SENATE FINANCE COMMITTEE April 2, 2019 9:00 a.m. 9:00:41 AM CALL TO ORDER Co-Chair Stedman called the Senate Finance Committee meeting to order at 9:00 a.m. MEMBERS PRESENT Senator Natasha von Imhof, Co-Chair Senator Bert Stedman, Co-Chair Senator Lyman Hoffman Senator Peter Micciche Senator Donny Olson Senator Mike Shower Senator Bill Wielechowski Senator David Wilson MEMBERS ABSENT Senator Click Bishop ALSO PRESENT Senator Cathy Giessel; Bruce Tangeman, Commissioner, Department of Revenue; Rob Johnson, Trustee, Alaska Retirement Management Board; Paul Erlendson, Callan Associates; Steve Center, Callan Associates; Ajay Desai, Director, Division of Retirement and Benefits, Department of Administration; Kevin Worley, Chief Financial Officer, Division of Retirement and Benefits. PRESENT VIA TELECONFERENCE David Kershner, Buck Firm SUMMARY PRESENTATION: ALASKA RETIREMENT MANAGEMENT BOARD UPDATE Co-Chair Stedman discussed the order of business for the meeting. He clarified that the state had been struggling with unfunded pension liability since the early 2000s. He thought it would take another decade to address the issue. He stressed that there was no imminent danger of health and retirement benefits not being paid. Co-Chair Stedman summarized that the state was facing a cashflow issue; and reiterated that there was no need for fear of missed payments. ^PRESENTATION: ALASKA RETIREMENT MANAGEMENT BOARD UPDATE 9:04:46 AM BRUCE TANGEMAN, COMMISSIONER, DEPARTMENT OF REVENUE, informed that the presentation was broken into three parts with testifiers from the Alaska Retirement Board, Callan LLC, the Department of Retirement and Benefits, and Buck Consulting. 9:05:48 AM ROB JOHNSON, TRUSTEE, ALASKA RETIREMENT MANAGEMENT BOARD, discussed the presentation "Informal Presentations from: Alaska Retirement Management Board/Callan LLC/Buck" (copy on file). Mr. Johnson looked at Slide 2, "ARMB Overview." Mr. Johnson turned to Slide 3, " Alaska Retirement Management Board": Primary Duties (AS 37.10.210(a)) ? Serve as trustee for pension and retiree health trusts, the State of Alaska Supplemental Annuity Plan, and Deferred Compensation programs ? Manage and invest assets in a manner that is sufficient to meet the liabilities and pension obligations of the systems Mr. Johnson discussed overarching obligations of the Alaska Retirement Management (ARM) Board. 9:07:03 AM Mr. Johnson showed Slide 4, " Alaska Retirement Management Board": Summary of Activities ? Establish Investment Policies ? Review Actuarial Earnings Assumptions ? Establish Asset Allocation ? Set Contribution Rates of Employers ? Provide Investment Options ? Monitor Performance Mr. Johnson asserted that the task of reviewing actuarial earnings assumptions was a critical feature and specified that the review of actuarial assumptions took place every five years. He said that the review of the assumption that had concluded in 2018 had been adopted and would take effect in FY 21. He shared that the most critical change to the assumptions was the earnings assumption, which had been reduced from 8 percent to 7.38 percent. He relayed that the assumption consisted of two things; the real investment earnings of 4.88 percent, which had not changed, and the inflation assumption, which had been dropped. This significant factor change had resulted in the creation of an increasing unfunded liability. Mr. Johnson discussed the importance of asset allocation. He said when setting contribution rates, the board considered actuarial assumptions, actuarial returns, and setting contribution rates. 9:09:19 AM Mr. Johnson continued discussing Slide 4. He said that the board provided investment options for defined contribution plans and set the investment options for SBS and deferred compensation programs. 9:10:27 AM Mr. Johnson displayed Slide 5, "Alaska Retirement Management Board": Composition of the Board Nine members o Commissioners of administration and revenue o Seven members appointed by the governor ? Qualify for permanent fund dividend ? Recognized competence in investment management, finance, banking, economics, accounting, pension administration or actuarial analysis o Two members of the general public o One member employed as finance officer for a political subdivision o Two PERS and two TRS members, each selected from a list of four nominees submitted from PERS and TRS bargaining units o Other than commissioners, members serve staggered, four-year terms Mr. Johnson shared that the composition of the board was defined in AS 37.10.210 and 220. Mr. Johnson discussed the PERS and TRS members. The seven governor appointees served on staggering terms and were subject to reappointment. 9:12:58 AM Mr. Johnson showed Slide 6, "Alaska Retirement Management Board": Composition of the Board (cont.) ? Rob Johnson, Chair, represents PERS bargaining units ? Gayle Harbo, Secretary, represents TRS bargaining units ? Lorne Bretz, finance officer ? Tom Brice, represents PERS bargaining units ? Allen Hippler, member of general public ? Bruce Tangeman, Commissioner of Revenue Designee ? Kelly Tshibaka, Commissioner of Administration Designee ? Norman West, member of general public ? Bob Williams, represents TRS bargaining units Mr. Johnson thought the board was a talented group that worked well together. He thought the statutory construct in the state prevented schisms and cliques and allowed the board to function well. He reiterated that the statutory constructs provided workable solutions and were a model for problem solving. He encouraged people to attend board meetings. 