By: Tia Goss Sawhney*
In the Illinois public pension world risk is the devil seldom acknowledged, let alone carefully considered. Everyone not living under a rock knows that Illinois has a public pension plan problem – one that poses an immense financial challenge. The public pension plan challenge, however, is overwhelming when the massive risk, currently born by the state, is carefully considered.
Actuarial calculations for pension plans are based on assumptions with respect to events that will occur over the next several decades. An actuarial valuation requires assumptions with respect to future investment earnings, future mortality, future retirement rates, and more. While some assumptions are more likely to be significantly wrong than others EVERY assumption will ultimately be wrong. No one, even the best actuary, has a crystal ball even for next year, let alone decades from now.
Perhaps the biggest fallacy is that if a pension plan is “big enough”, “everything” will work out “on average”, “over time”. That’s nonsense. A big, well-diversified plan, over time is less subject to the vagaries of individual events, such as the impact of owning equity in single company that goes bankrupt or the impact of a particularly bad flu season, than a smaller plan over a shorter time period. Size, diversification, and time, however, are not protective against systemic downside risks, such as economic recessions, deflation, mortality improvements, and even an unexpected surge in public pension plan retirements.[i]
Contributing to the fallacy is the frequent use of the word “expected value” when describing the selection of actuarial assumptions. In statistics “expected” is equivalent to “best estimate”. Statisticians understand that there is significant uncertainty with respect to expected, best estimate values. When used in the colloquial, however, “an expectation” is considered to be much more certain than “an estimate”. Actuaries would better communicate public pension plan risk if they referred to assumptions as estimates rather than expectations.
As we all know, the future, with all of its risk and vagaries, will unfold as it unfolds, irrespective of what we estimate today. Likewise, the ultimate cost of a pension plan will be the ultimate cost of future pension plan benefit payments, irrespective of today’s estimates of the liabilities associated with those payments. Yet there is pressure for pension plans to estimate a higher investment return rate in order to lower today’s estimates of pension liabilities.
It is not uncommon for actuaries and others to make statements such as “If it is assumed that assets will earn a higher amount of investment income over the years to come, then less money needs to be set aside to meet the future benefit obligation. The value today of future benefits, the liability, is lower.”[ii]
The quote, from a leading actuary[iii], is correct EXCEPT that it is missing a critical additional sentence. The sentence should be along the lines of “However, the higher the assumed investment income and therefore the lower today’s estimated liability, the greater the risk that the amount set aside for the pension payments will be too little and that a future generation will need to make up the difference in order to deliver the guaranteed benefit payments.”
The additional sentence then raises the questions of how often the estimated liability of the pension plans will be too low and by how much – questions regarding how much risk the State has assumed. You won’t find the answers to these questions within the documents posted on the websites of the Illinois pension plans. Multiple Freedom of Information Act (FOIA) requests are required. The State keeps the devil out of sight.
I have made FOIA requests of TRS (Teachers’ Retirement System) and SERS (State Employees’ Retirement System), plans that respectively account for about 60% and 20% of Illinois’ public pension plan liabilities. SURS (Illinois State University Retirement System) accounts for most of the remaining liability. Because FOIA requests and the review the resulting non-standardized documents is a painful process, I have not yet gathered similar documents from SURS. The following is what I have learned with respect to investment return assumptions.[iv] (The FOIA’d documents that I use for this discussion are available via WirePoints links found in the endnotes.)
