. Michael Ashton. TIPS, the Triple Duration, and the OPEB Liability: Hedging Medical Care Inflation in OPEB Plans. Tech. rep. 6/28/15. Society of Actuaries, May 2011. URL: http://ssrn.com/abstract=1838545.

53. Tom Sgouros. “The Manufactured OPEB Crisis.” In: RIFuture (Nov. 2013). 6/26/15. URL: http://www.rifuture.org/the-manu-factured-opeb-crisis.html.

54. For example: Peng, State and Local Pension Fund Management; Winklevoss, Pension Mathematics with Numerical Illustrations.

55. Alicia H. Munnell, Jean-Pierre Aubry, and Mark Cafarelli. An Update on Pension Obligation Bonds. Tech. rep. 6/20/15. Washington, DC: Center for State and Local Government Excellence, July 2014. URL: http://slge.org/publications/an-up-date-on-pension-obligation-bonds.

56. Ibid.

57. Government Finance Officers of Amer-

ica, Advisory, Pension Obligation Bonds, 6/26/16, http://www.gfoa.org/pension-ob-ligation-bonds. The advisory goes on to list five very good reasons to avoid POBs:

The invested proceeds might fail to earn more than the interest rate.

POBs commonly have a complex structure with derivatives embedded in them that introduce counterparty risk and credit risk.

POBs are taxable debt, which usually counts against a government’s overall debt burden for the ratings agencies.

The principal (re)payments may be missing, or structured over a longer term than the amortization period.

POBs are a red flag to ratings agencies, who generally take them as a sign of other trou-ble, potentially undisclosed.

58. Though this may have been mostly to do with the city’s ill-fated attempts to hedge the floating interest rate of the POBs it issued by swapping with banks for fixed-rate debt payments. See Farmer, “Detroit’s Pension Is Actually Well-Funded, So What’s All the Fuss?.”

59. Walsh, “How Plan to Help City Pay Pensions Backfired.”

60. CalPERS. The Impact of Closing the Defined Benefit Plan at CalPERS. 6/21/15. California Public Employees’ Retirement System (CalPERS). Mar. 2011. URL: http://www.calpers.ca.gov/eip-docs/closing-im-pact.pdf.

61. The plan has almost 60% of its assets in cash and fixed-income investments, so it takes a remarkable investment year for its assets to earn the 7.5% it needs to match its assumptions. In 2014, however, its assets did earn 8.28% gross (not counting investment fees). See the 2014 Woonsocket annual financial report at http://www.ci.woonsock-et.ri.us/FINancial%20Report%206-30-2014.pdf (6/25/16), pp.58-65. The annual contributions of the single remaining employee were noted in the 2013 report (http://www.municipalfinance.ri.gov/documents/data/audits/2013/Woonsocket_2013.pdf). See also Sandy

https://www.illinoispolicy.org/reports/cps-pensions-from-retirement-security-to-political-slush-fund/



http://www.nber.org/books/cost98-1

http://www.nber.org/books/cost98-1

http://www.gfoa.org/sites/default/files/GFR_FEB_11_28.pdf

http://www.gfoa.org/sites/default/files/GFR_FEB_11_28.pdf

http://ssrn.com/abstract=1838545

http://www.rifuture.org/the-manufactured-opeb-crisis.html

http://www.rifuture.org/the-manufactured-opeb-crisis.html

http://slge.org/publications/an-update-on-pension-obligation-bonds

http://slge.org/publications/an-update-on-pension-obligation-bonds

http://www.gfoa.org/pension-obligation-bonds

http://www.gfoa.org/pension-obligation-bonds

http://www.calpers.ca.gov/eip-docs/closing-impact.pdf



http://www.ci.woonsocket.ri.us/FINancial%20Report%206-30-2014.pdf



http://www.municipalfinance.ri.gov/documents/data/audits/2013/Woonsocket_2013.pdf




Seoane, “Despite concessions, Woonsock-et’s pension is forecast to run dry.” Valley Breeze, December 2, 2015. 6/26/16, URL: http://www.valleybreeze.com/2015-12-02/woonsocket-north-smithfield/despite-con-cessions-woonsocket-s-pension-fore-cast-run-dry

62. CalPERS, The Impact of Closing the Defined Benefit Plan at CalPERS.

63. Deloitte Consulting. Inside the Structure of Defined Contribution/401(k) Plan Fees, 2013: A study assessing the mechanics of the “all-in” fee. Tech. rep. 9/29/15. Washington, DC: Investment Company Institute, Aug. 2014. URL: http://www.ici.org/pdf/rpt_14_dc_401k_fee_study.pdf.

64. See, for example, Andrew G. Biggs, Josh McGee, and Michael Podgursky. Transition cost not a bar to pension reform. 6/21/15. Jan. 2014. URL: http://www.aei.org/publi-cation/transition-cost-not-a-bar-to-pension-reform/

65. Lance Christensen, Truong Bui, and Leonard Gilroy. Addressing Common Objections to Shifting from Defined- Benefit Pensions to Defined-Contribution Retirement Plans. 6/21/15. June 2014. URL: http://reason.org/news/show/pension-re-form-defined-contribution.

66. Fran Hawthorne. Pension Dumping: The Reasons, The Wreckage, The Stakes for Wall Street. New York: Bloomberg, 2008.

67. Ellen Schultz. Retirement Heist: How Companies Plunder and Profit from the Nest Eggs of American Workers. Portfolio/Pen-guin, 2011.

68. Ilana Boivie. Who Killed the Private Sector DB Plan? Issue Brief. 6/12/15. Washington, DC: National Institute on Retirement Security, Mar. 2011. URL: http://www.nirsonline.org/index.php?op-tion=com_content&task=view&id=607&Itemid=49.

69. In truth, there is no reason a system using GASB standards cannot also provide a depletion date estimate as well, beyond the fact that the GASB rules provide no guidance for calculating it and that ratings agencies and bond buyers are not looking for it.

70. Hal Dardick and Monique Garcia. “Emanuel’s pension plan: Relief on payments, casino to pay for it.” In: Chicago Tribune (May 2015). 6/21/15. URL: http://www.chicagotribune.com/news/ct-chicago-pension-payment-law-0527-20150526-sto-ry.html.

71. See, for example, Wei Lin Koo and Tracy Van Hoy, of Jones Lang Lasalle, a large Chicago real estate firm, who developed estimates for the return on investment for preventive maintenance of various real estate assets such as HVAC equipment. Their estimates are that simple maintenance expenses can be construed as investments whose return is over 500%: Determining the Economic Value of Preventive Maintenance, 6/24/16 http://www.pmmi.org/files/ms/certified/newsletters/preventivemainte-nance.pdf

http://www.valleybreeze.com/2015-12-02/woonsocket-north-smithfield/despite-concessions-woonsocket-s-pension-forecast-run-dry




http://www.ici.org/pdf/rpt_14_dc_401k_fee_study.pdf

http://www.ici.org/pdf/rpt_14_dc_401k_fee_study.pdf

http://www.aei.org/publication/transition-cost-not-a-bar-to-pension-reform/



http://reason.org/news/show/pension-reform-defined-contribution

http://reason.org/news/show/pension-reform-defined-contribution

http://www.nirsonline.org/index.php?option=com_content&task=view&id=607&Itemid=49



http://www.chicagotribune.com/news/ct-chicago-pension-payment-law-0527-20150526-story.html




http://www.pmmi.org/files/ms/certified/newsletters/preventivemaintenance.pdf




APPENDIX

Year Total Payments Employee Payments

Employer Payments

Investment Earn-ings

1942 119 65 54 36

1950 539 260 278 70

1960 1,819 802 1,017 398

1970 7,388 2,788 4,600 2,460

1975 13,604 4,488 9,116 5,294

1985 36,878 9,479 27,399 34,852

1986 39,185 10,586 28,599 48,965

1990 46,431 13,853 32,578 64,907

Table 1: Earnings from State and Local Pension Funds

Week Savings Payment from Savings

Payment from Other

Total Payment

Remain-ing Debt

Funding Ratio

0 $999.76 60.01%

1 $600.00 $11.76 $7.96 $19.72 981.00 59.96%

2 588.24 11.76 7.96 19.72 962.22 59.91%

3 576.48 11.76 7.96 19.72 943.43 59.86%

4 564.72 11.76 7.96 19.72 924.61 59.75%

5 552.96 11.76 7.96 19.72 905.78 59.69%

6 541.20 11.76 7.96 19.72 886.93 59.69%

7 529.44 11.76 7.96 19.72 868.07 59.64%

8 517.68 11.76 7.96 19.72 849.18 59.58%

9 505.92 11.76 7.96 19.72 830.28 59.52%

10 494.16 11.76 7.96 19.72 811.36 59.46%

15 435.36 11.76 7.96 19.72 716.48 59.12%

20 376.56 11.76 7.96 19.72 621.14 58.73%

25 317.76 11.76 7.96 19.72 525.34 58.25%

30 258.96 11.76 7.96 19.72 429.08 57.61%

35 200.16 11.76 7.96 19.72 332.36 55.69%

40 141.36 11.76 7.96 19.72 235.15 55.11%

45 82.56 11.76 7.96 19.72 137.51 51.49%

50 23.76 11.76 7.96 19.72 39.38 30.47%

51 12.00 11.76 7.96 19.72 19.70 1.22%

52 0.26 11.74 7.96 19.70 0.0 0.00%

Table 2: Amortizing a debt with both income and savings. The first column is the savings balance at the beginning of the week, and the last column is the present value of the debt remaining after that week’s payments. At no time does the funding ratio go above 60%—in fact it declines each week—and yet 100% of the debt is paid. The table assumes an annual discount rate of 5%.


haasinstitute.berkeley.edu

A fair distribution of economic power and resources is a necessary

component of an inclusive society.

http://berkeley.edu

http://berkeley.edu



February 2017 | NASRA ISSUE BRIEF: Public Pension Plan Investment Return Assumptions | Page 1

Figure 1: Public Pension Sources of Revenue, 1986-2015

Source: Compiled by NASRA based on U.S. Census Bureau

data

NASRA Issue Brief: Public Pension Plan Investment Return Assumptions

Updated February 2017

As of September 30, 2016, state and local government retirement systems held assets of $3.82 trillion.1 These assets are held in trust and invested to pre-fund the cost of pension benefits. The investment return on these assets matters, as investment earnings account for a majority of public pension financing. A shortfall in long-term expected investment earnings must be made up by higher contributions or reduced benefits. Funding a pension benefit requires the use of projections, known as actuarial assumptions, about future events. Actuarial assumptions fall into one of two broad categories: demographic and economic. Demographic assumptions are those pertaining to a pension plan’s membership, such as changes in the number of working and retired plan participants; when participants will retire, and how long they’ll live after they retire. Economic assumptions pertain to such factors as the rate of wage growth and the future expected investment return on the fund’s assets. As with other actuarial assumptions, projecting public pension fund investment returns requires a focus on the long-term. This brief discusses how investment return assumptions are established and evaluated, compares these assumptions with public funds’ actual investment experience , and the challenging investment environment public retirement systems currently face. Because investment earnings account for a majority of revenue for a typical public pension fund, the accuracy of the return assumption has a major effect on a plan’s finances and actuarial funding level. An investment return assumption that is set too low will overstate liabilities and costs, causing current taxpayers to be overcharged and future taxpayers to be undercharged. A rate set too high will understate liabilities, undercharging current taxpayers, at the expense of future taxpayers. An assumption that is significantly wrong in either direction will cause a misallocation of resources and unfairly distribute costs among generations of taxpayers. As shown in Figure 1, since 1985, public pension funds have accrued approximately $6.8 trillion in revenue, of which $4.3 trillion, or 63 percent, is from investment earnings. Employer contributions account for $1.7 trillion, or one-fourth of the total, and employee contributions total $805 billion, or 12 percent.2

1 Federal Reserve, Flow of Funds Accounts of the United States: Flows and Outstandings, Third Quarter 2016, Table L.120

2 US Census Bureau, Annual Survey of Public Pensions, State & Local Data


Figure 2: Annual change in contributions from prior year, corporate vs. public pensions

Most public retirement systems review their actuarial assumptions regularly, pursuant to state or local statute or system

policy. The entity responsible for setting the return assumption, as identified in Appendix B, typically works with one or

more professional actuaries, who follow guidelines set forth by the Actuarial Standards Board in Actuarial Standards of

Practice No. 27 (Selection of Economic Assumptions for Measuring Pension Obligations) (ASOP 27), which prescribes the

factors actuaries should consider in setting economic actuarial assumptions. ASOP 27 recommends that actuaries consider the context of the measurement they are making, as defined by such factors as the purpose of the

measurement, the length of time the measurement period is intended to cover, and the projected pattern of the plan’s

cash flows.

ASOP 27 also advises that actuarial assumptions be reasonable, defined in subsection 3.6 as being consistent with five

specified characteristics; and requires that actuaries consider relevant data, such as current and projected interest rates

and rates of inflation; historic and projected returns for individual asset classes; and historic returns of the fund itself.

For plans that remain open to new members, actuaries focus chiefly on a long investment horizon, i.e., 20 to 30 years, as

this is the length of a typical public pension plan’s funding period. One key purpose for relying on a long timeframe is to promote the key policy objectives of cost stability and predictability, and intergenerational equity among taxpayers.

The investment return assumption used by public pension plans typically contains two components: inflation and the

real rate of return. The sum of these components is the nominal return rate, which is the rate that is most often used

and cited. The system’s inflation assumption typically is applied also to other actuarial assumptions, such as the level of

wage growth and, where relevant, assumed rates of cost-of-living adjustments (COLAs). Achieving an investment return

approximately commensurate with the inflation rate normally is attainable by investing in securities, such as US Treasury

bonds, that are considered to be risk-free, i.e., that pay a guaranteed rate of return.

The second component of the investment return assumption is the real rate of return, which is the return on investment

after adjusting for inflation. The real rate of return is intended to reflect the return produced as a result of the risk taken

by investing the assets. Achieving a return higher than the risk-free rate requires taking some investment risk; for public

pension funds, this risk takes the form of investments in assets such as public and private equities and real estate, which

contain more risk than Treasury bonds.

