Preface: Earlier this year the California Policy Center published a study evaluating the Orange County Employee Retirement System (OCERS) to explore this same question: Are Annual Contributions into OCERS Adequate? That study adopted a unique focus, evaluating contributions into OCERS not based on percent of payroll, but by looking at the actual amount of cash being contributed each year. In particular, the study evaluated how much cash each year was being contributed to reduce the unfunded liability. This report performs the exact same analysis, using the exact same template. Different numbers; same story. Pension analysts and pension activists are encouraged to download the spreadsheets (CalSTRS, OCERS) used in both of these studies, and use them to perform similar analysis for whatever pension systems they are concerned about. For whatever pension fund they choose to analyze, it is quite likely they will find that the amount of money being contributed to reduce the unfunded liability is alarmingly low.
Summary: For the fiscal year ended 6-30-2012 the California State Teachers Retirement System, CalSTRS, collected $5.8 billion from employees and employers to invest in their pension fund. Of this $5.82 billion, $4.7 billion was the so-called “normal contribution,” which was a payment to cover the present value of future pensions earned during FYE 6-30-2012 by actively employed participants. The other $1.1 billion that was collected and invested in the fund was a “catch-up” payment to reduce the unfunded liability, which as of 6-30-2012 was officially estimated to be $71.0 billion.
Using evaluation formulas and unfunded liability payback terms recommended by Moody Investor Services in April 2013, this study shows that if the “catch-up” payment is calculated based on a level payment, 20 year amortization of the $71.0 billion unfunded liability – still assuming a 7.5% rate-of-return projection – this catch-up payment should be $7.0 billion per year. The study also shows that if the CalSTRS pension fund rate-of-return projection drops to 6.20%, the unfunded liability recalculates to $107.8 billion and the catch-up payment increases to $9.6 billion per year. At a rate-of-return projection of 4.81%, the unfunded liability recalculates to $154.9 billion and the catch-up payment increases to $12.2 billion per year.
This study also finds that the $4.7 billion normal contribution into CalSTRS for FYE 6-30-2012 was based on a rate-of-return assumption of 7.5%. The study shows that lowering the rate-of-return projection from 7.50% to 6.20% would require the normal contribution to increase by another $1.1 billion; lowering it from 7.50% to 4.81% would require the normal contribution to increase by another $2.5 billion. The rate of 6.2% represents the historical performance of U.S. equity investments (including dividends) between 1900 and 1999. The rate of 4.81% is the Citibank Pension Liability Index rate as of July 2013, which is the lower risk rate recommended by Moody’s Investor Services for pension liability forecasting.
The conclusion of this study is that CalSTRS relies on optimistic long-term earnings projections and very aggressive unfunded liability repayment schedules in order to pay the absolute minimum into their pension fund. If CalSTRS is required to even incrementally lower their rate-of-return projections – something that market conditions may eventually dictate – their funded ratio which is already only 67.0% will fall precipitously. Unless extremely favorable market conditions occur for the next several years without interruption, in order for CalSTRS to remain solvent, they need to dramatically cut retirement benefits, or increase their annual contributions by 50% or more per year.
* * *
The purpose of this brief study is to assess whether or not the $5.82 billion contributed during the fiscal year ended 6-30-2012 into the CalSTRS pension fund was sufficient to ensure the long-term solvency of the fund. Using methods recommended in July 2012 and finalized in April 2013 by Moody’s Investor Services, and elucidated in a recent CPPC study “A Method to Estimate the Pension Contribution and Pension Liability for Your City or County,” this study will perform what-if scenarios, estimating the size of the CalSTRS unfunded liability at various rates of return.
This study will also discuss whether or not the current contributions, both normal and catch-up, add sufficient assets to the CalSTRS fund to ensure solvency, and how much contributions would need to increase using more conservative assumptions regarding return on investment and payback periods. Finally, throughout this study an attempt will be made to discuss and explain key concepts of pension finance, in keeping with the CPPC’s mission to inform and involve more people in discussions of sustainable public finance. The tables in this study were produced on an Excel spreadsheet that can be downloaded here:
DOWNLOAD SPREADSHEET: Analysis-of-CalSTRS-Pension-Liability-and-Contributions.xlsx
The officially recognized amounts for key variables used in this study, including the total contribution to the pension fund, normal contribution, “catch-up” contribution to reduce the unfunded liability, as well as the total assets, total liabilities, and unfunded pension fund liability, were all drawn from publicly available CalSTRS financial reports.