9:16:52 AM Co-Chair Stedman commented that on the defined benefit side, the responsibility for paying the benefit lie with the state, rather than the employee. He cautioned that if a mistake was made that resulted in unfunded liability, the onus fell on the legislature for funding. He noted that the previous board had not included a state representative, which had led to a restructuring of the board. 9:18:21 AM Senator Hoffman asked Mr. Johnson to review the renumerations received by the board. Mr. Johnson stated that law entitled the board to per diem when traveling and an honorarium of $400 per day for days of meetings, and one travel day associated with meetings. All travel was reimbursed. 9:19:04 AM PAUL ERLENDSON, CALLAN ASSOCIATES, relayed that Callan and Associates had been in business since 1973, and had worked with several state pension funds and sovereign wealth funds. He relayed that the group helped in four main areas: asset allocation, implementation of policy, ascertain objectives, and ongoing education. STEVE CENTER, CALLAN ASSOCIATES, introduced himself. 9:21:35 AM Co-Chair von Imhof looked ahead to Slide 10, "25 Years of Capital Market Return History," which constrained a table entitled 'Returns for Periods ended December 31, 2018." Co-Chair Stedman interjected that the testifiers would be speaking to the current performance of the board after discussion the board's history. 9:22:31 AM Mr. Erlendson looked at Slide 10, which was divided by asset classes. He continued later slides would address policy setting and asset allocation. He thought there might be questions about the Permanent Fund, which would be addressed later in the presentation. He said that any information that was not presented today would be given in a follow up presentation. 9:23:37 AM Co-Chair Stedman thought there may be requests for additional information for the committee. 9:23:49 AM Mr. Erlandson noted that the figures on the table on slide 10 represented cumulative rates of return as of December 31, 2019. He pointed to the far-left column, which listed groups of asset classes. He noted the "risky assets": U.S Equity, Non-U.S. Equity, Fixed Income, and Real Estate. He said that "Alternatives" was a category and not an asset class, as they could not be indexed or bought passively. He said that the returns in the last three-month of last year reflected that equity markets had been down significantly. He stated that the highest returning strategies were equity-oriented strategies; fixed income returns were in the mixed single digits, real estate was the most stable, and the goal was to select the right combination in order to achieve long-term objectives. 9:25:31 AM Mr. Erlandson showed reviewed Slide 11, "Stock Market Returns by Calendar Year," which showed a graph entitled '2018 Performance in Perspective: History of the U.S. Stock Market (230 Years of Returns).' He directed attention to the tallest column and noted that the most frequently occurring annual return was from 0 to 10 percent. He pointed out that the graph depicted a classic bell curve distribution. The extreme outcomes were rare, but he pointed to 2008 as an example. He noted that the market had been relatively benign over the last few years. 9:27:05 AM Co-Chair Stedman thought the previous two years had robust returns. He asked whether there was an expectation that the market would stay robust. Mr. Erlandson stated that historically, lower returns would be expected in the future. He relayed that interest rates had dropped year after year, and the challenge to investors was to find the safest return on investment, which was cash; with equities on could suffer a 30 percent loss. He said that the hope for an investor was that longer the capital could work, the more risk that was taken, the higher the return would be achieved over a longer time period. He said that if the time period was short then risks could not be taken, which resulted in a portfolio with less risky assets and a lower return assumption. He stated that as interest rates went up there was no need to take as much risk with higher returning assets and the challenge became finding the right investment mix. He asserted that, based on history, returns would be lower than historical averages but would rise to equilibrium rates. He referred to Slide 10. He said that as interest rates increased, stable income could be achieved without taking riskier assets. 9:30:50 AM Co-Chair Stedman looked at Slide 11. He was under the impression that active markets started in the mid 1920's. He wondered how the presenters had found data to include from a time prior to the 1920's, when there were no active markets. Mr. Erlandson replied that the slide measured the riskiness of opportunity sets that were available at that point in time. He said that the information on the chart that related to the last decade, and not the last 100 years, was the relevant information. 