Unlike public pension plans, corporate

plans are required by federal regulations to

make contributions on the basis of current interest rates. As Figure 2 shows, this

funding method results in plan costs that

can be volatile and uncertain, often

changing dramatically from one year to the

next. This volatility is due partly to

fluctuations in interest rates and has been

identified as a leading factor in the decision

among corporations to abandon their

pension plans. By contrast, by focusing on the long-term and relying on a stable

investment return assumption, public plans

experience less contribution volatility.


Figure 3: Median public pension annualized investment returns for period ended 12/31/2016

Figure 3 plots median public pension fund annualized

investment returns for a range of periods ended December 31,

2016. As the higher investment returns achieved in the 1980s

and the 1990s are replaced by lower returns in more recent

years, average annualized returns for longer periods, such as 20 and 25 years, have begun to decline gradually. The steep

market declines of 2000-02 and 2008-09 have imposed a

particularly negative effect for measurement periods that

incorporate those events.

In the wake of the 2008-09 decline in capital markets, and

Great Recession, global interest rates and inflation have

remained low by historic standards, due partly to so-called

quantitative easing of central banks in many industrialized economies, including the U.S. Now in their eighth year, these

low interest rates, along with low rates of projected global

economic growth, have led to reductions in projected returns

for most asset classes, which, in turn, have resulted in an

unprecedented number of reductions in the investment return

assumption used by public pension plans. This trend is illustrated by Figure 4, which plots the distribution of investment

return assumptions among a representative group of plans since 2001. Among the 127 plans measured, nearly three-

fourths have reduced their investment return assumption since fiscal year 2010, resulting in a decline in the average

return assumption from 7.91 percent to 7.52 percent. If projected returns continue to decline, investment return assumptions are likely to also to continue their

downward trend. Appendix A lists the

assumptions in use or adopted for future use by

the 127 plans in this dataset.

One challenging facet of setting the investment

return assumption that has emerged more

recently is a divergence between expected

returns over the near term, i.e., the next five to 10 years, and over the longer term, i.e., 20 to 30

years3. A growing number of investment return

projections are concluding that near-term

returns will be materially lower than both

historic norms as well as projected returns over

longer timeframes. Because many near-term

projections calculated recently are well below

the long-term assumption most plans are using,

some plans face the difficult choice of either maintaining a return assumption that is higher

than near-term expectations, or lowering their

return assumption to reflect near-term

expectations.

3 Horizon Actuarial Services, “Survey of Capital Market Assumptions, 2016 Edition (July 2016) p4

Figure 4: Change in Distribution of Public Pension Investment Return Assumptions, FY 01 to FY 18


If near-term rates indeed prove to be lower than historic norms, plans that maintain their long-term return assumption

are likely to experience a steady increase in unfunded pension liabilities and corresponding costs. Alternatively, plans

that reduce their assumption in the face of diminished near-term projections will experience an immediate increase

unfunded liabilities and required costs. As a rule of thumb, a 25 basis point reduction in the return assumption, such as

from 8.0 percent to 7.75 percent, will increase the cost of a plan that has a COLA, by three percent of pay (such as from 10 percent to 13 percent), and a plan that does not have a COLA, by two percent of pay.

Conclusion The investment return assumption is the single most consequential of all actuarial assumptions in terms of its effect on a pension plan’s finances. The sustained period of low interest rates since 2009 has caused many public pension plans to re-evaluate their long-term expected investment returns, leading to an uprecedented number of reductions in plan investment return assumptions. Absent other changes, a lower investment return assumption increases both the plan’s unfunded liabilities and cost. The process for evaluating a pension plan’s investment return assumption should include abundant input and feedback from professional experts and actuaries, and should reflect consideration of the factors prescribed in actuarial standards of practice.

See Also: Actuarial Standards of Practice No. 27, Actuarial Standards Board

The Liability Side of the Equation Revisited, Missouri SERS, September 2006

Contact: Keith Brainard, Research Director, [email protected]

Alex Brown, Research Manager, [email protected]

National Association of State Retirement Administrators

Figure 5: Distribution of investment return assumptions

http://www.actuarialstandardsboard.org/asops/selection-economic-assumptions-measuring-pension-obligations/

http://www.mosers.org/~/media/Files/Adobe_PDF/About_MOSERS/Board-Newsletters/Operations-Outlook/operations_outlook_September06.ashx

mailto:[email protected]

mailto:[email protected]

http://www.nasra.org/


Appendix A: Investment Return Assumption by Plan

(Figures reflect the nominal assumption in use, or announced for use, as of February 2017)

Plan Rate (%) Alabama ERS1 7.875

Alabama Teachers1 7.875

Alaska PERS 8.0

Alaska Teachers 8.0

Arizona Public Safety Personnel 7.40

Arizona SRS 8.0

Arkansas PERS 7.5

Arkansas Teachers 8.0

California PERF2 7.375

California Teachers3 7.250

Chicago Teachers 7.750

City of Austin ERS 7.50

Colorado Affiliated Local 7.50

Colorado Fire & Police Statewide 7.50

Colorado Municipal 7.25

Colorado School 7.25

Colorado State 7.25

Connecticut SERS 6.9

Connecticut Teachers 8.0

Contra Costa County 7.25

DC Police & Fire 6.5

DC Teachers 6.5

Delaware State Employees 7.2

Denver Employees 7.75

Denver Public Schools 7.25

Duluth Teachers 8.0

Fairfax County Schools 7.5

Florida RS 7.6

Georgia ERS 7.5

Georgia Teachers 7.5

Hawaii ERS 7.0

Houston Firefighters4 8.5

Idaho PERS 7.0

Illinois Municipal 7.50

Illinois SERS 7.25

Illinois Teachers 7.0

Illinois Universities 7.25

Indiana PERF 6.75

Indiana Teachers 6.75

Iowa PERS 7.50

Kansas PERS 7.75

Kentucky County 6.75

Kentucky ERS 6.75

Kentucky Teachers 7.50

LA County ERS 7.50

Louisiana Parochial Employees 7.0

Louisiana SERS5 7.70

Louisiana Teachers5 7.70

Maine Local 6.875

Maine State and Teacher 6.875

Maryland PERS 7.55

Maryland Teachers 7.55

Massachusetts SERS 7.50

Massachusetts Teachers 7.50

Michigan Municipal 7.75

Michigan Public Schools 8.0

Michigan SERS 8.0

Minnesota PERF 8.0

Minnesota State Employees 8.0

Minnesota Teachers6 8.40

Mississippi PERS 7.75

Missouri DOT and Highway Patrol 7.75

Missouri Local 7.25

Missouri PEERS 7.75

Missouri State Employees 7.65

Missouri Teachers 7.75

Montana PERS 7.75

Montana Teachers 7.75

Nebraska Schools 7.5

Nevada Police Officer and Firefighter 8.0

Nevada Regular Employees 8.0

New Hampshire Retirement System 7.25

New Jersey PERS 7.90

New Jersey Police & Fire 7.90

New Jersey Teachers 7.90

New Mexico PERA 7.25

New Mexico Teachers 7.75

New York City ERS 7.0

New York City Teachers 7.0

New York State Teachers 7.50

North Carolina Local Government 7.25

North Carolina Teachers and State Employees 7.25

North Dakota PERS 8.0

North Dakota Teachers 7.75

NY State & Local ERS 7.0

NY State & Local Police & Fire 7.0


Ohio PERS 8.0

Ohio Police & Fire 8.25

Ohio School Employees 7.50

Ohio Teachers 7.75

Oklahoma PERS 7.25

Oklahoma Teachers 7.50

Oregon PERS 7.50

Pennsylvania School Employees 7.25

Pennsylvania State ERS 7.50

Phoenix ERS 7.50

Rhode Island ERS 7.50

Rhode Island Municipal 7.50

San Diego County 7.50

San Francisco City & County 7.46

South Carolina Police 7.50

South Carolina RS 7.50

South Dakota PERS 6.50

St. Louis School Employees 8.0

St. Paul Teachers 8.0

Texas County & District 8.0

Texas ERS 8.0

Texas LECOS 8.0

Texas Municipal 6.75

Texas Teachers 8.0

TN Political Subdivisions 7.50

TN State and Teachers 7.50

Utah Noncontributory 7.20

Vermont State Employees 7.95

Vermont Teachers 7.90

Virginia Retirement System 7.00

Washington LEOFF Plan 17 7.70

Washington LEOFF Plan 2 7.50

Washington PERS 17 7.70

Washington PERS 2/37 7.70

Washington School Employees Plan 2/37 7.70

Washington Teachers Plan 17 7.70

Washington Teachers Plan 2/37 7.70

West Virginia PERS 7.50

West Virginia Teachers 7.50

Wisconsin Retirement System 7.20

Wyoming Public Employees 7.75

1. The Retirement Systems of Alabama is reducing its plans’ return assumptions from 8.0 percent to 7.75 percent over a two-

year period.

2. CalPERS is reducing its investment return assumption from 7.50 percent to 7.0 percent over three years. In February 2017

the CalPERS Board adopted a risk mitigation policy, effective beginning FY 2021, that calls for a reduction in the system’s

investment return assumption commensurate with the pension fund achieving a specified level of investment return.

Details are available online: https://www.calpers.ca.gov/docs/board-agendas/201702/financeadmin/item-9a-02.pdf.

3. CalSTRS is reducing its investment return assumption from 7.50 percent to 7.0 percent over two years.

4. A proposal to reform pension plans sponsored by the City of Houston includes a reduction to the investment return assumption of the Houston Firefighters plan from its current level of 8.5 percent to 7.0 percent. This lower rate is pending approval of other elements of this proposal by the Texas Legislature during its 2017 Regular Session.

5. The Louisiana State Employees’ Retirement System and Teachers’ Retirement System are reducing their investment return

assumption from 7.75 percent to 7.50 percent by 2021 in annual increments of 0.05 percent.

6. Legislation approved by the Minnesota Legislature in 2016 would have reduced the return assumption of the Teachers’

Retirement Association to 8.0 percent, but was vetoed by the governor for reasons extraneous to the assumption.

7. For all Washington State plans except LEOFF Plan 2, the assumed rate of return is being reduced gradually, from 8.0 percent to 7.50 percent, over a 10-year period.

https://www.calpers.ca.gov/docs/board-agendas/201702/financeadmin/item-9a-02.pdf


Appendix B: Entity Responsible for Setting Investment Return Assumption for Selected State Plans

State System Investment Return Assumption Set By

AK Alaska Public Employees Retirement System Alaska Retirement Management Board

AK Alaska Teachers Retirement System Alaska Retirement Management Board

AL Retirement Systems of Alabama Retirement board

AR Arkansas Public Employees Retirement System Retirement board

AR Arkansas Teachers Retirement System Retirement board

AZ Arizona Public Safety Personnel Retirement System Retirement board

AZ Arizona State Retirement System Retirement board

CA California Public Employees Retirement System Retirement board

CA California State Teachers Retirement System Retirement board

CO Colorado Public Employees Retirement Association Retirement board

CO Fire & Police Pension Association of Colorado Retirement board

CT Connecticut State Employees Retirement System State Employees Retirement Commission

CT Connecticut Teachers Retirement Board Retirement board

DC District of Columbia Retirement Board Retirement board

DE Delaware Public Employees Retirement System Retirement board

FL Florida Retirement System FRS Actuarial Assumption Estimating Conference1

GA Georgia Employees Retirement System Retirement board

GA Georgia Teachers Retirement System Retirement board

HI Hawaii Employees Retirement System Retirement board

IA Iowa Public Employees Retirement System IPERS Investment Board

ID Idaho Public Employees Retirement System Retirement board

IL Illinois State Universities Retirement System Retirement board

IL Illinois State Employees Retirement System Retirement board

IL Illinois Municipal Retirement Fund Retirement board

IL Illinois Teachers Retirement System Retirement board

IN Indiana Public Retirement System Retirement board

KS Kansas Public Employees Retirement System Retirement board

KY Kentucky Retirement Systems Retirement board

KY Kentucky Teachers Retirement System Retirement board

LA Louisiana State Employees Retirement System Retirement board

LA Louisiana Parochial Employees’ Retirement System Retirement board

LA Louisiana Teachers Retirement System Retirement board

MA Massachusetts State Employees Retirement System

Collaborative between the legislature, state treasurer,

governor, and the Massachusetts Public Employee

Retirement Administration Commission

MA Massachusetts Teachers Retirement Board

Collaborative between the legislature, state treasurer,

governor, and the Massachusetts Public Employee

Retirement Administration Commission

MD Maryland State Retirement and Pension System Retirement board

ME Maine Public Employees Retirement System Retirement board

MI Michigan Public School Employees Retirement System Retirement board

MI Michigan State Employees Retirement System Retirement board

MI Municipal Employees' Retirement System of Michigan Retirement board

MN Minnesota Public Employees Retirement Association Legislature

MN Minnesota State Retirement System Legislature

MN Minnesota Teachers Retirement Association Legislature

MO Missouri Local Government Employees Retirement System Retirement board


MO Missouri Public Schools Retirement System Retirement board

MO Missouri State Employees Retirement System Retirement board

MO MoDOT & Patrol Employees' Retirement System Retirement board

MS Mississippi Public Employees Retirement System Retirement board

MT Montana Public Employees Retirement Board Retirement board

MT Montana Teachers Retirement System Retirement board

NC North Carolina Retirement Systems Retirement board

ND North Dakota Public Employees Retirement System Retirement board

ND North Dakota Teachers Fund for Retirement Retirement board

NE Nebraska Public Employees Retirement System Retirement board

NH New Hampshire Retirement System Retirement board

NJ New Jersey Division of Pension and Benefits Retirement board and state treasurer

NM New Mexico Educational Retirement Board Retirement board

NM New Mexico Public Employees Retirement Association Retirement board

NV Nevada Public Employees Retirement System Retirement board

NY New York State & Local Retirement Systems State comptroller

NY New York State Teachers Retirement System Retirement board

OH Ohio Police and Fire Pension Fund Retirement board

OH Ohio Public Employees Retirement System Retirement board

OH Ohio School Employees Retirement System Retirement board

OH Ohio State Teachers Retirement System Retirement board

OK Oklahoma Public Employees Retirement System Retirement board

OK Oklahoma Teachers Retirement System Retirement board

OR Oregon Public Employees Retirement System Retirement board

PA Pennsylvania Public School Employees Retirement System Retirement board

PA Pennsylvania State Employees Retirement System Retirement board

RI Rhode Island Employees Retirement System Retirement board

SC South Carolina Retirement Systems Legislature

SD South Dakota Retirement System Retirement board

TN Tennessee Consolidated Retirement System Retirement board

TX Teacher Retirement System of Texas Retirement board

TX Texas County & District Retirement System Retirement board

TX Texas Employees Retirement System Retirement board

TX Texas Municipal Retirement System Retirement board

UT Utah Retirement Systems Retirement board

VA Virginia Retirement System Retirement board

VT Vermont State Employees Retirement System Retirement board

VT Vermont Teachers Retirement System Retirement board

WA Washington Department of Retirement Systems Legislature

WI Wisconsin Retirement System Retirement board

WV West Virginia Consolidated Public Retirement Board Retirement board

WY Wyoming Retirement System Retirement board

1. The Conference consists of staff from the Florida House, Senate, and Governor’s office

Atlanta Cleveland Los Angeles Miami Washington, D.C.

www.horizonactuarial.com

Published: July 2016 © Copyright 2016, Horizon Actuarial Services, LLC

Survey of Capital Market Assumptions

2016 Edition

Horizon Actuarial Services, LLC is proud to serve as the actuary to over 90 multiemployer defined benefit pension plans across the United States and across various industries. As actuary to these plans, we must develop assumptions regarding future investment returns on plan assets. We then use those assumptions as we determine the actuarial values of the benefits promised by these plans to their participants and beneficiaries, as well as to project plan funding and solvency levels years into the future.