* * *
HOW MUCH WAS CalSTRS UNDERFUNDED AS OF 6-30-2012?
The underfunding of any pension fund is calculated by subtracting from the amount of the total invested assets the present value of the liability to pay pensions in the future. Because these future pension obligations are intended to be paid sometime between next year and 50-60 years from now, the liability today is calculated by adding up the net present values of the estimated payments to be made for each year in the future. If the total value of the pension fund’s invested assets is equal to the present value of all future pension payments, the fund is said to be 100% funded.
Using data from the CalSTRS “Defined Benefit Program Actuarial Valuation as of June 30, 2012,” here are the officially recognized amounts for CalSTRS assets, liabilities, and underfunding at the end of last year :
- Actuarial accrued liability (AAL) = $215.19 billion
- Valuation value of assets (VVA) = $144.23 billion
- Unfunded Actuarial Accrued Liability = $70.96 billion
* * *
HOW MUCH WAS CONTRIBUTED INTO CalSTRS IN FYE 6-30-2012?
Again using data from CalSTRS “Comprehensive Annual Financial Report, FYE 6-30-2012,” the total contribution into the pension fund in 2012, via direct payments by participating employers and withholding from participating employee paychecks, was $5.82 billion .
Annual pension fund contributions have two components, the “normal contribution,” and the “unfunded contribution,” which is the amount paid towards reducing the plan’s unfunded liability. The “normal contribution,” simply stated, needs to be an amount equal to the present value of future pension payment obligations earned in the current year.
Normal Contribution: The total employer and employee normal cost for FYE 6-30-2012 was $4.69 billion as documented on page 15 of the Milliman actuarial report for CalSTRS, “Defined Benefit Program Actuarial Valuation.” .
Unfunded Contribution: The amount paid into the CalSTRS pension fund during their FYE 6-30-2012 to reduce their unfunded liability is not explicitly disclosed on official documents. But logic dictates the total unfunded contribution to be the difference between the total contribution, $5.82 billion, and the normal contribution, $4.69 billion, or $1.13 billion.
- Total contribution = $5.82 billion
- Normal contribution = $4.69 billion
- Unfunded contribution = $1.13 billion
The remainder of this report will consider whether or not this level of contributions is adequate by estimating required contributions based on more aggressive payback terms, as well as more conservative rate-of-return projections. The first step is performing these what-ifs is to estimate how the unfunded liability is affected by changes to the rate-of-return projections.
* * *
HOW DO LOWER RATES OF RETURN AFFECT CalSTRS UNFUNDED PENSION LIABILITY?
To be 100% funded, a pension plan must have invested assets equal to the present value of all future pension payments. For every participating employee, whether they are active or retired, actuaries estimate their salary growth, their year of retirement, their initial pension, their subsequent pension payouts based on COLAs, and their life expectancy. The present value of all these future payments is how much a fully funded pension plan’s assets must be worth.
The rate at which these future payments are discounted per year must be equivalent to whatever rate the pension fund managers believe they will earn interest, on average, over the life of the fund. The theory is that if no future work were performed, and no future pension benefits were earned, and no additional money were contributed to the fund, the assets currently invested in a fully funded plan would earn enough interest to support every future pension payment until the last participant died of old age.
Since the amount of assets in a pension plan is a known, objective quantity, the debate over how much unfunded liability a plan may have centers on what assumptions are used to estimate the present value of the future payments, i.e., the “Actuarial accrued liability (AAL),” which CalSTRS valued as of 6-30-2013 at $215.2 billion. In order to assess whether or not that amount is overstated or understated, we can use a short-cut formulated by Moody’s Investor Services in their July 2012, “Moody’s Adjustments to US State and Local Government Reported Pension Data.”
In order to revalue a pension fund’s liabilities without having access to every actuarial calculation from every fund, what Moody’s does is estimate the midpoint of the future payments stream. They select 13 years into the future, which is quite conservative. Using a longer duration than 13 years will greatly increase the sensitivity of the liability to changes in the projected rate-of-return. Here is their rationale :
“To implement the discount rate adjustment, we propose using a common 13-year duration estimate for all plans. This is a measure of the time-weighted average life of benefit payments. Each plan’s reported actuarial accrued liability (“AAL”) is projected forward for 13 years at the plan’s reported discount rate, and then discounted back at 5.5%. This calculation results in an increase in AAL of roughly 13% for each one percentage point difference between 5.5% and the plan’s discount rate. For example, a plan with a $10 billion reported AAL based on a discount rate of 8% would have an adjusted AAL of $13.56 billion, or 35.6% greater than reported.