9:32:41 AM Co-Chair Stedman thought information from the 1800s did not have relevancy and the ability to measure that information was questionable. 9:32:58 AM Mr. Erlandson spoke to Slide 12, "Historical Public Fund Asset Allocation and Returns," which showed a line graph showing rolling 10-year returns, as well as a grouping of pie charts and a data table. He drew attention to the first pie chart, which showed that in 1985 the average pension fund was almost equally invested in domestic equity and domestic fixed income. He said that the 10-year return for the average pension fund in 1985 was over 10 percent. From left to right the pie charts indicated that public pension funds have become progressively more complex investment structures. He said that from the late 80s until the market crash in the early 2000s, the average public pension had returns in the low double digits. He said that once inflation was brought under control in the late 80s, money was plentiful for institutional investors, increasing appetites for risk. He noted that every additional asset class that was brought in was typically done at a higher management fee; the costs of the programs went up and the returns were robust. Over the 43-year history charted on the slide - the average 10-year return was over 9 percent per year. He noted the vertical line on the graph that charted the market and pointed to 2009, when the global financial crisis ebbed, and the market began to correct itself. He said that before that time the average pension fund had a return of approximately 10 percent, since that time the average return had been less than 6 percent. He reiterated that the asset allocation of the average pension fund had grown progressively more complex in order to find assets that will deliver higher expected returns. 9:36:34 AM Co-Chair Stedman looked at 2010 and 2015 and asked what the 10-year trail was for the board's administrable assets. Mr. Erlandson offered to provide the information later. Co-Chair Stedman recalled at that from 2007 to 2010, Callan had given several presentations that had indicated that the target rate should be 8.5 percent. He said that the committee had repeatedly expressed concerns about the rate of expected return being too high and was unachievable. He wanted to see the target rates for the time periods relative to the returns on the slide. He was concerned that year after year the aggregate dollar value of the portfolio was missing the target. He worried that the state could never catch up and wanted to know if the state was meeting the projections. He thought it was nice to see the historical public trust asset allocations but stressed that it was more important to see the returns on the state's current portfolio. 9:39:54 AM Co-Chair von Imhof added to Co-Chair Stedman's comments. She thought the basic material of the presentation was fine but suggested that the committee had expected a more sophisticated presentation that revealed specific information, rather than general trends. She said that when the state dropped form an 8 percent return estimate to 7.3 percent, payments increased. She thought it could be helpful to look at 10-year, or 5-year, return. She suggested examining the past 25-years in 5-year increments. She thought this would provide different perspectives on volatility. She felt that charting the volatility of the market would be helpful in knowing how the board payments would be affected. She requested a ratio of what the actual returns were and what future proforma payments could be. 9:41:32 AM Mr. Erlandson advanced to Slide 15, "Historical Return Projections: Major Asset Classes," which showed a line graph depicting return projections from 1989 to 2019. He noted that the first observation was in 1989. He drew attention to the 'Private Equity' line and observed that back then it had been assumed that private equity would return 15 percent per year over the decade. He summarized that return assumptions had been coming down and had influenced work with the board when deciding how to achieve long-term goals. 9:43:16 AM Mr. Erlandson turned to Slide 16, " Historical Risk Projections: Major Asset Classes," which showed a line graph depicting risk projections from 1989 to 2019. He said that the slide showed that risk levels declined a bit but not as much as the returns. He stated that this meant the if the targeted rate of return was at 1989 levels more money would need to be deployed into risker assets. He stressed that the challenge of asset allocation was how to balance the amount of variability on returns on the short- term basis, with the goal of achieving the long-term assumption. 