At Horizon Actuarial, we are actuaries, not investment professionals. Therefore, when developing assumptions as to what returns a pension plan’s assets might be expected to earn in the future, we look to our colleagues in the investment advisory community. Each year, as part of this survey, we ask different investment firms to provide their “capital market assumptions” – their expectations for future risk and returns for different asset classes in which pension plans commonly invest. The information gathered from this survey can help answer the common question: “Are my plan’s investment return assumptions reasonable?”

Of course, there are many factors to consider when evaluating a plan’s investment return assumptions, such as its asset allocation and the maturity of its participant population. Any of these factors can make the expected return for one plan very different from others. Therefore, this report does not opine on the reasonableness of any one plan’s investment return assumptions. Nevertheless, we hope this report will be a useful resource for trustees, actuaries, and investment professionals alike.

Horizon Actuarial sincerely thanks the 35 investment advisors who participated in this survey.

http://www.horizonactuarial.com/

Survey of Capital Market Assumptions: 2016 Edition

2 of 15

Table of Contents

Introduction 1

Summary 2

Survey Participants

A listing of advisors participating in the survey 3

Investment Horizons

A summary of assumptions by investment horizon 3

Short-Term vs. Long-Term

A comparison of expected returns over shorter time horizons versus over longer horizons

4

Differing Opinions

The distribution of expected returns and volatilities by asset class

5

Changing Outlooks: Last Five Years

A look at how expected returns and volatility have changed over the past five years

6

Evaluating the Return Assumption

Evaluating expected returns for a hypothetical multiemployer pension plan, using the results from the 2016 survey

7

Comparison with Prior Surveys

Reviewing the expected returns for the same hypothetical pension plan, using survey results over the past few years

9

Glossary

Basic definitions for certain investment terms

10

Methodology

A high-level description of the methodologies used in compiling the results of the survey

10

Appendix

Supplemental exhibits showing the detail behind the expected returns for the hypothetical plan, a summary of the average assumptions from the 2016 survey, and ranges of expected returns for 10-year and 20-year horizons

11

Summary

Horizon Actuarial first conducted this survey in 2010, and it included 8 investment advisors. In 2012, we first published a report on the survey results, which included 17 advisors. The survey has expanded considerably over the past few years; this 2016 edition of the survey includes assumptions from 35 different investment firms.

In general, expected returns have come down in recent years. When we focus on the 11 advisors who participated in each of the last five surveys, we see that expected returns for equity investments generally decreased from 2012 to 2014, with some stabilization (or perhaps increases) in 2015 and 2016. We also see expected returns for fixed income investments have generally decreased over the past few years, though they did increase from 2015 to 2016. Expected volatilities have remained relatively flat in recent years.

As we have seen in prior surveys, expected returns are noticeably lower over the short term than over the long term. This trend is apparent when we focus on the 12 advisors who provided assumptions for both the short term (up to 10 years) and long term (20 years or more). The difference is more pronounced for equity and fixed income investments, but less so for alternative investments such as real estate.

For ongoing pension plans without solvency issues, we believe a horizon of 20 years or more is appropriate for evaluating the reasonableness of the long-term investment return assumption. A shorter horizon, such as 10 years, may be more appropriate for evaluating the return assumption for a plan that is very mature or has solvency issues. Furthermore, even for plans with long-term investment horizons, it is important to understand the potential impact of lower expected returns over the short term. Therefore, this survey evaluates return expectations over horizons of both 10 years and 20 years.

For illustration, this report also constructs an asset allocation for a hypothetical multiemployer pension plan and uses the results from the survey to develop expected returns for the plan. When compared to the 2015 edition of the survey, the expected returns for this 2016 edition were virtually the same over a 10-year horizon and slightly higher over a 20-year horizon. These differences were driven both by changes to individual firms’ expectations, as well as new firms being added to the survey.

If you have questions about how the results of this survey relate to your multiemployer plan, please contact your consultant at Horizon Actuarial or visit the “contact us” page on our website, www.horizonactuarial.com. For questions about the survey itself, please contact Jason Russell at [email protected].

Horizon Actuarial Services, LLC is an independent consulting firm specializing in providing actuarial and consulting services to

multiemployer benefit plans. Horizon Actuarial does not provide investment, legal, or tax advice. Please consult with your

investment advisor, legal counsel, or tax advisor for information specific to your plan’s investment, legal, or tax implications.

http://www.horizonactuarial.com/


3 of 15

Survey Participants

Exhibit 1 below lists the 35 investment advisors whose capital market assumptions are included in the 2016 survey. This report will not attribute specific assumptions to individual firms, which is a precondition of the survey.

Originally, this survey was exclusive to the multiemployer plan community; it included only assumptions from investment advisors to multiemployer pension plans. The survey has expanded over the years, and it now includes assumptions from investment advisors outside of the multiemployer plan community.

Of the 35 sets of capital market assumptions included in the 2016 edition of the survey, 25 were provided by investment advisors to multiemployer plans, 7 were obtained from published white papers, and 3 were provided by investment advisors who do not consult with multiemployer plans. The different types of firms participating in the survey are indicated below.

Exhibit 1

2016 Survey Participants

AJ Gallagher

Alan Biller

Aon Hewitt

The Atlanta Consulting Group

Bank of New York Mellon*

BlackRock*

Bogdahn Group

Callan Associates

CapTrust*

Ellwood Associates

Envestnet**

Goldman Sachs Asset Management

Graystone Consulting

Investment Performance Services, LLC (IPS)

J.P. Morgan Asset Management*

Marco Consulting Group

Marquette Associates

Meketa Investment Group

Merrill Lynch Global Institutional Consulting

Morgan Stanley Wealth Management

New England Pension Consultants (NEPC)

Pavilion Advisory Group**

Pension Consulting Alliance

The PFM Group

RVK

Segal Rogerscasey

SEI

Sellwood Consulting

Summit Strategies Group

SunTrust Investment Advisory Group*

UBS

Verus

Voya Investment Management*

Wells Fargo Investment Institute*

Willis Towers Watson**

* Assumptions obtained from published white paper

** Advisor from outside multiemployer community

Investment Horizons

When evaluating the expected return assumption for an active, ongoing multiemployer pension plan, the plan actuary will usually consider investment returns over a long-term investment horizon of 20 years or more. A shorter time horizon, say over the next 10 years, may be more appropriate when evaluating the return assumption for a very mature plan that has unusually high negative cash flows or is projected to become insolvent.

It is also important to understand the sensitivity of plan funding to changes in future investment returns. For example, the actuary for an active, ongoing pension plan will typically set the plan’s investment return assumption based on expectations over a long-term horizon. However, it would still be instructive for the actuary to evaluate the sensitivity of funding results to the extent that short-term investment returns are expected to be higher or lower than the long-term assumption.

Survey participants were requested to provide their most recent capital market assumptions: expected returns for different asset classes, standard deviations for those expected returns, and a correlation matrix. The survey participants were also requested to indicate the investment horizon(s) to which their assumptions apply. If the participant develops separate assumptions for different time horizons, they were requested to provide each set of assumptions.

In the 2016 edition of the survey, 23 advisors provided one set of assumptions: of those, 21 specified a time horizon of 10 years and 2 specified a time horizon of 10 to 15 years. The remaining 12 advisors provided assumptions over both shorter-term (5 to 10 years) and longer-term (20 years or more) horizons.

Exhibit 2 below summarizes the time horizons specified by each advisor, grouped by type. Note that of the 12 advisors who provided both short-term and long-term assumptions, 11 of them are advisors to multiemployer pension plans.

Exhibit 2

Investment Time Horizons

Advisor Type

10 Years

10 to 15 Years

Both Short and Long-Term

Total

(A)

13

1

_11_

25

(B)

6

1

_-_

7

(C)

2

-

_1_

3

Total

21

2

_12_

35

(A) Multiemployer plan investment advisor (B) Published white paper (C) Advisor from outside multiemployer community


4 of 15

Short-Term vs. Long-Term

As noted in the previous section, survey participants provided expected returns over different time horizons. Given current market conditions, many investment advisors may expect returns for certain asset classes to be different in the short term than over the long term.

For comparability, this survey groups expected returns into two time horizons: 10 years and 20 years. We often refer to the 10-year expected returns as “short-term” and the 20-year expected returns as “long-term.” (Note however that many investment firms consider 10-year expectations to be “long-term.”)

When comparing the expected returns for the 12 advisors who provided both short-term and long-term assumptions,1 we see some interesting differences. See Exhibit 3 below. Expected returns are geometric and are generally considered to be indexed and net of fees.

Exhibit 3

The consensus among these 12 advisors was that returns are expected to be lower in the short term compared to the long term. In general, the difference between long

1 In cases where an advisor indicated a time horizon shorter than 10 years (for example, 5 or 7 years), the shorter-term expected returns

were combined with the longer-term expected returns to achieve a 10-year horizon. Similarly, if an advisor indicated a time horizon longer than 20 years (for example, 30 years), the longer-term expected returns were combined with the shorter-term expected returns to achieve a 20-year horizon.

term and short term returns is more pronounced for US equity and fixed income investments.

As noted earlier, the results shown in Exhibit 3 are based on a subset of 12 advisors. If we include all 35 survey advisors, the short-term and long-term expected returns do not change dramatically. See Exhibit 4 below.

Exhibit 4

The 10-year expected returns shown above include assumptions from all 35 advisors, while the 20-year expected returns include assumptions from only the 12 advisors who provided longer-term assumptions.

Prior year survey reports showed “blended” assumptions, which included 20-year expected returns from the advisors who provided longer-term assumptions and 10-year expected returns from the others. A significant drawback to using the blended assumptions is that they are not internally consistent and comparable with respect to investment horizons. Therefore, this 2016 survey no longer presents blended assumptions. It instead focuses only on the different expected returns over 10-year and 20-year investment horizons.

10-Year 20-Year

Asset Class Horizon Horizon Difference

US Equity - Large Cap 6.85% 7.89% 1.04%

US Equity - Small/Mid Cap 7.07% 8.23% 1.16%

Non-US Equity - Developed 7.33% 8.02% 0.69%

Non-US Equity - Emerging 8.62% 9.11% 0.50%

US Corporate Bonds - Core 3.41% 4.58% 1.17%

US Corporate Bonds - Long Dur. 3.72% 4.87% 1.15%

US Corporate Bonds - High Yield 6.18% 6.81% 0.64%

Non-US Debt - Developed 2.52% 3.70% 1.18%

Non-US Debt - Emerging 5.80% 6.43% 0.63%

US Treasuries (Cash Equivalents) 2.19% 3.15% 0.96%

TIPS (Inflation-Protected) 3.12% 3.94% 0.82%

Real Estate 6.52% 6.75% 0.23%

Hedge Funds 5.56% 6.16% 0.60%

Commodities 4.22% 4.84% 0.62%

Infrastructure 6.43% 7.12% 0.69%

Private Equity 9.45% 10.33% 0.89%

Inflation 2.22% 2.31% 0.08%

The 10-year and 20-year returns shown above are the averages for the 12

advisors who provided both short-term and long-term assumptions.

Expected returns are annualized (geometric).

Average Expected Returns: Short-Term vs. Long-Term

Subset of 12 Survey Respondents

10-Year 20-Year

Asset Class Horizon Horizon Difference

US Equity - Large Cap 6.64% 7.89% 1.25%

US Equity - Small/Mid Cap 7.00% 8.23% 1.23%

Non-US Equity - Developed 7.12% 8.02% 0.90%

Non-US Equity - Emerging 8.48% 9.11% 0.64%

US Corporate Bonds - Core 3.41% 4.58% 1.17%

US Corporate Bonds - Long Dur. 3.82% 4.87% 1.05%

US Corporate Bonds - High Yield 5.90% 6.81% 0.92%

Non-US Debt - Developed 2.43% 3.70% 1.27%

Non-US Debt - Emerging 5.77% 6.43% 0.66%

US Treasuries (Cash Equivalents) 2.14% 3.15% 1.01%

TIPS (Inflation-Protected) 2.80% 3.94% 1.14%

Real Estate 6.36% 6.75% 0.38%

Hedge Funds 5.41% 6.16% 0.75%

Commodities 3.98% 4.84% 0.86%

Infrastructure 6.59% 7.12% 0.53%

Private Equity 9.22% 10.33% 1.11%

Inflation 2.16% 2.31% 0.15%

Expected returns are annualized (geometric).

Average Expected Returns: Short-Term vs. Long-Term

All Survey Respondents

20-year horizon results include a subset of 12 survey respondents.

10-year horizon results include all 35 survey respondents.