TRS currently assumes a 7.5% rate of return for purposes of estimating liabilities.[v] TRS actuaries estimate that while the mean compound asset return over the next 20 years will be 8.7%, 25% of the time the return over the next 20 years will be 6.6% or less.[vi] Investment advisors engaged by TRS are less optimistic than the actuaries with respect to their projections and project that the mean long term asset return will be 6.9%, implying that 50% of the time the long term return will be 6.9% or less and that well more than 50% of the time the return will be less than the assumed 7.5% rate of return.[vii]
SERS (and also SURS) currently assumes a 7.25% rate of return for purposes of estimating liabilities.[viii] SERS actuaries estimate that the mean return over 30 years will be 6.8%, that nearly 60% of the time the 7.25% return will not be achieved over 30 years, and that 25% of the time the 30 year return will be less than 5.4%.[ix]
While the above estimates are inconsistent (not surprisingly as we expect that people will have different estimates of the future), the estimates should be enough to convince anyone that there is a significant risk that today’s investment rate return assumptions of 7.5% (TRS) and 7.25% (SERS and SURS) could be 1.5% too high over the next two or three decades. What is the impact of a 1.5% miss?
Huge! As of June 30, 2014 TRS, SERS, and SURS had $78 billion of assets.[x] A 1.5% miss is worth $1.2 billion for just this year – a non-trivial percentage of the already severely strained state budget. Furthermore, since pension assets are projected to increase faster than the state budget due to the pay-down of the unfunded liability, the strain created by a consistent 1.5% miss will get proportionally bigger over time.
Should a 1.5% long-term miss occur, however, it most definitely will not emerge consistently over time. It will likely occur via one or several significant market downturns, downturns that may or may not be proceeded or followed by significant market upturns. For example, the State lost $14 billion in fiscal year 2009 instead of earning the assumed (estimated) $6 billion return – closing the year with $20 less than the assumed market return. On the upside, the State earned nearly $6 billion more than the assumed market return in each of fiscal years 2007 and 2011.[xi]
The pension funds, invested as they are in 75% equities, derivatives, and real estate[xii], are subject to sometimes dramatic investment return ups and downs that, over time, will create compounded returns that are higher or lower than estimated. It is rather like playing at the casinos – with a given play you may win or lose a substantial amount and, while losses and winnings may somewhat smooth out over many plays, at any given point you are guaranteed to be up or down and never exactly even.
While responsible gamblers only gamble what they can afford to lose, the State seldom acknowledges pension plan risk and does not have a plan or the apparent capacity to absorb the risk of losses while maintaining guaranteed benefits.
The lack of a plan is apparent in the State’s defense of SB1, the pension reform bill passed in 2013 that is now on appeal to the Illinois Supreme Court. Entering fiscal year 1998 all three plans (TRS, SERS, and SURS) assumed a long-term 8.5% investment rate of return. In the interim 17 years, the plans collectively achieved just over a 7.0% investment rate of return – approximately 1.5% less than expected.[xiii]
The State’s defense of SB1 is built around the central argument that everything was under control in 1997 and that the State has no choice but to use police powers to reduce the benefits of current and future retirees because of negative returns in fiscal years 2008 and 2009 – even though the 17 year 7.0% return inclusive of these years is well within the probable range of results for a portfolio with an estimated 8.5% long term return.
In defending SB1, neither the State, nor its expert actuary, acknowledge the massive investment risks assumed by large defined benefit pension plans. Instead the State describes the fiscal year 2008 and 2009 market losses and the impact of the losses on the plans that were (and still are[xiv]) invested 75% in equities, derivatives, and real estate as “extraordinary adverse events”, “large, unforeseen”, “unanticipated and substantial”, “unexpected and unanticipated”, and “unforeseen and unintended”.[xv] Supporting the State’s assertions, the actuarial expert categorizes the 2008 and 2009 market losses and their impact as “unexpected and devastating external circumstances”, “significant unexpected developments”, and “unexpected and unanticipated effects”.[xvi]
These statements are at best misleading. The possibility of a significant market downturn resulting in market losses for a portfolio invested 75% in equities, derivatives, and real estate was and is foreseen, anticipated, and expected. And although the 2008 and 2009 fiscal year impact was substantial and somewhat extraordinary, the impact, when bundled with 15 other years of returns, it not extraordinarily adverse compared to other possible long-term results. The financial impact is devastating because the State knowingly took risks that it was not prepared to absorb. The State continues to take the same risks.