9:44:50 AM Senator Shower was concerned with inflation. He asked where the presenter expected inflation to be currently, in relation to past levels. Mr. Erlandson stated that the long-term average going back to the 1950s was about 4 percent per year, but much of the inflation occurred during the 70s and 80s. He thought the inflation average was more misleading than informative. He said that over the last 20 years inflation had been at less than 3 percent, closer to 2 over the past decade. He related that his assumption going forward was 2.25 percent. Co-Chair Stedman asked about what the rate of inflation was over the previous century. Mr. Erlandson said he did not know. 9:46:16 AM Co-Chair von Imhof considered an overlay of actual returns from 1989 to now. Mr. Center returned to Slide 15 and noted that the chart represented Callan's capital market projections for each calendar year. He relayed that the only thing that could be promised about the projections was that they would be wrong. He thought guessing what the next 10-year return was an art rather than a science. He stated that the only scientific fact that could be considered would be the current interest rates and their impact on historical returns. He offered to send a chart that showed the information. Co-Chair von Imhof thought it would be interesting to see some volatility overlaid on the chart. She spoke of the permanent fund and mentioned the percent of market value (POMV) draw, which was a five-year lookback. She noted that volatility materially affected the draw. She thought it would be helpful to see the variations in the potential market scenarios. 9:48:59 AM Mr. Center stated that one reason a five-year lookback was used was to minimize the impact of volatility. He stated that volatility in the market could result in substantial variations in the draw amount. He stated that in the most previous study he had done, there was less than a 10 percent chance that the ERA would be depleted due to a multiple year period of negative returns. He said that further finding would be presented by June 30, 2019. 9:50:13 AM Co-Chair von Imhof thought while the ERA balance was important, it was less so than the what the potential budget number could be with the POMV draw. In a true pure POMV, there was no distinction between the corpus of the fund and the ERA. She considered that the amount of the POMV was more important that what was in the ERA. 9:51:09 AM Co-Chair Stedman asked about correlations as shown on Slide 16. He asked for further explanation of the correlations. Mr. Erlandson reminded that Slide 15 showed projected returns and Slide 16 showed projected risk. Because of the way models were built, a single number was plugged into the equation, which was rarely accurate. He relayed that correlation calculations used numbers rarely used in the modeling. The model was meant to demonstrate trends and were only tools used to examine knowable facts and ranges in outcomes. He thought that the narrowing of the probables was the best that could be done. 9:53:46 AM Mr. Erlandson displayed Slide 17, "Projected Allocations required to Achieve 7.5% Expected Return: Increasing Levels of Risk Required to Obtain the Same Expected Rate of Return," which showed a flow chart including three pie charts. The charts illustrated asset mixes that would have been suggested to investors in 1989, 2004, and 2019, to achieve a 7.5 percent return. He said that using the assumptions in 1989, risky assets would have been unnecessary to reach the 7.5 percent return. He noted that in 2004, achieving 7.5 percent would have required a movement to riskier assets, tripling the level of risk to achieve the wanted return. The slide detailed the progression: • In 1989, our expectations for cash and broad U.S. fixed income were 6.80 percent and 9.35 percent, respectively. • 15 years later, and investor would have needed half of the portfolio in public equities to achieve 7.5 percent, nearly tripling the portfolio volatility of 1989. • Today an investor is required to include 96 percent in return-seeking assets to earn 7.5 percent at almost 6 times the volatility compared to 1989. Mr. Erlandson questioned whether it was more important to achieve the 7.5 percent return or, was it more important to avoid and 18 percent risk level. 9:55:57 AM Co-Chair Stedman suggested that if Mr. Erlandson provided a bugle chart it could help the committee to visualize future performance expectations. Mr. Center stated that he was currently working through an asset allocation study and an asset liability analysis which would inform future overall asset allocation for PERS and TRS. He agreed to provide the requested information in a bugle chart format. Co-Chair Stedman recalled that the committee had great difficulty in the past getting the similar charts produced. He believed that the cart would be useful in examining the magnitude of potential outcomes. 9:57:43 AM Co-Chair von Imhof thought Slide 17 was very telling. She noted that in 1989, the portfolio assumed less risk, but inflation was more significant. She thought it would be helpful for a national and global analysis of what affected different asset choices, with recommendations. Mr. Center stated that the process was ongoing. 