5 of 15

Exhibit 5

Differing Opinions

Exhibit 5 below shows the distribution of expected returns and standard deviations (i.e., volatilities) for each asset class in the survey, as provided by the 35 individual advisors in the survey. Expected returns are geometric and apply to a 10-year investment horizon.2 Average assumptions from the 2016 survey are listed in brackets for each asset class. As noted earlier, returns are assumed to be indexed and net of fees.

The exhibit below shows that there are often significant differences in expected returns and standard deviations between investment advisors. As the saying goes, “reasonable people may differ.”

2 The above exhibit focuses on a 10-year horizon in order to include assumptions from all 35 advisors. See Exhibit 15 in the appendix to

this report for the assumptions over a 20-year horizon, based on the 12 advisors who provided longer-term assumptions. Also note that the exhibit above considers both expected returns and standard deviations. Prior editions of this survey showed exhibits that focused only on the ranges of expected returns, without considering the standard deviations that accompanied them. The ranges of expected returns by asset class can be found in the appendix as Exhibits 16 and 17.

The differences in assumptions are more pronounced for alternative investments such as real estate, hedge funds, and private equity. A contributing factor may be differences in the underlying strategies different advisors apply to these alternative investments (for example, opportunistic versus defensive). To contrast, the differences in expected returns and volatilities are smaller for more traditional investments, such as US equity and US fixed income.

A summary of the average survey assumptions can be found in the appendix to this report as Exhibit 14. This summary includes expected returns, standard deviations, and a correlation matrix.

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

10%

11%

12%

13%

0% 5% 10% 15% 20% 25% 30% 35%

Exp

ecte

d R

etu

rn

Standard Deviation

US Equity - Large Cap [ 6.6% | 16.9% ]

US Equity - Small/Mid Cap [ 7.0% | 21.0% ]

Non-US Equity - Developed [ 7.1% | 19.5% ]

Non-US Equity - Emerging [ 8.5% | 26.3% ]

US Corporate Bonds - Core [ 3.4% | 6.0% ]

US Corporate Bonds - Long Duration [ 3.8% | 10.5% ]

US Corporate Bonds - High Yield [ 5.9% | 11.0% ]

Non-US Debt - Developed [ 2.4% | 7.6% ]

Non-US Debt - Emerging [ 5.8% | 11.6% ]

US Treasuries (Cash Equivalents) [ 2.1% | 2.8% ]

TIPS (Inflation-Protected) [ 2.8% | 6.5% ]

Real Estate [ 6.4% | 14.7% ]

Hedge Funds [ 5.4% | 8.4% ]

Commodities [ 4.0% | 18.5% ]

Infrastructure [ 6.6% | 13.8% ]

Private Equity [ 9.2% | 23.1% ]

Alt

ern

ativ

esFi

xed

Inco

me

Equ

itie

s

2016 Survey: Distribution of Expected Returns and Standard Deviations

Asset Class [ Avg. Exp. Return | Avg. Std. Dev. ]

10-Year Horizon | Geometric Returns

SOURCE: Horizon Actuarial 2016 Survey of Capital Market Assumptions

Expected returns over a 10-year horizon include all 35 survey participants.


6 of 15

Changing Outlooks: Last Five Years

In recent years, there has been much discussion about whether it is reasonable to expect future investment returns to be as high as they have been historically. Following the financial market collapse of 2008, economic uncertainty and historically low interest rates painted a gloomy outlook for future investment returns. Even with improving economic conditions, interest rates remain low, and concerns about future investment returns remain.

Exhibit 6 below shows average expected returns for selected asset classes each year from 2012 to 2016. For consistency, this exhibit includes only the 11 advisors who participated in the survey in each of these years.

Note that the expected returns shown below are based on a 20-year horizon for advisors who provided longer-term assumptions and a 10-year horizon for others.3 For that reason (as well as the fact that we are dealing with a subset of advisors), the expected returns shown below are not directly comparable with those in other sections.

Exhibit 6

For this subset of advisors, average expected returns for equity-type investments such as US large cap equity, non-US developed equity, and private equity have generally decreased over the last five years, though they have stayed relatively flat (or perhaps increased) since 2014.

3 Of the 12 survey advisors who provided both shorter-term and longer-term assumptions, 10 of them indicated no difference in the

standard deviations of the expected returns over the short term versus the long term. For the other 2 advisors, the differences between short-term and long-term standard deviations were very minor.

4 The increase in average standard deviation for hedge funds from 2015 to 2016 is due mostly to changes in the subgroup: in 2016, one advisor stopped providing assumptions for hedge funds, while another provided assumptions for hedge funds for the first time.

For fixed income-type investments like US core bonds and US Treasuries, expected returns increased from 2013 to 2014, perhaps in anticipation of rising interest rates. Since then, expected returns declined or remained flat.

In addition to expected returns, it is also important to consider expected volatility of the returns, measured by standard deviations. Average standard deviations over the last five years are shown in Exhibit 7 below.

Exhibit 7

Overall, average standard deviations have stayed relatively flat from 2012 to 2016. This may imply that, on average, these 11 advisors expect the financial markets to remain volatile, but they do not necessarily expect volatility to increase.

In general, from one year to the next, these 11 advisors made relatively small adjustments (or no adjustment at all) to their standard deviations. Notable exceptions are US large cap equity and hedge funds, for which a few advisors adjusted their standard deviations significantly downward from 2013 to 2015.4 Other exceptions include US core bonds and US Treasuries, for which a few advisors adjusted their standard deviations upward over the past two years.

2%3%4%5%6%7%8%9%

10%11%

2012 2013 2014 2015 2016

Private Equity 10.6% 10.6% 10.0% 10.0% 10.2%

US Eq. (Large Cap) 8.9% 8.8% 8.2% 8.2% 8.1%

Non-US Eq. (Dev) 8.5% 8.3% 8.0% 8.0% 8.0%

Real Estate 7.0% 6.9% 6.7% 6.6% 6.6%

Hedge Funds 7.6% 7.0% 6.7% 6.0% 6.1%

US Bonds (HY) 7.0% 6.4% 6.5% 6.6% 6.8%

US Bonds (Core) 4.8% 3.8% 4.2% 4.1% 4.5%

US Treasuries 3.1% 2.4% 2.8% 2.6% 2.9%

Figures are average geometric returns for selected asset classes for the 11 advisors who participated in each of the surveys from 2012 through 2016.

Average Expected Returns: 2012 - 2016

0%

5%

10%

15%

20%

25%

30%

2012 2013 2014 2015 2016

Private Equity 24.2% 24.7% 24.6% 23.6% 23.9%

US Eq. (Large Cap) 22.9% 22.6% 20.8% 20.5% 20.5%

Non-US Eq. (Dev) 19.9% 19.7% 19.3% 19.6% 19.3%

Real Estate 13.0% 13.1% 13.6% 13.4% 13.8%

Hedge Funds 9.1% 9.5% 9.7% 8.1% 8.7%

US Bonds (HY) 11.4% 11.2% 10.6% 10.7% 10.3%

US Bonds (Core) 5.0% 4.9% 5.0% 5.9% 5.9%

US Treasuries 1.5% 1.5% 1.3% 1.9% 1.9%

Figures are average standard deviations for selected asset classes for the 11 advisors who participated in each of the surveys from 2012 through 2016.

Average Standard Deviations: 2012 - 2016


7 of 15

Evaluating the Return Assumption

Multiemployer pension plans are usually invested in a well-diversified mix of stocks, bonds, real estate, and alternative investments structured to maximize returns over the long term while minimizing return volatility.

The actuary of a multiemployer pension plan must evaluate the plan’s asset allocation and, based on expectations of future returns, develop an assumption for what plan assets are projected to earn over the long term. This assumption is then used (along with others) to determine the actuarial present value of the benefits promised by the plan to its participants and beneficiaries.

The actuary will often rely on the future return expectations of the plan’s investment advisor in developing the plan’s investment return assumption. However, as noted earlier, different investment advisors often have widely differing opinions on what future returns will be. Therefore, it can be beneficial to keep in mind other advisors’ expectations when setting the investment return assumption.

In the following exhibits, we will evaluate the investment return assumption for a hypothetical multiemployer pension plan. Exhibit 8 below shows the asset allocation for this hypothetical plan. The asset allocations are arbitrary, except for the fact that we made sure to include at least a small allocation to every asset class in the survey.

Exhibit 8

The following exhibits show expected annualized (geometric) returns for the hypothetical multiemployer pension plan over two different investment horizons.

Exhibit 9 evaluates the return expectations for the hypothetical plan over a 10-year horizon. These results may be appropriate for modeling sensitivities of future funding results to short-term investment returns, or for evaluating the return assumption for a very mature plan with severely negative cash flows.

Exhibit 10 evaluates the expected returns over a 20-year horizon based on assumptions from the 12 advisors who provided longer-term assumptions. These results may be more appropriate for evaluating the return assumption for an ongoing plan with no projected solvency issues.

Exhibit 9

Exhibit 10

Hypothetical Multiemployer Plan

Asset Class Weight

US Equity - Large Cap 20.0%

US Equity - Small/Mid Cap 10.0%

Non-US Equity - Developed 7.5%

Non-US Equity - Emerging 5.0%

US Corporate Bonds - Core 7.5%

US Corporate Bonds - Long Duration 2.5%

US Corporate Bonds - High Yield 5.0%

Non-US Debt - Developed 5.0%

Non-US Debt - Emerging 2.5%

US Treasuries (Cash Equivalents) 5.0%

TIPS (Inflation-Protected) 5.0%

Real Estate 10.0%

Hedge Funds 5.0%

Commodities 2.5%

Infrastructure 2.5%

Private Equity 5.0%

TOTAL PORTFOLIO 100.0%

2.5%3.5%4.5%5.5%6.5%7.5%8.5%9.5%

10.5%

ConservativeAdvisor

SurveyAverage

OptimisticAdvisor

7.34% 8.73% 9.98%

3.08% 4.11% 5.26%

- - - - - - - - - - - - - - -

- - - - - - - - - - - - - - -

18.8% 32.3% 45.7%

23.4% 37.7% 51.4%

28.5% 43.3% 57.0%

75th percentile

25th percentile

Annualized Expected ReturnsHypothetical Multiemployer Pension Fund

8.00% per Year

10-Year Horizon

7.50% per Year

7.00% per Year

Probability of Meeting or Exceeding:

2.5%3.5%4.5%5.5%6.5%7.5%8.5%9.5%

10.5%

ConservativeAdvisor

SurveyAverage

OptimisticAdvisor

8.01% 9.08% 10.36%

4.72% 5.78% 6.70%

- - - - - - - - - - - - - - -

- - - - - - - - - - - - - - -

25.1% 40.8% 57.8%

32.0% 48.8% 64.8%

39.7% 57.0% 71.4%

75th percentile

25th percentile

Annualized Expected ReturnsHypothetical Multiemployer Pension Fund

8.00% per Year

20-Year Horizon

7.50% per Year

7.00% per Year

Probability of Meeting or Exceeding:


8 of 15

It is important to keep in mind that the expected returns shown in Exhibits 9 and 10 apply only to the hypothetical asset allocation shown in Exhibit 8. The expected returns will be different – perhaps significantly – for different asset allocations.

Exhibit 13 in the appendix to this report shows more detail regarding the derivation of the expected returns for this hypothetical pension plan.

The following are points to consider when reviewing the results in Exhibits 9 and 10:

Reasonable Range: When setting the investment return assumption for pension valuations, actuaries traditionally constructed a “reasonable range” of assumptions and then selected a best-estimate point within that range. Actuaries would often consider the reasonable range to be the middle 50 percent of possible results, bounded by the 25th and 75th percentiles.

The applicable actuarial standards of practice were updated in 2013, and the new standards de-emphasize use of the reasonable range when setting the investment return assumption. Nevertheless, considering this range remains instructive; it may be difficult for an actuary to justify an assumption outside of this range.

Based on the average assumptions in this 2016 survey, the middle 50 percent range for this hypothetical pension plan is very wide: 5.78% to 9.08% over the next 20 years. Note that the reasonable range is even wider for a 10-year horizon: 4.11% to 8.73%. This is due to the fact that, while returns may be volatile from one year to the next, deviations will be lower when returns are annualized (in other words, smoothed out) over longer horizons.

Probability of Meeting/Exceeding the Benchmark: For example, say that the actuary for this hypothetical pension plan expects its investment returns to be 7.50% per year, represented by the gold lines in Exhibits 9 and 10. Based on the average assumptions in the 2016 survey, there is a 48.8% probability the plan will meet or beat its 7.50% benchmark on an annualized basis over a 20-year period. The probability is lower, 37.7%, that the plan will meet or beat its benchmark over the next 10 years.

Also note that over a 20-year period, the probability that the annualized investment return will exceed 8.00% (arbitrarily, 50 basis points above the benchmark return) is 40.8%. The probability that the annualized return will exceed 7.00% (50 basis points below the benchmark) is 57.0%. These probabilities are a bit lower when focusing on a 10-year horizon rather than a 20-year horizon.

Optimistic and Conservative Assumptions: As previously noted, different investment advisors have sometimes widely varying future capital market expectations. Therefore, it may also be interesting to consider the range of expected returns based on the assumptions provided by the most conservative and most optimistic advisors in the survey.

For this hypothetical asset allocation, the assumptions from the most conservative advisor indicate that the probability of beating the 7.50% benchmark assumption over the next 20 years is 32.0%. However, using assumptions from the most optimistic advisor results in a probability of 64.8%. Again, reasonable people may differ.

Limitations: The following are some important limiting factors to keep in mind when reviewing these results:

The asset classes in this survey do not always align perfectly with the asset classes provided by the investment advisors. Adjustments were made to standardize the different asset classes provided by each of the advisors.

Many of the advisors develop their future assumptions based on investment horizons of no more than 10 years, and some returns are generally expected to be lower in the short term. The typical multiemployer pension plan will have an investment horizon that is much longer than 10 years.

The return expectations included in the survey are based on indexed returns. In other words, they do not reflect any additional returns that may be earned due to active asset managers outperforming the market (“alpha”), net of investment expenses.

The return expectations do not adjust for plan size. Specifically, they do not take into account the fact that certain investment opportunities are more readily available to larger plans, as well as the fact that larger plans may often receive more favorable investment fee arrangements than smaller plans.

