Public Pension Oversight Board




<MeetMDY1> March 27, 2017


Call to Order and Roll Call

The<MeetNo2> 3rd meeting of the Public Pension Oversight Board was held on<Day> Monday,<MeetMDY2> March 27, 2017, at<MeetTime> 1:00 PM, in<Room> Room 169 of the Capitol Annex. Senator Joe Bowen, Chair, called the meeting to order, and the secretary called the roll.


Present were:


Members:<Members> Senator Joe Bowen, Co-Chair; Senators Jimmy Higdon, Christian McDaniel, Dennis Parrett, and Wil Schroder; Representatives Ken Fleming, James Kay, Jerry T. Miller, Arnold Simpson, and Russell Webber; J. Michael Brown, John Chilton, Mike Harmon, James M. "Mac" Jefferson, and Sharon Mattingly.


Guests: Jim Waters, Aaron Ammerman, and William F. Smith, Bluegrass Institute; Brad Gross, LRC; Dolly Guenthner, Retiree.


LRC Staff: Brad Gross, Jennifer Black Hans, Bo Cracraft, and Angela Rhodes.


Approval of Minutes

Senator Higdon moved that the minutes of the February 27, 2017, meeting be approved. Mac Jefferson seconded the motion, and the minutes were approved without objection.


Senator Bowen welcomed Senators McDaniel, Parrett, and Schroder, and Representatives Fleming, Simpson, and Russell, as new members to the Public Pension Oversight Board (PPOB).


Bluegrass Institute

Jim Waters, President, Bluegrass Institute discussed how the Bluegrass Institute could offer information to help with the pension crisis. The Bluegrass Institute is a state-based, free market think tank that was started in 2003 and is a member of the state policy network, which offers free market solutions and ideas to Kentucky’s greatest challenges.


Aaron Ammerman, Investment Analyst, Bluegrass Institute explained that pension reform has been a big part of the Bluegrass Institute’s goal over the last 7 to 8 years. Mr. Ammerman addressed the benefit structure of the pension plans and stated that the Bluegrass Institute is not against defined benefit plans, but if a defined benefit is incorporated into a state structure, there are a set of rules that need to be followed in order to obtain and keep a fully funded status. Over the last 30 years, Kentucky has had multiple rules violated that has resulted in the severe underfunded status of the state plans. Two rules that most likely caused the most damage are benefit enhancements that were given retroactively to employees and retirees and the lack of an actuarial analysis on the benefit enhancements.


William F. Smith, MD, Systems Analyst, Bluegrass Institute discussed the mathematical analysis of the system. He stated that it is important that certain rules are followed with a defined benefit plan, and the pension system has been used improperly. The pension system has been used as a pay-as-you-go system instead of an actuarial reserve system. The systems are designed to support a very specific set of actuarially prefunded benefits that are assigned, earned, and funded to a specific year with payroll contributions. The payroll contributions that are used to prefund these benefits actuarially are called the normal cost. The reason for prefunded benefits is that it offers protection against unfunded liabilities. The scenario is that if the normal cost payroll is made and the actuarial assumptions are achieved, there will be a fully funded pension system. The scenario only applies if there are legitimate and reasonable actuarial assumptions and beneficiaries receive the actuarially prefunded benefits.


Mr. Smith explained that if these benefits are retroactively enhanced after being prefunded, the intended strategy is being defeated. Prefunding means creating an actuarial reserve for these benefits with normal cost payroll contributions, thereby investing that money so it will grow to pay the future benefit obligations. So, if there is failure giving beneficiaries the benefit that is used to calibrate the system, there will no longer be a calibrated system, and there will no longer be an actuarial reserve system. The actuarial assumptions and normal costs are no longer relevant. The system has one set of benefits that are earned and actuarially prefunded and then a second completely separate set of benefits that have been artificially enhanced. The enhanced set of benefits are what people receive when they retire, and there is no prefunding for those benefits. So, now it is more of a pay-as-you-go system instead of an actuarial reserve system.


Mr. Smith discussed actuarial funding of benefits and stated that the benefit factor is the quantitative measure of benefits as a percentage of final compensation and are assigned by legislators and are the defined benefits. The actuaries use that benefit to calibrate the system which gives the normal cost. Making the normal cost payroll contributions as the benefit is being earned is important so there will be an actuarial reserve to pay those benefits in the future. If the benefit is retroactively enhanced, there is no actuarial reserve. The primary funding mechanism for benefits is normal cost and the ARC is there to amortize debt that occurs when the actuarial assumptions are not achieved. The ARC is not intended to be used as a primary funding mechanism for benefits.


There are other elements of benefits that can be enhanced and applied retroactively as actuarially prefunded benefits, such as, enhanced final compensation formula (high3/high5), enhanced final compensation (spiking), excessive COLAs, and health insurance premiums.