We recognize this duration estimate may be higher than warranted for some plans and lower than warranted for others. Each pension plan has a unique benefit structure and demographic profile that affects the time-weighted profile (duration) of future benefit payment liabilities. However, plan durations are not reported, and calculating duration individually for each plan is not feasible. Our proposed 13-year duration is the median calculated from a sample of pension plans whose durations ranged from about 10 to 17 years. Plans with shorter durations usually are closed or have a preponderance of older or retired members.”
Here is the formula that governs this readjustment:
Adj PV = [ PV x ( 1 + official %i ) ^ years ] / ( 1 + adjusted %i ) ^ years
Table 1, below, shows how much the CalSTRS unfunded liability will increase based on two alternative rate-of-return projections, both of which are lower than the official rate of 7.50% currently used by CalSTRS. Here they are:
Case 1, 6.2% – here is the rationale for this rate-of-return, excerpted from the report “Pension Math: How California’s Retirement Spending is Squeezing The State Budget”  authored by Joe Nation, a Ph.D., Stanford Institute for Economic Policy Research, and former California Democratic assemblyman: “This 6.0 to 6.5 percent figure is based on the performance of a hypothetical fund containing 80 percent equity and 20 percent income instruments between 1900 and 1999. It assumes an equity rate based on the 20th-century Dow Jones industrial annual average of 5.3 percent, plus 2 percent in dividends, less 0.5 percent in fees. Combined with income instruments with a net rate of return of 4.5 percent, this hypothetical fund would have earned an average annual rate of 6.2 percent.”
Case 2, 4.81% – the rationale for this rate-of-return comes from Moody’s Investor Services “Moody’s Revised New Approach to Adjusting Reported State and Local Government Pension Data,” released in April 2013 : “For adjustments to pension data, Moody’s will use the Citibank Pension Liability Index (Index) posted on the date of the valuation instead of its original proposal to apply a single rate for all plans each year.” Citigroup Pension Discount rate as posted by the Society of Actuaries in July 2013 was 4.81% .
Table 1 makes the revaluation method specified by Moody’s transparent. To download the spreadsheet that contains all of these tables, click here: Analysis-of-CalSTRS-Pension-Liability-and-Contributions.xlsx, and refer to the first tab “unfunded liability.” To make the logic of these calculations as plain as possible, the spreadsheet cells where assumptions are input are shaded in yellow, and the result cells at the bottom are shaded in green. Column 1 shows the baseline case, using the official projected interest rate of 7.50%, meaning the end result is unchanged. Column two uses the “case 1″ lower rate of 6.20%, column three uses the “case 2″ rate of 4.81%. The first three rows show how the officially reported unfunded liability is calculated. The first four rows of the second second section, “Revaluation of Unfunded Pension Liability,” projects the liability forward 13 years to develop a future value at the official rate of return, 7.50%. The final three rows of the second section then calculate the present value using the baseline rate of 7.50%, the case 1 rate of 6.20%, and the case 2 rate of 4.81%.
As can be seen, even with what is probably an unrepresentative duration of only 13 years, if the pension fund is only projected to earn 6.2% instead of 7.50%, the unfunded liability estimate jumps from $70.0 billion to $107.8 billion, and at a projection of 4.81%, more than doubles to $154.93 billion. These are not implausible scenarios.
TABLE 1 – RECALCULATING CalSTRS UNFUNDED PENSION LIABILITY
* * *
HOW DO LOWER RATES OF RETURN AFFECT CalSTRS UNFUNDED CONTRIBUTION?
Table 2, below, shows how much CalSTRS should be paying back each year to reduce their unfunded liability if they were to pay it back using the terms of a conventional amortized loan with level payments each year. And this “level payment” method is what has been adopted by Moody’s Investor Services in their Moody’s Revised New Approach to Adjusting Reported State and Local Government Pension Data, “Moody’s adjusted net pension liability will be amortized over a 20-year period on a level-dollar basis using the interest rate provided by the Index.” 
When reviewing Table 2, bear in mind that the payment made in FYE 6-30-2012 into the CalSTRS pension fund towards reducing their unfunded liability was $1.13 billion. As the baseline case in column 1 shows, if this payment were calculated using a level payment, 20 year term as recommended by Moody’s Investor Services, they would have had to pay $6.96 billion during 2012, more than six times as much. This observation merits repetition: By applying repayment terms that Moody’s Investor Services – the largest credit ratings agency in the world – has recommended public sector pension funds adopt, and without changing the return-on-investment assumptions that many analysts (including Moody’s who recommend using the Citibank Pension Index rate which typically is under 5.0%), CalSTRS is underpaying their unfunded contribution by a factor of more than six times.