The ranges of expected annualized returns were constructed using basic, often simplified, formulas and methodologies. More sophisticated investment models – which may consider various economic scenarios, non-normal distributions, etc. – could produce significantly different results.

In most cases, adjustments made to account for these limitations tended to slightly lower the expected returns in the survey, for the sake of conservatism.


9 of 15

Comparison with Prior Surveys

Exhibits 6 and 7 showed how expected returns and standard deviations for certain asset classes have changed over the past few years. Similarly, Exhibits 11 and 12 below show how return expectations for the hypothetical multiemployer pension plan whose asset allocation is shown in Exhibit 8 have changed from 2013 to 2016. (Note that 2013 was the first year this survey developed separate 10-year and 20-year expected returns.)

Both exhibits show the probabilities that the hypothetical pension plan will meet or exceed its 7.50% benchmark return on an annualized basis over the given time horizon. Exhibit 11 focuses on expected returns over a 10-year period, and Exhibit 12 focuses on expected returns over a 20-year period. Probabilities are shown for the survey average for each year from 2013 through 2016. For comparison, probabilities are also shown for the most conservative and optimistic advisors in each survey.

Exhibit 11

Exhibit 12

As shown in Exhibits 11 and 12, the probabilities that this hypothetical pension plan would meet or beat a benchmark return of 7.50% have generally come down from 2013 to 2015. From 2015 to 2016, the probability of meeting or beating a 7.50% annualized return over a 10-year period, based on the average survey assumptions, stayed relatively flat. The probability increased, however, when based on a 20-year horizon.

For example:

Based on the average assumptions from the 2016 survey, the probability of this hypothetical plan meeting or exceeding an annualized return of 7.50% over the next 10 years is 37.7%. For comparison, the probability was virtually the same (37.8%) based on average 2015 assumptions.

Based on the average assumptions from the 2016 survey, the probability of this hypothetical plan meeting or exceeding an annualized return of 7.50% over the next 20 years is 48.8%. The probability was somewhat lower (45.7%) based on average 2015 assumptions.

Other points of note when comparing the results from the 2016 survey to those from prior years:

In general, expected returns for the next 10 years remained relatively consistent from the 2015 survey to the 2016 survey. Much of the decline in expected 10-year returns is due to the addition of new advisors to the survey. In particular, the two advisors with the most conservative assumptions in the 2016 survey are first-time participants.

Average expected returns over the next 20 years are generally higher in the 2016 survey than in the 2015 survey. Some continuing survey participants who provide longer-term assumptions adjusted their expected returns upward. Another factor is that the advisor with the most optimistic 20-year assumptions is a first-time participant in 2016.

10%

20%

30%

40%

50%

60%

2013 2014 2015 2016

52.4% 50.2% 51.5% 51.4%

41.2% 40.6% 37.8% 37.7%

33.6% 34.0% 28.2% 23.4%

Most Optimistic

Survey Average

Most Conservative

Probability of Meeting 7.50% BenchmarkHypothetical Multiemployer Pension Fund

10-Year Horizon

Survey Year

20%

30%

40%

50%

60%

70%

2013 2014 2015 2016

63.0% 65.4% 61.4% 64.8%

51.0% 50.0% 45.7% 48.8%

39.4% 41.9% 32.1% 32.0%

Most Optimistic

Survey Average

Most Conservative

Probability of Meeting 7.50% BenchmarkHypothetical Multiemployer Pension Fund

20-Year Horizon

Survey Year


10 of 15

Glossary

The following are basic definitions of some of the investment terminology used in this report.

Expected Return

The expected return is the amount, as a percentage of assets, that an investment is expected to earn over a period of time. Expected returns presented in this survey are generally assumed to be indexed and net of investment fees.

Arithmetic vs. Geometric Returns

The arithmetic return is the average return in any one year; in other words, it has a one-year investment horizon. A geometric return is the annualized return over a multi-year period. In general, when evaluating expected returns over multi-year horizons, we find it more appropriate to focus on geometric returns.

Since 2013, the survey has presented expected returns on a geometric basis. Many advisors provide both arithmetic and geometric expected returns. For advisors who provided expected returns only on an arithmetic basis, we converted them to geometric returns for consistency. The following formula was used in making this conversion.

E[RG] = ((1 + E[RA])2 - VAR[R])1/2 - 1

In this formula, E[RG] is the expected geometric return, E[RA] is the expected arithmetic return, and VAR[R] is the variance of the expected annual return.

Standard Deviation

The standard deviation is a measure of the expected volatility in the returns. Generally, the standard deviation expresses how much returns may vary in any one year. Assuming that returns are “normally distributed,” there is about a 68% probability that the actual return for a given year will fall within one standard deviation (higher or lower) of the expected return. There is about a 95% probability that the actual return will fall within two standard deviations of the expected return.

Correlation

An important aspect of capital market assumptions is the degree to which the returns for two different asset classes move in tandem with one another: this is their correlation. For example, if two asset classes are perfectly correlated, their correlation coefficient will be 1.00; in other words, if one asset class has a return of X% in a given market environment, then the other asset class is expected to also have a return of X%. A portfolio becomes better diversified as its asset classes have lower (or even negative) correlations with each other.

Methodology

The following is a high-level description of the methodology used in compiling the survey results.

Standardized Asset Classes

Not all investment advisors use the same asset classes when developing their capital market assumptions. Some are very specific (more asset classes), while others keep things relatively simple (fewer asset classes).

We exercised judgment in classifying each advisor’s capital market assumptions into a standard set of asset classes. In the event that an advisor did not provide assumptions for a given asset class, the average assumptions from the other advisors was used when developing expected returns for that advisor.

Investment Horizons

This survey considers “short-term” expected returns to apply to 10-year investment horizon, and “long-term” expected returns to apply to a 20-year horizon.

In the 2016 survey, 23 of the 35 advisors provided only short-term assumptions, indicating a horizon of no more than 10 years. Included in this group are 2 advisors who provided assumptions over a horizon of 10 to 15 years.

All 12 advisors who provided long-term assumptions over horizons of 20 years or more also provided short-term assumptions. In cases where such an advisor indicated a horizon shorter than 10 years, the shorter-term expected returns were combined with the longer-term expected returns to achieve a 10-year horizon. If an advisor indicated a horizon longer than 20 years, the expected returns were assumed to also apply to a 20-year horizon.

No Adjustment for Alpha

No adjustment was made to reflect the possible value added by an active investment manager outperforming market returns (earning “alpha”).

Normally-Distributed Returns

This survey assumes that investment returns will be normally distributed according to the capital market assumptions provided. The survey also assumes that the investment return in one year does not affect the investment return in the following year.

Equal Weighting

Each advisor was given equal weight in developing the average assumptions for the survey, regardless of factors such as total assets under advisement, number of clients common with Horizon Actuarial, etc.

Survey of Capital Market Assumptions: 2016 Edition APPENDIX

11 of 15

Exhibit 13

The following exhibit evaluates the investment return assumption for a hypothetical multiemployer pension plan. It reflects the same hypothetical asset allocation as shown in Exhibit 8, and it provides more detail than Exhibits 9 and 10. Note that the most conservative and optimistic advisors for the 10-year horizon are not necessarily the same as the most conservative and optimistic advisors for the 20-year horizon. This hypothetical pension plan has a benchmark return of 7.50% per year, which is indicated by the gold line in the exhibit below.

Hypothetical Multiemployer Plan2016 Survey of Capital Market Assumptions

Individual Asset Classes Average Survey Assumptions Total Portfolio Results 10-Year Horizon 20-Year Horizon

Portfolio 10-Year 20-Year Standard Conservative Survey Optimistic Conservative Survey Optimistic

Asset Class Weight Horizon Horizon Deviation Advisor Average Advisor Advisor Average Advisor

US Equity - Large Cap 20.0% 6.64% 7.89% 16.92% Expected Returns

US Equity - Small/Mid Cap 10.0% 7.00% 8.23% 21.01% Average Annual Return (Arithmetic) 5.68% 6.97% 8.19% 6.92% 7.98% 9.21%

Non-US Equity - Developed 7.5% 7.12% 8.02% 19.50% Annualized Return (Geometric) 5.21% 6.42% 7.62% 6.36% 7.43% 8.53%

Non-US Equity - Emerging 5.0% 8.48% 9.11% 26.35% Annual Volatility (Standard Deviation) 9.97% 10.83% 11.05% 10.91% 10.94% 12.13%

US Corporate Bonds - Core 7.5% 3.41% 4.58% 5.96%

US Corporate Bonds - Long Duration 2.5% 3.82% 4.87% 10.49% Range of Expected Annualized Returns

US Corporate Bonds - High Yield 5.0% 5.90% 6.81% 11.01% 75th Percentile 7.34% 8.73% 9.98% 8.01% 9.08% 10.36%

Non-US Debt - Developed 5.0% 2.43% 3.70% 7.58% 25th Percentile 3.08% 4.11% 5.26% 4.72% 5.78% 6.70%

Non-US Debt - Emerging 2.5% 5.77% 6.43% 11.58%

US Treasuries (Cash Equivalents) 5.0% 2.14% 3.15% 2.79% Probabilities of Exceeding Certain Returns

TIPS (Inflation-Protected) 5.0% 2.80% 3.94% 6.51% 8.00% per Year, Annualized 18.8% 32.3% 45.7% 25.1% 40.8% 57.8%

Real Estate 10.0% 6.36% 6.75% 14.74% 7.50% per Year, Annualized 23.4% 37.7% 51.4% 32.0% 48.8% 64.8%

Hedge Funds 5.0% 5.41% 6.16% 8.39% 7.00% per Year, Annualized 28.5% 43.3% 57.0% 39.7% 57.0% 71.4%

Commodities 2.5% 3.98% 4.84% 18.50%

Infrastructure 2.5% 6.59% 7.12% 13.78%

Private Equity 5.0% 9.22% 10.33% 23.12%

Inflation N/A 2.16% 2.31% 1.78%

TOTAL PORTFOLIO 100.0% Expected returns are geometric.

Considerations and Limitations

- Allocations may be approximated if certain asset classes are not included in the survey.

- Many investment advisors provided only shorter-term assumptions (10 years or less).

- Assumptions are based on indexed returns and do not reflect anticipated alpha.

- Assumptions do not reflect investment opportunities or fee considerations available to larger funds.



Expected returns over a 20-year horizon are based a subset of 12 survey participants who provided longer-term assumptions.

2.5%

3.5%

4.5%

5.5%

6.5%

7.5%

8.5%

9.5%

10.5%

11.5%

Conservative

Advisor

Survey

Average

Optimistic

Advisor

Conservative

Advisor

Survey

Average

Optimistic

Advisor

10-Year Horizon 20-Year Horizon

Ranges of Expected Annualized Returns


12 of 15

Exhibit 14

The following exhibit provides the average capital market assumptions for all 35 investment advisors in the 2016 survey. Each of the 35 advisors was given equal weight in determining the average assumptions. For reference, expected returns are shown over 10-year and 20-year horizons. Expected returns are also provided on both an arithmetic basis (one-year average) and geometric basis (multi-year annualized). The standard deviations (volatilities) and correlations apply to both arithmetic and geometric expected returns.

Horizon Actuarial 2016 Survey of Capital Market AssumptionsAverage Survey Assumptions

Expected Returns

10-Year Horizon 20-Year Horizon Standard Correlation Matrix

Asset Class Arith. Geom. Arith. Geom. Deviation 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1 US Equity - Large Cap 7.98% 6.64% 9.25% 7.89% 16.92% 1 1.00

2 US Equity - Small/Mid Cap 9.07% 7.00% 10.40% 8.23% 21.01% 2 0.90 1.00

3 Non-US Equity - Developed 8.90% 7.12% 9.77% 8.02% 19.50% 3 0.82 0.76 1.00

4 Non-US Equity - Emerging 11.68% 8.48% 12.46% 9.11% 26.35% 4 0.73 0.72 0.79 1.00

5 US Corporate Bonds - Core 3.59% 3.41% 4.75% 4.58% 5.96% 5 0.15 0.11 0.15 0.10 1.00

6 US Corporate Bonds - Long Duration 4.37% 3.82% 5.58% 4.87% 10.49% 6 0.13 0.09 0.14 0.11 0.91 1.00

7 US Corporate Bonds - High Yield 6.49% 5.90% 7.40% 6.81% 11.01% 7 0.64 0.64 0.62 0.64 0.36 0.35 1.00

8 Non-US Debt - Developed 2.74% 2.43% 4.01% 3.70% 7.58% 8 0.11 0.06 0.30 0.17 0.57 0.53 0.21 1.00

9 Non-US Debt - Emerging 6.42% 5.77% 7.20% 6.43% 11.58% 9 0.55 0.51 0.59 0.66 0.44 0.38 0.63 0.40 1.00

10 US Treasuries (Cash Equivalents) 2.22% 2.14% 3.18% 3.15% 2.79% 10 (0.08) (0.10) (0.06) (0.06) 0.30 0.22 (0.04) 0.23 0.10 1.00

11 TIPS (Inflation-Protected) 3.03% 2.80% 4.27% 3.94% 6.51% 11 0.02 0.01 0.08 0.13 0.68 0.63 0.27 0.51 0.40 0.29 1.00

12 Real Estate 7.48% 6.36% 7.75% 6.75% 14.74% 12 0.38 0.38 0.36 0.31 0.07 0.11 0.31 0.06 0.25 0.04 0.13 1.00

13 Hedge Funds 5.77% 5.41% 6.59% 6.16% 8.39% 13 0.62 0.62 0.65 0.64 0.14 0.10 0.54 0.11 0.48 (0.01) 0.11 0.27 1.00

14 Commodities 5.62% 3.98% 6.47% 4.84% 18.50% 14 0.30 0.30 0.40 0.44 0.07 0.03 0.32 0.20 0.36 0.02 0.26 0.22 0.43 1.00

15 Infrastructure 7.52% 6.59% 8.26% 7.12% 13.78% 15 0.53 0.51 0.56 0.50 0.23 0.21 0.48 0.23 0.37 0.02 0.16 0.27 0.46 0.32 1.00

16 Private Equity 11.77% 9.22% 12.94% 10.33% 23.12% 16 0.77 0.76 0.72 0.64 0.02 0.05 0.54 0.06 0.46 (0.05) (0.04) 0.39 0.60 0.29 0.43 1.00

Inflation 2.16% 2.16% 2.31% 2.31% 1.78%





13 of 15

Exhibit 15

Earlier in this report, Exhibit 5 showed the distribution of expected returns and standard deviations over an investment horizon of 10 years. Exhibit 15 below shows the same distribution, but for a horizon of 20 years. Note that while Exhibit 5 included assumptions for all 35 advisors in the survey, the exhibit below includes only assumptions for the 12 advisors who provided longer-term assumptions (horizons of 20 years or more).