Mr. Smith stated the Teachers’ Retirement System’s (TRS) benefit enhancements have a different strategy and do not shuffle benefit factors, but do have benefit enhancements that inflate the value of actuarially prefunded benefits. The benefits that are not covered by the inviolable contract are, the 3 percent factor for years of service over 30, the high 3 final compensation formula at 27 years/age 55, benefit spiking with sick days, pay-as-you-go health insurance.


Mr. Smith discussed KRS 6.350 that was passed in 1980 that requires an actuarial analysis for benefit enhancements. The statute states “A bill which would increase or decrease the benefits…of any state-administered retirement system shall not be reported from a legislative committee of either house of the General Assembly for consideration by the full membership of that house unless the bill is accompanied by an actuarial analysis.” Mr. Smith stated that an actuarial analysis determines how much a benefit enhancement will cost.


Mr. Smith stated the one benefit enhancement that was accompanied by an actuarial analysis was in 1998, SB 142. Further, the final compensation formula was changed from high 5 to high 3, with zero additional funding for two years. Mr. Smith stated that actuary, Steve Gagel, performed the actuarial analysis for SB 142 and was opposed to the legislation because any increase in benefits should reflect a true need or meet a real shortfall in retirement benefits, that it was unclear whether such a shortfall exists, that there is real danger that spendable income after retirement could exceed pre-retirement spendable income, and that it was not an effective use of taxpayer dollars. Mr. Smith said that Steve Gagel thought it was going to cost $300 million over a 30 year period of time and assumed retirement patterns would remain unchanged.


Mr. Smith discussed a couple issues regarding the inviolable contract, stating that it is important to understand based on the wording that the beneficiaries do not contribute to retroactively enhanced benefits and that all the funding comes from future employers. Also, future benefit accrual rates have not been earned and not funded.


Mr. Smith gave recommendations for all plans and stated it is important to enact a constitutional amendment prohibiting retroactive benefit enhancements, providing complete transparency for all benefits received by every retiree and how these benefits were determined, providing that plan governance and board representation should reflect the risk assumed by each stakeholder, and imposing a hard freeze if legitimate reform efforts prove unsuccessful.


Senator Bowen asked if anyone attending the meeting or on the board wished to refute the information the Bluegrass Institute presented to the PPOB with no response.


In response to a question from Representative Kay, Mr. Smith stated that by not paying the ARC exacerbated the problem, but it did not cause the problem. The system design is to avoid an ARC altogether. The ARC is normal cost plus amortized debt and the role of the ARC is to amortize debt as created when the actuarial assumptions are not met, it is not intended to be a primary funding mechanism for the benefit enhancements.


In response to a question from Representative Kay, Mr. Ammerman stated that Bluegrass Institute is an advocate for transparency in all levels of government. TRS’ performance over the last 30 years with extremely low levels of fees is very impressive. KRS performance has lagged the last few years, and KRS is in the process of extracting itself from hedge funds, but the performance is not the reason for the current situation.


In responding to questions from Representative Miller, Mr. Waters stated that the Bluegrass Institute is presenting this information and analysis across the state and will look into whether or not there are court cases about inviolable protection over retroactive benefits.


In response to a question from Senator Higdon, Mr. Smith stated that the debt is being amortized and being charged to future employers as a percentage of their payroll, which make the ARC payments depend on payroll growth assumptions and payroll growth rates.


Payroll Growth Assumptions

Brad Gross, LRC discussed the actuarial data background and stated that with the actuarial valuation there are a lot of components that go into it, such as, actuarial assumptions and methods (including payroll growth), benefits and funding provisions, financial experience, and demographic data and experience. Total amount of benefits paid determines the output of the plan, such as, funding levels, unfunded liabilities, and employer contribution rates. The assumptions are reviewed every five years through an experience study. A secondary review, called the actuarial audit, which is conducted by a different actuary, is also done every five years. An asset liability modeling study is typically performed following the experience study and evaluates various asset allocations against projected system liabilities with the ultimate goal of selecting a target asset allocation for the investment portfolio.


Mr. Gross discussed the funding levels and unfunded liabilities from the 2016 actuarial valuation results. The KERS non-hazardous unfunded liability for pension is -$11.112 billion and -$1.713 billion for retiree health. The County Employees Retirement System (CERS) non-hazardous unfunded liability for pension is -$4.541 billion and -$0.908 billion for retiree health. The TRS unfunded liability for pension is -$14.531 billion and -$2.839 billion for retiree health. For all state-administered pension and retiree health funds the combined total unfunded liability is -$38.652 billion.