Columns 2 and 3 in Table 2 help illustrate why it isn’t excessive to specify a 20 year, level payment plan to eliminate a pension plan’s unfunded liability. Because the amount of the unfunded pension liability, the principle being repaid, will fluctuate according to how much fund managers think the invested assets are going to earn. And as Table 1 shows, the amount of the unfunded liability is extremely sensitive to these changes.
In column 2, case 1 shows the impact of a 6.20% rate of return projection for the CalSTRS pension fund. The unfunded pension liability increases from the official $70.0 billion to $107.8 billion, which more than offsets the using the lower 6.20% rate to calculate the required payments. In case 1, a 6.20% projected rate of return for the CalSTRS pension assets will equate a $9.55 billion annual payment to reduce their unfunded liability. In case 2, a 4.81% projected rate of return for the CalSTRS pension assets will equate a $12.2 billion annual payment to reduce their unfunded liability.
Note: The first table in the Appendix to this report shows more detail on how CalSTRS is currently incurring negative amortization, i.e., how the unfunded pension liability will increase each year if only $1.13 billion is paid annually toward reducing that liability, as well as how the unfunded pension can be eliminated within 20 years using the higher payments calculated using the level payment method.
TABLE 2 – RECALCULATING CalSTRS UNFUNDED CONTRIBUTION
* * *
HOW DO LOWER RATES OF RETURN AFFECT CalSTRS NORMAL CONTRIBUTION?
No discussion of whether or not sufficient funds were contributed to CalSTRS during FYE 6-30-2012 would be complete without considering the “normal contribution,” which was $4.69 billion. The normal contribution is defined as how much future pension obligations were earned by actively employed participants during the current year, in this case, during the 12 month period ended 6-30-2012. If a pension plan is considered to be 100% funded, the normal contribution is the only payment necessary.
Table 3, below, revalues the normal contribution using methods recommended by Moody’s Investor Services. Again, a shortcut of this type is necessary because it is impossible with publicly available information to apply various rates of return to estimated future annual pension payments accrued during 2012 according to the actuarial profile for every actively employed participant in CalSTRS. Instead, the normal payment, representing the present value of the entire stream of future pension payments earned by actively working participants in the most recent year, is recalculated, using the official rate-of-return projection, 7.50%, at a future value set 17 years from now, representing the average remaining service life of active employees. This future value is then discounted back to present value using the lower rate-of-return, in our cases, at 6.20% and 4.81%. This process will yield a higher required normal contribution. Here’s how Moody’s describes this method in their July 2012 proposal “Moody’s Adjustments to US State and Local Government Reported Pension Data.” 
“The ENC [employer’s normal cost] adjustment reflects the lower assumed discount rate and the use of a 17-year active employee duration estimate for all plans – i.e., each plan’s normal cost is projected forward for 17 years at the plan’s reported discount rate, and then discounted back at 5.5% [based on the Citibank Pension Liability Index, as Moody’s specifies, which has dropped subsequently dropped to 4.81% – this study uses two lower rate scenarios, 6.20% and 4.81%], after which employee contributions are deducted to determine the adjusted ENC. The 17-year duration assumption reflects our estimate of the average remaining service life of employees based on a sample of public pension plans. We acknowledge that this is a simplifying assumption that may be too long or too short for different plans. Using this approach, a reported ENC payment of $100 million based on an 8% discount rate would grow to $149 million based on a 5.5% discount rate.”
As case 1 and 2 show on Table 3, lowering the CalSTRS pension fund’s rate-of-return projection from 7.50% to 6.20% increases the normal contribution by $1.1 billion; if it is lowered from 7.50% to 4.81% the normal contribution increases by $2.5 billion.
TABLE 3 – RECALCULATING CalSTRS NORMAL PENSION CONTRIBUTION
* * *
Using evaluation formulas and unfunded liability payback terms recommended by Moody Investor Services in April 2013, this study shows that if the CalSTRS “catch-up” payment is calculated based on a level payment, 20 year amortization of the $70.0 billion unfunded liability – still assuming a 7.50% rate-of-return projection – this catch-up payment should be $6.96 billion per year, rather than the $1.1 billion unfunded payment that was actually made. The study also estimates that if the CalSTRS pension fund rate-of-return projection drops to 6.20%, the unfunded liability recalculates to $107.8 billion and the catch-up payment increases to $9.6 billion per year. At a rate-of-return projection of 4.81%, the unfunded liability recalculates to $154.9 billion and the catch-up payment increases to $12.2 billion per year.