9:58:57 AM Senator Micciche asked whether 1989 represented an extremely unusual case. He asked whether the state had ever been in the same situation at another point in time. He wondered how the 1989 numbers for risk could be compared with those of 2004 and forward. Mr. Erlandson stated that there had been shorter times when the cash return had been relatively high; he noted that in 2007, the return on cash investment was nearly 5 percent. He thought it was unusual and inexplicable that interest rates had been down while there was so much capital in the market. He said that inflation was not viewed as a potential issue and that interest rates were predicted to rise. He discussed the investment climate in 1989 and how it informed decisions at that point in time. 10:01:08 AM Mr. Erlandson referenced Slide 19, "2019 Callan Capital Market Projections," which showed a data table entitled ' Expected risk and return (20192028).' He pointed out the Asset Class column on the left. He noted the Index column and explained the column reflected if one were to buy a passive exposure, which index would be purchased. He said that the capital market projections were net of fee, index returns; without active management, except for "alternatives" that could not be passively bought. He continued to discuss the thought process for an investor as they sought return on their investment. He stated that the policy assumed passive implementation and if a return was hoped for that exceeded the target, active manager opportunities needed to be explored. 10:02:53 AM Co-Chair Stedman thought it was possible to measure statistically the stock selection and timing ability within the portfolio. Mr. Erlandson confirmed that on the Callan website there was a quarterly exhibit that measured the net of fee effectiveness of managers. Co-Chair Stedman asked for an explanation of the difference between the 1-year arithmetic and 10-year annualized projected return. 10:03:33 AM Mr. Erlandson stated that the arithmetic return was from the beginning, day zero, to the end of the 10-year period, without variability. The geometric return assumed variability of return. Co-Chair Stedman thought the easy way to consider the difference was that the arithmetic was a 1-year return, the geometric was a 10-year compounding return that took risk level into account. Co-Chair Stedman asked for an explanation of 'duration'. Mr. Center stated that for fixed income investments "duration" had to do with the length of time that the bond had to maturity. Shorter duration maturity could be 3 years or less, longer duration dated securities, or market duration, could take 5 or 6 years to maturity. Co-Chair Stedman thought the duration time would indicate the timing of the cash flows. Mr. Center added that shorter duration securities tended to have lower risk because the likelihood of the default was lower than that of a longer duration bond. 10:06:16 AM Mr. Erlandson discussed Slide 20, " Expanding the Length of the Forecast Horizon": 10-Year vs. Equilibrium Capital Market Expectations ?As the time horizon grows beyond 10 years, our capital market expectations increasingly incorporate "equilibrium returns". Equilibrium returns reference long-term historical mean results, with an overlay of informed judgment. Key elements to consider: Nominal returns Inflation Real returns Risk premium bonds over cash, stocks over bonds, long duration over short Long-term underlying economic growth (real GDP) ? 10-Year expectations: Large Cap Stocks: 7.0% nominal, 4.75% real, 3.25% premium over bonds Bonds: 3.75% nominal, 1.50% real, 1.25% premium over cash Cash: 2.50% nominal, 0.25% real Inflation: 2.25% Underlying economic growth (real GDP) 2 to 2.5% per year ?Equilibrium expectations: Large Cap Stocks: 8.25% nominal, 6.0% real, 3.25% premium over bonds Bonds: 5% nominal, 2.75% real, 1.75% premium over cash Cash: 3.25% nominal, 1.0% real Inflation: 2.25% Underlying economic growth (real GDP) 3% per year Mr. Erlandson said that the slide put into narrative format the assumption that lower returns were being assumed for the next decade but, 15, 25, 30 years out, returns would gradually increase. He said that everything was priced off the risk-free rate and anything that was purchased would be scrutinized for cash return and risk. He argued that longer time horizons should assume higher rates of return but, if the return horizon was shorter then lower return expectations should be made for the same asset allocation. 10:09:12 AM Mr. Erlandson turned to Slide 21, "Comparison of 10-year Returns with Equilibrium Returns," which showed a data table that compared the returns. 10:09:31 AM Mr. Erlandson referenced Slide 22, "2019 Capital Market Projections versus Other Firms," which showed a data table listing the 11 organizations that had made projections like Callan. He noted that the return projects did not all use the same asset classes and the time horizons differed. He gleaned that the longer the time horizon, generally the higher return the organization listed were predicting. He pointed out the four columns on the far right that showed the high and the low returns, with the Callan assumptions highlighted in yellow. 