I am an actuary and, as an actuary, I am particularly disappointed by the actuarial expert’s statements. Actuaries are trained to understand, evaluate, and communicate risk. The other actuaries employed by the State acknowledge investment risk.[xvii]
So, are you scared of the risk devil yet? Regardless of whether you are reading this essay primarily from a taxpayer, pension plan participant, or bondholder perspective, you should be. The fear likely makes you want to think about other things. Ignoring a fearsome devil, however, doesn’t make it go away.
Reducing benefits via legislation police powers when risks don’t pay off is not a solution. We won’t emerge from the Illinois public pension mess until we tame the devil of risk by assessing and valuing plan risks, assuming only the risks we are prepared to bear, and distributing the resulting gains and losses equitably. The process starts with acknowledging risk.
*Tia Goss Sawhney, DrPH, FSA, MAAA. Dr. Sawhney is a qualified health insurance actuary with an active interest in Illinois public pension plans; she is not a pension actuary. These opinions are her own. She may be reached at email@example.com.
[i] For discussions of public pension plan risks see 1) Report of the Blue Ribbon Panel on Public Pension Funding, Society of Actuaries, February 2014, https://www.soa.org/blueribbonpanel/ and 2) The Blinken Report: Strengthening the Security of Public Sector Defined Benefit Plans, Donald J. Boyd and Peter J. Kiernan, The Rockefeller Institute of Government, January 2014, http://www.rockinst.org/pdf/government_finance/2014-01-Blinken_Report_One.pdf.
[ii] Expert Report in Defense of Illinois PA-98-599 (aka, SB1), as filed by the IL Attorney General’s Office, by Thomas S. Terry, Actuary, revised September 26, 2014, page 10, paragraph 55, italics as they appear in the report.
[iii] As noted in his Declaration, at the time that he prepared the expert report, Tom Terry was President of the American Academy of Actuaries.
[iv] Since the investment rate of return assumption is the single-most critical assumption in valuing public pension plan liabilities, I have focused on the investment rate of return for this essay. There is, however, a similar story to tell with respect to other risks, particularly mortality. I may discuss mortality risk in another essay.
[v] State Actuary’s Report, prepared by Cheiron, for Auditor General, December, 2014, http://trs.illinois.gov/pubs/actuarial/2014-State-Actuary-Rpt.pdf, page 7.
[vii] Asset/Liability Study for Teachers’ Retirement System of the State of Illinois, by RVK, April 2014, page 25, “Return (compound)” line, “Long Term Target Portfolio” column.
[viii] State Actuary’s Report, citation above, page 7.
[ix] 2014 Experience Review, for State Employee’s Retirement System of Illinois, by Gabriel Roeder Smith & Company (GRS), April 2014, WirePointsLink, page 11.
[xii] See Comprehensive Annual Financial Reports for each of the plans and also: Public Pensions Need Gamblers Anonymous, by Andrew G. Biggs, Wall Street Journal, December 2, 2014, http://www.wsj.com/articles/andrew-g-biggs-public-pensions-need-gamblers-anonymous-1417563447.
[xiii] Excel Spreadsheet ReconcileTTExpertReport 2-1-2015, cited above.
[xiv] See historical and current Comprehensive Annual Financial Reports for each of the plans.
[xv] Memorandum in Support of Defendant’s Motion for Summary Judgment, Filed in the Circuit Court for the Seventh Judicial Circuit, Sangamon County, IL, by Lisa Madigan, Attorney General, October 2, 2014. Respectively: page 1 (first 4 quotes), page 10, and section header for pages 34-36.
[xvi] Expert Report, citation above. Respectively: page 3, paragraph 12; page 45, header 9; page 45, table 21 header.
[xvii] As described above actuaries with GRS and Buck acknowledge investment risk. Furthermore, actuaries with Cheiron, the actuarial consulting firm engaged by the Illinois Auditor General, encourage the plans to carefully consider risk associated with significant market downturns. State Actuary’s Report, citation above, page 2.