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

10%

11%

12%

13%

14%

0% 5% 10% 15% 20% 25% 30% 35%

Exp

ecte

d R

etu

rn

Standard Deviation

US Equity - Large Cap [ 7.9% | 17.2% ]

US Equity - Small/Mid Cap [ 8.2% | 21.6% ]

Non-US Equity - Developed [ 8.0% | 19.4% ]

Non-US Equity - Emerging [ 9.1% | 27.0% ]

US Corporate Bonds - Core [ 4.6% | 5.9% ]

US Corporate Bonds - Long Duration [ 4.9% | 12.2% ]

US Corporate Bonds - High Yield [ 6.8% | 11.1% ]

Non-US Debt - Developed [ 3.7% | 7.5% ]

Non-US Debt - Emerging [ 6.4% | 12.7% ]

US Treasuries (Cash Equivalents) [ 3.2% | 1.8% ]

TIPS (Inflation-Protected) [ 3.9% | 7.5% ]

Real Estate [ 6.7% | 14.5% ]

Hedge Funds [ 6.2% | 9.1% ]

Commodities [ 4.8% | 18.4% ]

Infrastructure [ 7.1% | 15.6% ]

Private Equity [ 10.3% | 23.8% ]

Alt

ern

ativ

esFi

xed

Inco

me

Equ

itie

s

2016 Survey: Distribution of Expected Returns and Standard Deviations

Asset Class [ Avg. Exp. Return | Avg. Std. Dev. ]

20-Year Horizon | Geometric Returns




14 of 15

Exhibit 16

The exhibit below shows the ranges of expected annual returns for different asset classes over a 10-year investment horizon. The ranges shown below include assumptions for all the 35 advisors in the 2016 survey. Expected returns shown below are annualized (geometric). To illustrate the distribution of expected returns, the exhibit shows the range of the middle 50 percent of results: the range between the 25th and 75th percentiles. It also shows the median expected return for each asset class: the 50th percentile. Note that the expected returns for the median advisor shown below are not the same as the average expected returns shown elsewhere in the report. In most cases, however, the differences between median and average expected returns are relatively small.

0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0%

Inflation [ 1.5% | 2.0% | 2.2% | 2.3% | 2.8% ]

Private Equity [ 7.4% | 8.5% | 9.0% | 10.0% | 11.8% ]

Infrastructure [ 5.5% | 6.3% | 6.6% | 6.8% | 8.0% ]

Commodities [ 1.8% | 3.0% | 3.8% | 4.8% | 6.8% ]

Hedge Funds [ 4.2% | 4.7% | 5.3% | 6.0% | 7.6% ]

Real Estate [ 4.1% | 5.5% | 6.2% | 6.8% | 10.0% ]

TIPS (Inflation-Protected) [ 1.9% | 2.5% | 2.7% | 3.0% | 5.0% ]

US Treasuries (Cash Equivalents) [ 0.3% | 2.0% | 2.2% | 2.4% | 3.5% ]

Non-US Debt - Emerging [ 3.9% | 5.3% | 5.6% | 6.3% | 7.7% ]

Non-US Debt - Developed [ 0.9% | 1.7% | 2.3% | 3.1% | 4.6% ]

US Corporate Bonds - High Yield [ 3.7% | 5.4% | 5.8% | 6.8% | 7.5% ]

US Corporate Bonds - Long Duration [ 2.3% | 3.1% | 3.9% | 4.6% | 5.3% ]

US Corporate Bonds - Core [ 2.1% | 3.0% | 3.3% | 3.7% | 5.3% ]

Non-US Equity - Emerging [ 5.8% | 7.5% | 8.5% | 9.6% | 11.3% ]

Non-US Equity - Developed [ 3.0% | 6.6% | 7.3% | 7.7% | 9.2% ]

US Equity - Small/Mid Cap [ 5.0% | 6.3% | 7.1% | 7.7% | 9.0% ]

US Equity - Large Cap [ 5.1% | 6.1% | 6.9% | 7.2% | 8.0% ]

2016 Survey: Expected Returns by Asset Class (10-Year Horizon)

Asset Class [ Min | 25th | 50th | 75th | Max ]

Alt

ern

ativ

e Fi

xed

Inco

me

Equ

itiy


Expected returns are annualized over 10 years (geometric).

Expected Annual Return


15 of 15

Exhibit 17

The exhibit below shows the ranges of expected annual returns for different asset classes over a 20-year investment horizon. The ranges shown below are based on the assumptions for 12 advisors who provided longer-term assumptions (horizons of 20 years or more). Expected returns shown below are annualized (geometric). Note that the ranges of expected returns are somewhat narrower when the investment horizon is longer. To illustrate the distribution of expected returns, the exhibit shows the range of the middle 50 percent of results: the range between the 25th and 75th percentiles. It also shows the median expected return for each asset class: the 50th percentile. Note that the expected returns for the median advisor shown below are not the same as the average expected returns shown elsewhere in the report. In most cases, however, the differences between median and average expected returns are relatively small.

0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 14.0%

Inflation [ 2.0% | 2.1% | 2.3% | 2.5% | 2.8% ]

Private Equity [ 8.3% | 9.2% | 10.0% | 11.9% | 12.4% ]

Infrastructure [ 6.7% | 6.9% | 7.2% | 7.4% | 7.5% ]

Commodities [ 4.1% | 4.5% | 4.8% | 5.3% | 5.4% ]

Hedge Funds [ 5.2% | 5.6% | 5.9% | 6.6% | 7.3% ]

Real Estate [ 5.3% | 5.9% | 6.3% | 7.2% | 11.2% ]

TIPS (Inflation-Protected) [ 2.9% | 3.3% | 3.8% | 4.6% | 5.7% ]

US Treasuries (Cash Equivalents) [ 2.1% | 2.9% | 3.1% | 3.3% | 5.3% ]

Non-US Debt - Emerging [ 5.4% | 5.8% | 6.1% | 6.8% | 8.7% ]

Non-US Debt - Developed [ 2.3% | 2.4% | 3.5% | 4.9% | 5.5% ]

US Corporate Bonds - High Yield [ 4.7% | 6.3% | 6.7% | 7.7% | 8.3% ]

US Corporate Bonds - Long Duration [ 3.1% | 3.8% | 4.6% | 6.2% | 6.6% ]

US Corporate Bonds - Core [ 3.1% | 3.7% | 4.9% | 5.1% | 6.5% ]

Non-US Equity - Emerging [ 7.6% | 7.9% | 9.3% | 9.6% | 12.1% ]

Non-US Equity - Developed [ 6.7% | 7.5% | 7.9% | 8.7% | 9.3% ]

US Equity - Small/Mid Cap [ 6.9% | 7.1% | 7.7% | 9.8% | 10.4% ]

US Equity - Large Cap [ 6.5% | 7.3% | 7.7% | 8.7% | 9.0% ]

2016 Survey: Expected Returns by Asset Class (20-Year Horizon)

Asset Class [ Min | 25th | 50th | 75th | Max ]

Alt

ern

ativ

e Fi

xed

Inco

me

Equ

itiy


Expected returns are annualized over 20 years (geometric), based on a subset of 12 advisors who provided longer term assumptions.

Expected Annual Return

GASB Outlook From the Chairman››

Calculating an

appropriate

discount rate to

measure the net

liability for

postemployment

benefits is a critical

financial

accounting and

reporting issue for

state and local

governments.

Calculating an appropriate discount rate to measure the

net liability for postemployment benefits is a critical

financial accounting and reporting issue for state and

local governments. The long-term expected rate of

return is a fundamental component used in developing

the discount rate. As can be seen by the sensitivity

disclosures required by the postemployment benefits

standards, a change of just 1 percentage point in the

discount rate can have for many plans a significant

impact on the net liability.

In justifying the long-term expected rate of return, one often hears

“historical investment performance supports that rate.” The

standards, however, address the long-term expected rate of return.

Historical data can be inconsistent with the forward-looking nature

of this expectation and is not a complete source for the development

of long-term anticipations about future economic phenomena.

The long-term expected rate of return should be based upon the

nature and mix of current and expected postemployment benefit

investments. That means the postemployment benefit investments

must be expected to be invested using a strategy to achieve that

return.

During the development of the postemployment benefits standards

the Board concluded that it was not within the scope of the Board’s

activities to set standards that establish a specific funding method

for postemployment benefits–that is a policy decision for

government officials or other responsible authorities to make.

Accordingly, the postemployment benefits standards set

requirements in the context of accounting, not funding. This is a very important distinction, as one

also often hears “we will reduce the discount rate gradually over time, that’s all we can afford now.”

Affordability is a funding issue, not an accounting issue.

The accounting standards require the use of the long-term expected rate of return to develop the

discount rate–funding affordability is not a component to be considered in determining the

From the Chairman http://www.gasb.org/cs/ContentServer?c=Page&pagename=GASB/Pag...

1 of 2 3/6/2017 1:18 PM

long-term expected rate of return when developing the discount rate for financial accounting and

reporting purposes. To appropriately comply with the postemployment benefits standards for financial

reporting purposes, it is critical that governments measure the net liability for postemployment benefits

using a discount rate based on an accounting perspective–one that appropriately incorporates the

long-term expected rate of return–not a rate based on a funding affordability perspective.

From the Chairman http://www.gasb.org/cs/ContentServer?c=Page&pagename=GASB/Pag...

2 of 2 3/6/2017 1:18 PM

1

Funding Risk Mitigation Policy

Effective Date November 18, 2015

This policy is effective immediately upon adoption.

Introduction The Funding Risk Mitigation Policy (Policy) seeks to reduce CalPERS

funding risk over time. It establishes a mechanism whereby CalPERS investment performance that significantly outperforms the discount rate triggers adjustments to the discount rate, expected investment return, and strategic asset allocation targets. Reducing the volatility of investment returns will increase the long-term sustainability of CalPERS pension benefits for members.

Policy This Policy applies to the Public Employees’ Retirement Fund (PERF).

If a Funding Risk Mitigation Event occurs, the discount rate and expected investment return shall be adjusted as set forth in Table 1 below, and the strategic asset allocation targets shall be adjusted consistent with such new discount rate and expected investment return. The current CalPERS strategic asset allocation targets can be found in the CalPERS Total Fund Investment Policy, and are defined or approved during the periodic Asset Liability Management process undertaken by CalPERS, subject to adjustments per this Policy.

Table 1: Funding Risk Mitigation Event Thresholds and Impacts

Excess Investment Return

Reduction in Discount Rate

Reduction in Expected

Investment Return

If the actual investment returns exceed the discount rate by:

Then the discount rate will be reduced by:

And the expected investment return will be reduced by:

4.00% 0.05% 0.05%

7.00% 0.10% 0.10%

10.00% 0.15% 0.15%

13.00% 0.20% 0.20%

17.00% 0.25% 0.25%

2

Discount rate reduction shall be governed by the following parameters: a) Reduced by increments of five (5) basis points (0.05%) b) Maximum reduction per year of 25 basis points (0.25%) c) The discount rate/expected investment return shall not be

reduced to the point where the estimated investment return volatility drops below eight percent (8%) according to the Capital Market Assumptions most recently adopted by the Investment Committee.

Upon the occurrence of a Funding Risk Mitigation Event:

1. Staff shall report the annual net investment return for the given fiscal year ending June 30th to the CalPERS Board of Administration.

2. Staff shall implement new strategic asset allocation targets based on the reduction in investment return indicated in Table 1 in accordance with the current schedule of asset allocation ranges and targets adopted by the Investment Committee.

3. The new strategic asset allocation targets shall take effect on

October 1 of the fiscal year immediately following the Event Year.

4. The total fund policy benchmark shall be adjusted consistent

with the new strategic asset allocation targets and Staff shall report the new strategic asset allocation targets, total fund policy benchmark and expected investment return to the Investment Committee.

5. The discount rate shall be adjusted and reported to the Finance & Administration Committee.

6. Member calculations, including optional factors and service credit purchase, shall reflect the reduced discount rate effective October 1 of the fiscal year immediately following the Event Year.

7. The effect of any reduction in discount rate for a given Event Year shall be included in the actuarial valuations calculated as of June 30 for such year.

3

Reviews Funding risk mitigation progress will be reported in the Annual Funding

Levels and Risks Report. CalPERS staff periodically conducts the ALM process with the Board of Administration. This policy shall be reviewed as part of the CalPERS cyclical Asset Liability Management (ALM) process. Staff will review funding risk mitigation actions taken with the Board and the Board will assess the progress that has been made.

Glossary of CalPERS Specific Terms

Italicized terms appearing in the Policy are CalPERS specific in nature and are defined below:

Term Definition

Funding Risk Mitigation Event The achievement of a time-weighted annual investment return net of investment expenses for a given fiscal year, as first publicly reported following the end of such fiscal year, that exceeds the CalPERS discount rate by 4.00% or more.

Event Year The fiscal year in which the funding risk mitigation event occurred.

Threshold The time-weighted annual investment return, net of investment expenses, in excess of the discount rate required for a funding risk mitigation event to occur.

CalPERS to Lower Discount Rate to Seven Percent Over the Next Three... https://www.calpers.ca.gov/page/newsroom/calpers-news/2016/calpers-...

1 of 3 2/16/2017 12:00 PM

CalPERS to Lower Discount Rate to Seven Percent Over the Next Three... https://www.calpers.ca.gov/page/newsroom/calpers-news/2016/calpers-...