Mr. Gross discussed financing unfunded liabilities and stated that in the ARC calculation there is normal cost, which is the anticipated cost of the next year’s benefit, and an amortized unfunded liability payment. For the 2015 valuation, the KERS non-hazardous combined pension and retiree health normal cost was 5.95 percent of pay, the unfunded liability payment was 41.33 percent of pay, for a combined total of 47.28 percent of pay as an employer contribution rate. The amortization method is how the unfunded liabilities are paid off over time. The key components are the amortization period length, whether the period is closed or open, and whether it is financed using the level percent of payroll method or level dollar method. The KRS unfunded liability is amortized over a 30 year closed period using the level percent of payroll method (with an assumed 4 percent payroll growth assumption). The TRS pension unfunded liability is amortized over a 30 year closed period for legacy liabilities as of 2014 with separate 20 year closed amortization periods on new sources of unfunded liabilities. TRS also uses the level percent of payroll method with a 3.5 percent payroll growth assumption. The Judicial Form Retirement System (JFRS) unfunded liability is not impacted by payroll growth. Rather, the JFRS amortization method is a statutory formula of interest on the liabilities plus one percent of the unfunded liability.


Mr. Gross discussed the level dollar versus level percent of pay methods and stated the level dollar method sets the amortization payment as a set dollar amount, meaning no payroll growth is assumed in the future, thus resembling most home loans. In terms of dollars, this method will be more expensive in the short run and less expensive in the long run as compared to the level percent of pay method. The level percent of pay method, which is what KRS and TRS utilize, will be less expensive in the short run, but will require more dollars in the future because the method sets the payment as a percent of total payroll over the amortization period and assumes future payroll growth. In this method a potential problem is negative amortization. LRC did a 50 state peer group survey and 41 states use the level percent of pay method while 9 use the level dollar method. Of those that are using the level percent of pay method, the payroll growth assumption varied between 2.25 percent to 6.5 percent of pay with a median of 3.5 percent of pay.


Mr. Gross discussed how the payroll growth assumption is developed and stated the number ultimately comes from the actuarial experience study. KRS had their last experience study in 2014, which was made effective in the 2015 valuation. TRS had their last experience study in 2016. Typically, the system actuaries use a building block approach that incorporates the assumed inflation rate in the payroll growth assumption. Both systems had an actuarial audit in 2015 by the Segal Company and, in both audits, Segal noted that the 4.0 percent payroll growth assumption used by KRS and TRS at the time was aggressive and detailed that actual payroll growth of the systems was below the assumption for the period being evaluated.


Mr. Gross discussed the payroll growth data for a five year period from 2012 to 2016. KERS non-hazardous payroll has dropped from $1,816.39 million to $1,662.37 million; KERS hazardous increased from $142.85 million to $158.33 million; CERS non-hazardous increased from $2,300.71 million to $2,413.66 million; CERS hazardous increased from $497.46 million to $526.15 million; State Police Retirement System (SPRS) dropped from $51.19 million to $46.69 million; and TRS increased from $3,369.04 to $3,456.41.


Mr. Gross discussed the payroll growth for the KERS non-hazardous system and the payroll values and employee count by types of KERS agencies for fiscal years 2012 to 2016 and noted the drop in numbers from certain types of agencies.


Mr. Gross discussed the negative amortization and stated that it is a potential problem in the level percent of payroll method, and it occurs when the amortization payment (unfunded liability payment) is less than the interest on the unfunded liability. This phenomenon typically occurs earlier in amortization period, but can be prolonged if payroll doesn’t grow as anticipated, and results in growing unfunded liability even if the full ARC is paid and all other assumptions are met. Both KERS and TRS have experienced negative amortization in recent history.


Mr. Gross discussed the 2015 projections for the KERS nonhazardous fund comparing the level percent of pay method (with a 4 percent payroll growth assumption) versus the level dollar method (no payroll growth). Mr. Gross also discussed the payroll growth sensitivity analysis from the 2016 valuation for the KERS non-hazardous, CERS non-hazardous, and TRS funds and provided data on the impact to plan statistics including the employer contribution rate.


In response to questions from Senator McDaniel, Mr. Gross stated that the level dollar method is a more conservative approach since it does not rely upon payroll growth to help finance the liability.


In response to a question from Auditor Harmon, Mr. Gross stated that Kentucky has not looked at going from a 30 year closed to a 15 to 20 year open amortization period and that KRS is bound by statute to adhere to a 30 year amortization period.


Public Comments

Dolly Guenthner, Retiree, testified that she was a whistleblower against retirement fraud, and there is corruption in Elizabethtown in regards to reemployed retirees. Ms. Guenthner wanted answers from the board as to what steps would be taken to resolve the matter. Senator Bowen responded that the PPOB was not a court of law, and the information brought forth would need to be litigated in another capacity.


With no further business, the meeting was adjourned. The next regularly scheduled meeting is Monday, April 24, 2017.