This study also estimates that the $4.7 billion normal contribution into CalSTRS for FYE 6-30-2012, based on a rate-of-return assumption of 7.50%, would have to increase to $5.5 billion based on lowering the rate-of-return assumption to 6.20%. Further, the study shows that by lowering the rate-of-return assumption from 7.50% to 4.81% would require the normal contribution to increase to $6.9 billion.
The 2nd table in the appendix, “Required rate-of-return at various levels of underfunding,” shows why underfunding is a slippery slope by illustrating how much higher rates have to be just to keep the underfunding from getting worse. Pension spokespersons have consistently stated that annual earnings in any mature fund will always greatly exceed annual contributions. On the table, the 5th column “baseline” shows how much earnings have to increase at CalSTRS current official level of 67.02% funded. At that level of funding, as can be seen, just in order for the unfunded liability to not increase, CalSTRS currently has to earn an annual return of 11.2%. At that sustained rate-of-return, the surplus earnings beyond the projected 7.50% do not reduce the unfunded liability at all, they merely prevent it from getting larger.
To recap, for the fiscal year ended 6-30-2012, here are some CalSTRS financial highlights as determined in this study:
- Lowering the earnings projection to 6.20% increases the normal contribution by $1.1 billion per year; lowering it to 4.81% increases the normal contribution by $2.5 billion per year.
- The unfunded “catch-up” contribution of $1.1 billion did not lower the officially recognized unfunded liability of $71.0 billion, in fact, it grew by $4.2 billion (ref. Appendix 1, baseline case).
- If the earnings projection is lowered from 7.50% to 6.20% the unfunded liability increases from $71.0 billion to $107.8 billion; if it is lowered to 4.81% the unfunded liability increases to $154.9 billion.
- At the official return projection of 7.50%, if the unfunded liability is paid back according to 20 year level payment amortization terms, the annual catch-up payment would have been $6.9 billion.
- Using 20 year level payment amortization terms, at a return projection of 6.20% the annual catch-up payment should have been $9.6 billion; at 4.81%, it should have been $12.2 billion.
- An annual return projection of 6.20% represents the long-term appreciation of equities , an annual return projection of 4.81% represents the Citibank pension liability index rate as of July 2013 .
The conclusion of this study is that CalSTRS relies on optimistic long-term earnings projections and very aggressive unfunded liability repayment schedules in order to contribute the absolute minimum each year into their pension fund. As a result, their unfunded liability increased during FYE 6-30-2012 by over $4.0 billion. If CalSTRS is required to even incrementally lower their rate-of-return projections further – something that market conditions may eventually dictate – their funded ratio which is already only 67.0% will fall precipitously. Unless extremely favorable market conditions occur for the next several years without interruption, in order for CalSTRS to remain solvent, they need to dramatically cut retirement benefits, or increase their annual contributions by 50% or more per year.
* * *
(1) CalSTRS Defined Benefit Program Actuarial Valuation as of June 30, 2012, page 32, Table 8
(2) CalSTRS “Comprehensive Annual Financial Report, FYE 6-30-2012, page 41, “Statement of Changes in Fiduciary Net Assets.”
(3) Milliman: Defined Benefit Program Actuarial Valuation, as of June 30, 2012, page 15, Section 4 “Actuarial Obligation, Normal Cost.”
(4) Moody’s Adjustments to US State and Local Government Reported Pension Data, request for comment, July 2nd, 2012, page 6.
(5) Pension Math: How California’s Retirement Spending is Squeezing The State Budget, Stanford Institute for Economic Policy Research, page 13, December 2011.
(6) Moody’s Revised New Approach to Adjusting Reported State and Local Government Pension Data,” ref. “The Adjustments,” item 1, April 2013.
(7) Citigroup Pension Discount Curve, Society of Actuaries, July 2013
(8) Moody’s Revised New Approach to Adjusting Reported State and Local Government Pension Data, ref. “The Adjustments,” item 3, April 2013.
(9) Moody’s Adjustments to US State and Local Government Reported Pension Data, ref. page 8, section four, part one “New discount rate applied to normal cost.”
* * *
APPENDIX – TABLE 1 – UNFUNDED LIABILITY AMORTIZATION SCENARIOS
APPENDIX – TABLE 2 – REQUIRED RETURN AT VARIOUS LEVELS OF UNDERFUNDING