10:10:49 AM Mr. Center displayed Slide 23, "PERS and TRS Asset Allocation Over Time": ? Table shows asset allocations for PERS (solid line) and TRS (dashed line) over time. Dashes are only visible when differences arise. ? Through much of the1990's, PERS and TRS had slightly different asset allocations ? PERS had a moderately higher Fixed Income allocation ? TRS had a slightly higher allocation to Domestic and Non-US Equities ? Asset Allocations have been effectively similar since 2000 Mr. Center stated that the slide was in response to a question from the committee after a presentation that had showed historical returns for PERS and TRS; a return difference had been observed for the two plans over the long term. He noted that there were two sets of lines for each asset class. The slide showed asset allocations for PERS (solid line) and TRS (dashed line) over time; dashed were only visible when differenced arose. Through much of the 1990s, PERS and TRS had slightly different asset allocations, PERS had a moderately higher fixed income allocation, while TRS had a slightly higher allocation to domestic and non-US equities. The slide concluded that asset allocations had been effectively similar since 2000. 10:13:04 AM Co-Chair Stedman thank the presenters and invited the next set of speakers to the table. 10:13:50 AM AJAY DESAI, DIRECTOR, DIVISION OF RETIREMENT AND BENEFITS, DEPARTMENT OF ADMINISTRATION, introduced himself. KEVIN WORLEY, CHIEF FINANCIAL OFFICER, DIVISION OF RETIREMENT AND BENEFITS, introduced Mr. Kershner and showed slide 24, "Buck Analysis." DAVID KERSHNER, BUCK FIRM (via teleconference), relayed that the primary function of Buck had been to help the board assess the funded status of the plans each year based on the annual evaluations. The evaluations were used to set the contribution rates for the various employers and the state. 10:15:24 AM Mr. Kershner referenced Slide 25, "Impact of Actuarial Assumption Changes," which showed a data table the illustrated actuarial accrued liability for PERS and TRS. He explained that contributions for the plan came from three sources: active participants, participating employers, and the additional state contribution. He discussed the statutorily set contribution rates for each source. Mr. Kershner continued to address Slide 25. He informed that Buck was in the process of the 2018 evaluation, which would be presented to the board in May 2019. Over the previous year, the firm had performed an experience study to assess the reasonableness of the actuarial assumptions and whether changes were warranted. He continued that the firm made a series of assumptions in order to project benefits for participants decades into the future. Demographic assumptions were used to project the plan population each year into the future. Mr. Kershner discussed economic assumptions used to devise an investment return rate, which was composed of an inflation rate assumption and a real rate of return. He continued that the inflation assumption was common to other economic assumptions, i.e.; the assumption regarding future pay increases for active members because pension benefits are a function of average pay of retirement. Merit and productivity were also considered. He spoke to healthcare and said that the liabilities were a function of the medical and prescription drug claims generated by retirees and their covered dependents. He stated that assumptions were made as to how those expenses were expected to increase in the future due to inflation, changes in utilization, changes in technology, and new programs that were implemented to control healthcare costs. 10:19:40 AM Mr. Kershner shared that the board had adopted a study in January, and 2018 valuations would be used to set contribution rates for FY 21. Mr. Kershner noted that Slide 25, as well as the following slide would address the overall impact on PERS and TRS of the newly adopted assumption changes. He relayed that the top half of slide 25, showed the actuarial accrued liability from the most recent valuation prior to the new assumption being reflected in the 2018 evaluations. Mr. Kershner pointed to line 1, which showed PERS liability for pension and healthcare at approximately $22 billion and $10.1 billion for TRS. The actuarial accrued liability was the present value of future benefits, based on assumptions, attributable to service earned as of the evaluation date of June 30, 2017. He said that this was the denominator of the funded status of the plan and was used to set the contribution rate. 10:22:47 AM Mr. Kershner related that lowering the interest rate assumption from 8 percent to 7.38 percent had been the primary driver of the increase in the actuarial accrued liability. He said that the reason that the accrued liability increased, while the interest rate assumption decreased rate, was because more needed to be set aside in the present in order to provide the same benefits into the future. Mr. Kershner emphasized that the overall cost of the plans over time would be whatever the benefits were, administrative expenses, and whatever invested assets earned over time. All the actuarial assumptions did was help allocate the assets over time and determine a pattern of contributions. He stressed that the goal was to provide as much stability over time as possible, while recognizing that there was no control over what happened to invested assets, and little control over liability due to changes in assumptions. 