2 of 3 2/16/2017 12:00 PM

Finance and Administration Committee Agenda Item 9a

February 14, 2017 Item Name: Funding Risk Mitigation Policy Program: ASSET LIABILITY MANAGEMENT Item Type: Action Recommendation Staff recommends the California Public Employees’ Retirement System (“CalPERS” or the “System”) Board adopt the revised Funding Risk Mitigation Policy (“Policy”), which has been updated per the Board’s instructions at the December 2016 meetings of the Finance and Administration Committee (“FAC”) and of the Board. Per the Board’s additional instructions, staff recommends that the Board suspend the implementation of the Policy until Fiscal Year 2020-21 because of the Board’s decision at the December 2016 meeting to lower the Discount Rate pursuant to a phase-in schedule through Fiscal Year 2019-20. Executive Summary The Board’s action to lower the Discount Rate has implications for the Policy, which prescribes how the Discount Rate is to be lowered if the Public Employees Retirement Fund (“PERF”) achieves an investment return in a fiscal year that exceeds the Discount Rate. As a result of its deliberations, the Board directed staff to revise the Policy by lowering the first threshold for the percentage by which the actual investment return exceeds the discount rate in any fiscal year in order to trigger a discount rate reduction from 4% to 2%, as depicted in the table below:

Table 1: Funding Risk Mitigation Event Thresholds and Impacts

Excess Investment Return

Reduction in Discount Rate

Reduction in Expected Investment Return

If the actual investment returns exceed the discount rate by:

Then the discount rate will be reduced by:

And the expected investment return will be reduced by:

2.00% 0.05% 0.05% 7.00% 0.10% 0.10% 10.00% 0.15% 0.15% 13.00% 0.20% 0.20% 17.00% 0.25% 0.25%

All other excess return thresholds and rate reductions remain the same as in the current Policy.

Agenda Item 9a Finance and Administration Committee

Page 2 of 4

The Board also directed staff to revise the Policy to state that member calculations, including optional factors and service credit purchases, shall reflect the reduced Discount Rate effective immediately upon the occurrence of a Funding Risk Mitigation Event rather than on October 1 of the Fiscal Year following the Funding Risk Mitigation Event Year. With respect to revising the Policy, CalPERS has implemented a new standard policy format, and the Policy has been updated to comply with this format.

Finally, at its December 2016 meeting the FAC discussed introducing a motion to the Board to suspend the implementation of the Policy until Fiscal Year 2020-21 because of the Board’s decision to lower the Discount Rate pursuant to a phase-in schedule through Fiscal Year 2019-20. Strategic Plan This agenda item supports the following strategic plan goals:

Goal A – Improve long-term pension … sustainability by funding the System through an integrated view of pension assets and liabilities and actively assessing and managing funding risk through an asset liability management framework to guide investment strategy and actuarial policy.

Goal B – Cultivate a high-performing, risk-intelligent and innovative organization.

Background In February 2015, a workshop was conducted by staff for the Board to provide an overview of the risks faced in the funding of the System and how these risks are changing, and introduced several concepts that could be applied to mitigate these risks. During the discussion that took place at the workshop, the Board expressed a desire to further explore two possible risk mitigation strategies. In a May 2015 workshop, staff provided detailed information on the two risk mitigation strategies. In August 2015, additional information was presented by staff related to the two risk mitigation strategies under consideration by the Board, including an update on the outreach and engagement activities conducted to ensure feedback from stakeholders. During the October 2015 FAC meeting, staff presented the draft policy for the 1st reading. Based on feedback from the Board, staff was directed to develop a second policy with an update of the first threshold level that would trigger a risk mitigation event from 4% to 2%. Both policies were then presented for a 2nd reading during the November 2015 FAC meeting for review and approval of one policy. The FAC adopted the Policy with the initial threshold of 4%. At the December 2016 meetings of the FAC and the full Board, the Board directed staff to amend the Policy to lower the initial threshold to 2% and to make the other changes as outlined above. Analysis Staff presented the following analysis of the difference between using a 2% initial threshold versus 4% when the current policy was adopted November 2015. Using the 2% threshold:

• The pace of risk mitigation would increase slightly. The probability of reaching an 8% level of volatility (6.5% discount rate) in 30 years using 4% as the initial threshold is


Page 3 of 4

91%. With a 2% initial threshold it would be 96%, and the average time to reach an 8% level of volatility would be 19 years instead of 21 years.

• In the original proposed policy, assuming a drop in volatility levels to 8%, PEPRA miscellaneous members would see their contributions rise by approximately 1.5% (3% for safety) over a 30-year period. These levels would not change under the 2% threshold, but members could expect to see the increases sooner depending on the pace of risk mitigation.

• With a 2% initial threshold, employers would see modest increases in rates as compared to the 4% proposal, but those are expected to be less than 0.3% for miscellaneous and 0.5% for safety in any given year during risk mitigation.

• The only time the 2% initial threshold would produce a difference in a risk mitigation event would be when there is an investment return between 2-4% above the discount rate at that time. Based on modeling, this happens about 6.6% of the time, or about once every 15 years.

Budget and Fiscal Impacts There are no budget or fiscal impacts at this time. Benefits and Risks As stewards of the System, CalPERS must ensure that the pension fund is sustainable over multiple generations by taking steps to mitigate risks over the long-term through an integrated view of assets and liabilities. Benefits of a risk mitigation strategy include:

Strengthens the long-term sustainability of the fund and security of future benefit payments

Protects the fund from volatility of short-term investment returns or changing demographics that could reduce CalPERS long term funded status

Reduces the level of future risk in the investment portfolio

Reduces the volatility in contribution rates for employers

Risks associated with a risk mitigation strategy include: Increases financial stress on employers as a result of increased contributions when the

discount rate is reduced


Page 4 of 4

Attachments Attachment 1 – Current Funding Risk Mitigation Policy

Attachment 2 – Proposed Revised Funding Risk Mitigation Policy with Key Changes Highlighted

_________________________________ SCOTT TERANDO Chief Actuary _________________________________ THEODORE H. ELIOPOULOS Chief Investment Officer _________________________________ MARLENE TIMBERLAKE D'ADAMO Interim Chief Financial Officer

Asset Allocation Review

PERS Board Education

March 28,2017

James W. Van Heuit Capital Market Research Group

William C. Howard, CFA Denver Fund Sponsor Consulting

John P. Jackson, CFA Chicago Fund Sponsor Consulting

1 PERS Asset Allocation Review Knowledge. Experience. Integrity.

Agenda

● Building Blocks of Asset Class Returns – Components of returns – Contributions to investment forecasts

● Economic Outlook – Historical, current and prospective conditions – Implications for asset market forecasts

● Asset Class Forecasts – Equity – Fixed Income

● Asset Allocation Analysis – Alternative asset mix construction – Expected performance – Range of performance expectations

Topics for Discussion


Building blocks

● Investment markets are complex

● It is advantageous for the forecasting process to break markets down into “building blocks”

● Risk Premia Building Blocks – Investments span the risk spectrum from very safe Treasury bills to volatile emerging markets equity – Theoretically, investors should require more compensation in terms of higher returns in order to invest in riskier

markets – Forecasts are based on historical return “spreads” across the range of risky assets

● Economic and Financial Building Blocks – Generally speaking, investments are either debt or equity

– Investors are either lenders or owners – Investors are compensated for their contributions to economic activity (e.g., businesses, governments,

mortgage borrowers) – Economic and financial building blocks look at the sources of funds to compensate investors

– Revenues to pay government bond investors – Cash flow to pay interest to corporate bondholders and dividends to shareholders – Reinvestment to promote profit growth

– Forecasts are based on the opportunities available to generate cash flow and how much investors are willing to pay to receive these cash flows

Contributors to Return


Asset Class Returns

Risk Premia Building Blocks

● Investors require greater return for taking greater levels of risk: “risk premia”

● Risk premia have varied widely historically – In some periods, e.g., during and immediately after the financial crisis, equities returned less than bonds – What are the “right” levels for risk premia? How should they change through time?

● Treasury bills provide “risk-free” returns – If purchased and held to maturity the return is known in advance – The U.S. government is not going to default

● Bond risk premia – Term Risk: More time to maturity means greater uncertainty – Credit Risk: Higher probability of default means more uncertainty for future interest payments, and the

repayment of principal – The additional risk means that bonds need to offer a higher return than Treasury bills

● Equity risk premia – Stocks pay dividends which are not fixed in advance like bond interest rate payments – Stock values can fluctuate substantially, more widely than most bonds – The additional risk means that stocks need to offer a higher return than bonds

Building Blocks


Asset Class Returns

● For 30 years the S&P 500 has averaged a 6.33% return premium over 3-month Treasury bills

● The highest return premium was 14.5% at the peak of the tech bubble

● The lowest return premium was -6.4% at the nadir of the financial crisis

● The forecast should be somewhere in between these extremes, but it is unclear where – The historical time period is decisive in determining the forecast

Risk Premia Building Blocks

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

(10)

(5)

0

5

10

15

20

for 30 Years Ended December 31, 2016Rolling 40 Quarter Excess Return Relative To CE Fed 3 Mo Bill

Exc

ess

Ret

urn

US Eq S&P 500 US Eq S&P 500 Average

6.33


Asset Class Returns

Economic and Financial Building Blocks

● Economic factors play an important role – GDP growth and its underlying components – Inflation and its underlying components – These values tend to move up and down together

– More growth leads to more inflation – Less growth leads to less inflation

● Treasury bills are still “risk-free” returns – Yields depend on inflation over their brief time to maturity – Investors normally want a positive after-inflation “real” return

● Projected bond returns will be high if investors anticipate high growth and inflation – Bond investors also want a positive real rate of return – Credit spreads will be low if growth reduces the probability of corporate default

● Projected equity returns will be higher if investors anticipate high growth and inflation – Higher economic growth rates generally mean higher profits, dividends and capital appreciation

● Still, uncertainty over “right” levels of real bond and stock returns fluctuate as conditions change over time

● Investors still require greater return for taking greater levels of risk

Building Blocks


Economic Outlook

● Real Gross Domestic Product (GDP) Growth and Consumer Price Inflation (CPI) are forecast

● Callan forecasts are based on forecasts provided by the Federal Reserve and the International Monetary Fund (IMF) – Forecasts are country-specific for non-US markets – Aggregated non-US forecasts are weighted averages using World ex USA and Emerging Markets index

weights

● Forecasts are intertwined – GDP and inflation tend to rise and fall together

● Forecasts form a starting point for projections – GDP forecasts provide a very rough estimate of future earnings growth – Inflation forecasts provide an approximate path for short-term yields – Inflation is added to the real return forecasts for equity and fixed income

Role of Economic Variables


Economic Outlook

● Employment compensation as a percentage of national income has been falling for decades

● Falling compensation limits consumption which can limit economic growth

Employment Compensation Contribution to Gross Domestic Income (GDI)


Economic Outlook

● Corporate profits have been growing as a percentage of national income

● The impact of higher levels of corporate profits depends on how profits are used

Corporate Profits Contribution to Gross Domestic Income


Economic Outlook

● Gross private domestic nonresidential investment measures investments by companies

● While profits have been growing, investments by firms have been trending downward

● Investments increase productivity which leads to higher levels of economic growth

Private Domestic Investment Contribution to Gross Domestic Product


Economic Outlook

● Gross private domestic personal consumption expenditures measures the total amount of what consumers pay to live their daily lives

● Personal consumption expenditures as a percentage of GDP have experienced limited growth since over the most recent 15 years

Personal Consumption Expenditures Contribution to GDP


Economic Outlook

● Real GDP growth rebounded dramatically after the Global Financial Crisis

● Growth since the GFC has been lower than over previous long-term periods

● Growth is likely to be more muted going forward

Developed Markets Real GDP Growth

*

Source: International Monetary Fund


Economic Outlook

● Consumer price inflation fell in the aftermath of the Global Financial Crisis

● It has been a low levels since in spite of monetary and fiscal stimulus designed to increase it

● There are recent indications that inflation in the US is starting to pick up

Developed Markets Consumer Price Inflation

Source: International Monetary Fund


Economic Outlook The Federal Reserve Dot Chart

5 — — 5

4.25 — — 4.25 4.00 — — 4.00 3.75 — • • — 3.75 3.50 — • • — 3.50 3.25 — • •• — 3.25 3.00 — •• •••• •••••••• — 3.00 2.75 — •• •••••• — 2.75 2.50 — •• ••• • — 2.50 2.25 — •• •• — 2.25 2.00 — • ••• • — 2.00 1.75 — ••••• — 1.75 1.50 — •••• • — 1.50 1.25 — •••••• — 1.25 1.00 — •••• — 1.00 0.75 — •• • • — 0.75 0.50 — ••••••••••••••••• — 0.50 0.25 — — 0.25 0.00 — — 0.00

2016 2017 2018 2019 Longer run

Target fed funds rate at year-end (December projections)

Each shaded circle indicates the value of an individual participant’s judgment of the midpoint of the target federal funds rate at the end of the specified calendar year and over the longer run. The number in each column represents the lower bound of an 0.25 percentage point range.


Economic Outlook U.S. Treasury Yield Curves

Source: Federal Reserve and Callan


Economic Outlook U.S. Treasury Yield Curves


Economic Outlook Trailing P/E for the S&P 500

Trailing earnings as reported for the fiscal year; includes negative earnings from 1998 onward. Source: Standard & Poor’s and Callan

0

5

10

15

20

25

30

35

40

45

54 58 62 66 70 74 78 82 86 90 94 98 02 06 10 14

Pric

e/E

arni

ngs

Rat

io

Price to Earnings Ratio for S&P 500 (1954 - 2016)

S&P 500 P/E Ratio Long-Run Average

+ 2 Std. Dev.

- 2 Std. Dev.


Economic Outlook

● 2% to 2.5% for the US – Higher growth rate than the post financial crisis time period but lower than the last half century average – Factors which could lead to upper end of forecast

– Strong labor market – Expansive fiscal policies of the incoming administration – Restrictive trade policies

– Factors which could lead to lower end of forecast – Tighter monetary policy, unexpectedly high interest rates or both – Congressional resistance to the fiscal policies of the incoming administration – Strong dollar

● 1.5% to 2.0% for Developed Non-US Markets – Lower than the US due to concerns about political, fiscal and monetary policy as well as the banking system – Factors which could lead to upper end of forecast

– Additional stimulative monetary policies – Controlled growth of government budget deficits – Reduced levels of austerity – Improvements in bank balance sheets – Reduced unemployment

– Factors which could lead to lower end of forecast – Political uncertainty – Trade issues – Declining health of the financial sector

GDP Growth Forecasts


Economic Outlook

● 4% to 5% for Emerging Markets – Absolute size of many emerging economies (primarily China and India) limits growth to below historical values – Growth rates still substantially exceed those of the developed markets – Factors which could lead to upper end of forecast

– Improving export markets – Expanding internal demand

– Factors which could lead to lower end of forecast – Domestic government policies – Developed market trade restrictions

GDP Growth Forecasts


Economic Outlook

● 2% to 2.5% for the US – Headline values ticking up but still less than 1.75% while core inflation has been above 2% since late 2015 – Factors which could lead to upper end of forecast

– Rising energy prices – Tight labor markets – Fiscal stimulus – Declining dollar drives up import prices

– Factors which could lead to lower end of forecast – Poor overseas economic performance strengthens the dollar – Competitive labor market keeps wages in check – The Fed implements tight monetary policy

● 1.75% to 2.25% for Developed Non-US Markets – Inflation is starting to tick up but is very low in largest economies – Factors which could lead to upper end of forecast

– More expansive fiscal policies – Faster economic growth

– Factors which could lead to lower end of forecast – Opportunities for additional fiscal and monetary stimulus are limited – Slack in European labor markets

Inflation Forecasts


Economic Outlook

● 2.5% to 3.5% for Emerging Markets – Future inflation is uncertain – Path of prices depends on government policies, relative currency strength, trade policies, the balance of

internal supply and demand, and commodity prices

Inflation Forecasts


Equity Forecasts

● Fundamental Relationship

● Forecast capital appreciation depends on projected future earnings – Long-term earnings tend to correspond to long-term GDP growth

– Weak short-term relationship – Relationship more robust in developed than emerging markets economies

– Investors will pay more for stocks with better future earnings potential – Prices don’t depend on historical or current earnings

● Forecast income also depends on projected future earnings – Income is related to earnings via the payout ratio – Income also influenced by

– Prospects for future corporate investments – Interest rates

● Valuations have limited impact on forecasts – Average P/Es over different market cycles differ markedly

– Oil Boycott – Tech Bubble – Global Financial Crisis

– Capital market projections only impacted when markets reach extreme valuations

Overview

Equity Return = Capital Appreciation + Income


Equity Forecasts

● Return = 6.85%, Risk = 18.25%

● Broad US equity is represented by the Russell 3000 index

● Earnings growth likely to be moderate – Earnings growth is likely to be higher than it has been since 2014 – Stronger GDP growth than would have been the case without more expansive economic policies – GDP growth still likely to lag its longer-term average

● Dividend yield consistent with recent history – Payout ratios close to historical norms – Yields have been stable for 20 years in the face of changing interest rates

● Any additional return from share buybacks likely offset by dilution from additional issuance

Broad US Equity


Equity Forecasts

● Return = 7.00%, Risk = 21.00%

● Global ex US equity is represented by the MSCI All-Country World ex USA index

● Earnings growth likely to be moderate – Improving outlook for developed markets economies but from a low starting point – Emerging market growth declining but still substantial – Significant uncertainty in future economic policies

● Relatively high dividend yields will support returns – Developed markets yields are measurably higher than those in the US

– High even relative to history – Dividends to make up differences in capital appreciations between US and developed non-US markets

– Emerging markets yields not as high as in developed markets but still higher than in US

Global ex US Equity


Fixed Income Forecasts

● Fundamental Relationship

● Forecast capital appreciation depends on projected future interest rates – Inflation – Central bank policy – Credit conditions

● Income = yield

● Roll return reflects capital appreciation from declines in yields as bonds move toward maturity with upward sloping yield curves

Overview

Bond Return = Capital Appreciation + Income + Roll Return


Fixed Income Forecasts

● Return = 3.00%, Risk = 3.75%

● Broad US fixed income is represented by the Bloomberg Barclays Aggregate index

● Interest rates expected to rise – Most of the increase is expected over the next 3 years – Our path is consistent with that forecast by the Fed

● Yield curve expected to flatten – Yield curve currently steep – Long rates are still expected to increase but not as much as short rates

● Higher yields expected to be earned over most of the forecast horizon

● Capital losses expected as yields increase in early years – Losses consistent with moderate duration

– Historically about 5 but currently closer to 6 – Little impact from changing credit spreads

● Roll return expected to decline – Current steep yield curve provides relatively high roll return – Flatter curve will reduce its contribution to total return

Broad US Fixed Income


Return and Risk Projections 10-Year Time Horizon

PROJECTED RETURN PROJECTED RISK

Asset Class Index1-Year

Arithmetic10-Year

Geometric* RealStandard Deviation

Projected Yield

EquitiesBroad Domestic Equity Russell 3000 8.30% 6.85% 4.60% 18.25% 2.00%Large Cap S&P 500 8.05% 6.75% 4.50% 17.40% 2.10%Small/Mid Cap Russell 2500 9.30% 7.00% 4.75% 22.60% 1.55%Global ex-US Equity MSCI ACWI ex USA 8.95% 7.00% 4.75% 21.00% 3.10%International Equity MSCI World ex USA 8.45% 6.75% 4.50% 19.70% 3.25%Emerging Markets Equity MSCI Emerging Markets 10.50% 7.00% 4.75% 27.45% 2.65%

Fixed IncomeShort Duration Bloomberg Barclays 1-3 Yr G/C 2.60% 2.60% 0.35% 2.10% 2.85%Domestic Fixed Bloomberg Barclays Aggregate 3.05% 3.00% 0.75% 3.75% 3.50%Long Duration Bloomberg Barclays Long G/C 3.75% 3.20% 0.95% 10.90% 4.50%TIPS Bloomberg Barclays TIPS 3.10% 3.00% 0.75% 5.25% 3.35%High Yield Bloomberg Barclays High Yield 5.20% 4.75% 2.50% 10.35% 7.75%Non-US Fixed Bloomberg Barclays Glbl Agg xUSD 1.80% 1.40% -0.85% 9.20% 2.50%Emerging Market Debt EMBI Global Diversified 4.85% 4.50% 2.25% 9.60% 5.75%

OtherReal Estate Callan Real Estate Database 6.90% 5.75% 3.50% 16.35% 4.75%Private Equity TR Post Venture Capital 12.45% 7.35% 5.10% 32.90% 0.00%Hedge Funds Callan Hedge FoF Database 5.35% 5.05% 2.80% 9.15% 2.25%Commodities Bloomberg Commodity 4.25% 2.65% 0.40% 18.30% 2.25%Cash Equivalents 90-Day T-Bill 2.25% 2.25% 0.00% 0.90% 2.25%

Inflation CPI-U 2.25% 1.50%

* Geometric returns are derived from arithmetic returns and the associated risk (standard deviation).


Correlations

● Low correlations reduce portfolio risk

Diversification

Correlation Matrix 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 191 Broad Domestic Equity 1.002 Large Cap 1.00 1.003 Small/Mid Cap 0.97 0.94 1.004 Global ex-US Equity 0.87 0.87 0.84 1.005 International Equity 0.84 0.84 0.80 0.99 1.006 Emerging Markets Equity 0.87 0.86 0.85 0.94 0.87 1.007 Short Duration -0.25 -0.24 -0.27 -0.27 -0.25 -0.29 1.008 Domestic Fixed -0.11 -0.10 -0.14 -0.13 -0.11 -0.16 0.87 1.009 Long Duration 0.13 0.14 0.12 0.10 0.12 0.07 0.73 0.93 1.0010 TIPS -0.05 -0.05 -0.08 -0.05 -0.03 -0.09 0.53 0.60 0.53 1.0011 High Yield 0.64 0.64 0.61 0.63 0.61 0.62 -0.14 0.02 0.22 0.06 1.0012 Non-US Fixed 0.01 0.05 -0.10 0.01 0.06 -0.09 0.48 0.51 0.54 0.34 0.12 1.0013 EMD 0.57 0.57 0.56 0.58 0.55 0.58 -0.04 0.10 0.16 0.18 0.60 0.01 1.0014 Real Estate 0.73 0.73 0.71 0.68 0.66 0.65 -0.17 -0.03 0.19 0.00 0.56 -0.05 0.44 1.0015 Private Equity 0.95 0.95 0.92 0.93 0.90 0.91 -0.26 -0.20 0.02 -0.11 0.64 -0.06 0.57 0.72 1.0016 Hedge Funds 0.80 0.80 0.77 0.76 0.73 0.76 -0.13 0.08 0.30 0.08 0.57 -0.08 0.54 0.61 0.78 1.0017 Commodities 0.15 0.15 0.15 0.16 0.16 0.16 -0.22 -0.10 -0.04 0.12 0.10 0.05 0.19 0.20 0.18 0.21 1.0018 Cash Equivalents -0.04 -0.03 -0.08 -0.04 -0.01 -0.10 0.30 0.10 -0.05 0.07 -0.11 -0.09 -0.07 -0.06 0.00 -0.07 0.07 1.0019 Inflation -0.01 -0.02 0.02 0.01 0.00 0.03 -0.20 -0.28 -0.29 0.18 0.07 -0.15 0.00 0.10 0.06 0.20 0.40 0.00 1.00


Asset Mixes

● The current and target mixes are expected to return just under 6.6%

● The alternative asset mixes have the lowest risks for their targeted rates of return (“optimal”) – Mix returns range from 6% to 7% in 0.25% increments – There is a 1% minimum for cash but all of the other asset classes are unconstrained

● The current and target mixes have allocations, returns and risks between mixes 3 and 4

Current and Range of Alternative Asset Mixes


Range of Returns

● The exhibit above shows the range forecast returns for any given year in the 10-year horizon – The center dotted line is the median asset value while the bottom of the bar is the 95th percentile – The range is created by 5,000 simulations based on the returns, risks and correlations shown earlier

● The positions and sizes of the bars depend on forecasted asset class performances

● Asset mixes are below a 50% probability of achieving the 7.75% return goal in any one year

● All of the asset mixes have more than a 5% probability of a double digit one-year loss

Any Single Year in the Next Decade

Current Target M ix 1 M ix 2 M ix 3 M ix 4 M ix 5(30%)(20%)(10%)

0%10%20%30%40%50%

of R

etur

n (%

)A

nnua

l Rat

es

Average

5th Percentile25th PercentileM edian75th Percentile95th Percentile

Prob > 7.75%

6.84%

36.14%17.98%

6.74%(3.65%)

(16.71%)

47.5%

6.86%

36.32%18.01%

6.82%(3.69%)

(16.78%)

47.6%

6.16%

28.15%14.68%

6.08%(1.94%)

(12.26%)

44.3%

6.45%

31.29%16.00%

6.33%(2.53%)

(13.89%)

45.8%

6.74%

34.55%17.43%

6.68%(3.24%)

(15.65%)

47.1%

7.05%

38.35%18.95%

7.06%(4.05%)

(17.86%)

48.3%

7.38%

42.79%20.80%

7.43%(5.00%)

(20.21%)

49.3%

7.75%48 48 44 46 47 48 49


Range of Returns

● The ten-year ranges of return result from continuing the 5,000 simulations for a full 10 years

● The median return is lower over ten years than over any one year due to “volatility drag”

● The probability of achieving the return target falls to about 40% for the current, target and mix 3

● The annualized range of returns is compressed over 10 years relative to one year – Very good or poor returns in any one year are unlikely in other years diluting the extreme outcomes – There is still more than a 5% probability of a negative annualized return over the entire time period

10 Years, Annualized

Current Target M ix 1 M ix 2 M ix 3 M ix 4 M ix 5(30%)(20%)(10%)

0%10%20%30%40%50%

of R

etur

n (%

)A

nnua

l Rat

es

Average


Prob > 7.75%

6.57%

15.28%9.98%6.53%3.21%

(1.46%)

40.8%

6.58%

15.31%10.02%

6.56%3.21%

(1.49%)

41.0%

6.00%

12.55%8.62%5.98%3.47%

(0.16%)

32.3%

6.25%

13.60%9.18%6.22%3.40%

(0.59%)

36.5%

6.50%

14.74%9.75%6.47%3.31%

(1.15%)

39.8%

6.75%

16.01%10.38%

6.70%3.21%

(1.75%)

42.7%

7.00%

17.45%11.12%

6.93%2.99%

(2.45%)

45.1%

7.75%41 41 32 36 40 43 45


Range of Returns

● The exhibit above is based on the same data as the prior exhibits but with unannualized returns

● Over the next ten years the assets would be expected to grow by from about 80% to 95% in the median outcomes if cash flows are not considered

● Double digit nominal cumulative losses are possible in poor markets. – Not only are these absolute losses but they also represent considerable underperformance relative to the

expected returns

10 Years, Cumulative

Current Target M ix 1 M ix 2 M ix 3 M ix 4 M ix 5(100%)

0%

100%

200%

300%

400%

500%

Cum

ulat

ive

Rat

es o

f R

etur

n (%

)


314.4%158.8%

88.3%37.2%

(13.7%)

315.6%159.9%

88.8%37.1%

(14.0%)

226.2%128.7%

78.8%40.6%(1.6%)

257.8%140.6%

82.8%39.7%(5.8%)

295.4%153.4%

87.1%38.6%

(10.9%)

341.6%168.5%

91.3%37.1%

(16.2%)

399.4%186.9%

95.5%34.3%

(22.0%)


Disclaimers

This report is for informational purposes only and should not be construed as legal or tax advice on any matter. Any decision you make on the basis of this content is your sole responsibility. You should consult with legal and tax advisers before applying any of this information to your particular situation.

This report may consist of statements of opinion, which are made as of the date they are expressed and are not statements of fact.

Reference to or inclusion in this report of any product, service or entity should not be construed as a recommendation, approval, affiliation or endorsement of such product, service or entity by Callan.

Past performance is no guarantee of future results.

The statements made herein may include forward-looking statements regarding future results. The forward-looking statements herein: (i) are best estimations consistent with the information available as of the date hereof and (ii) involve known and unknown risks and uncertainties such that actual results may differ materially from these statements. There is no obligation to update or alter any forward-looking statement, whether as a result of new information, future events or otherwise. Undue reliance should not be placed on forward-looking statements.

Funding Public Pensions - Mississippi Workshop/2017...Funding Public Pensions •...

Documents

Transcript of Funding Public Pensions - Mississippi Workshop/2017...Funding Public Pensions •...