10:24:59 AM Mr. Kershner pointed out that the second line on Slide 25 reflected the increase in the actuarial accrued liability due to the new assumptions in dollar amounts, the next line reflected the percentage basis. For PERS there was approximately a 6 percent increase in the actual accrued liability, under 1 percent for TRS. 10:25:28 AM Mr. Kershner noted that line 3 introduced another change that would be reflected in the 2018 evaluations, the adoption of the Employer Group Waiver Plus program (EGWP). This was a program offered by the federal government that provided incentives to employers to cover prescription drugs for their participants that were covered under Medicare. Until EGWP was implemented, there was a Retiree Drug Subsidy (RDS) program that the state had decided to stop because EGWP provided a greater subsidy. The EGWP program reduced the liability of the plan. He pointed out that the third line on slide showed that showed a decrease in healthcare liabilities for PERS and TRS. He reiterated that the slide reflected estimations as of July 30, 2017 and were currently being refined to include the new program. 10:27:56 AM Mr. Kershner continued to discuss Slide 25. He discussed additional state contributions as shown on the slide. He noted that there were two components to the contribution rate: the cost of benefits occurring in the current year and the funding of the unfunded liability. Beginning in 2014, stature required that the unfunded liability for PERS and TRS be funded over a closed, 25-year period, starting in 2014, through 2039. He said that the state had 21 years left to amortize the unfunded liability. He said that work was being done with the board to modify the way the unfunded liability would be amortized, with an emphasis on minimizing the potential volatility in those additional state contributions. He offered a hypothetical using the current method of funding the unfunded liability that would result in an increase in state contribution. He said that the amortization modification method of 25-year layering would help to determine projected state contributions into the future. 10:32:47 AM Mr. Kershner continued to address slide 25 and noted that the projections assumed that all the assumptions over time would be realized. He considered current assumptions projected for FY 21 - FY 39; $4.1 billion for PERS and $3.7 billion for TRS. He discussed the changes under the new assumptions before recognizing EGWP and the 25-year layering. 10:36:33 AM Co-Chair Stedman asked Mr. Kershner to clarify why the burden fell to the employer and not the state. Mr. Kershner clarified that the first use of 25-year layering would start in 2018, and projected contribution rates would be less than the 22 percent employer contribution for PERS, and 12.56 for TRS. The additional state contributions only kicked in when the contribution rates exceeded the employer limits. 10:38:08 AM Co-Chair von Imhof thought EGWP had a tremendous positive effect on the unfunded liability balance. She wondered whether Mr. Kershner knew of any other factors such as pooling mechanisms that could aid in reducing liability. Mr. Kershner confirmed that he was not a healthcare actuary and offered to follow up later. 10:39:16 AM Senator Hoffman also saw the benefits of the EGWP program. He asked what authority the board had to implement EGWP. He asked what other states had implemented the program. Mr. Desai stated that the process of EGWP began in 2017. Initially the study with a health consulted had shown potential savings. When it was realized that the savings would be achieved through the program - it was implemented. Mr. Desai offered to provide information later regarding the question of authority to implement the program. 10:41:24 AM Senator Hoffman wanted a list of other states participating in the program. He asked what impact the program had on prescription drug prices for health trusts and end users. Co-Chair Stedman asked the testifiers to provide the information later. 10:41:56 AM Senator Micciche asked why excluding and including EGWP were stacked in the assumptions on the second half of the slide. Mr. Kershner stated that the changes were listed separately to provide more information on the impact of the three significant changes being implemented for the current valuations. 10:42:53 AM Senator Micciche looked at the net increase and decrease on unfunded liability. He thought that both influenced the actuarial calculation. Mr. Kershner stated that the new assumptions and the 25- year layering of the unfunded liability were separate decisions. The board could have approved the new assumptions, and kept the current, closed 25-year amortization, without introducing layering. The implementation of EGWP was a decision made by DOA and was independent of the assumptions or the method used to amortize the unfunded liability. 10:44:28 AM Senator Micciche considered the decisions that were made, and understood that the actuarial accrued liability, with new assumptions, was $22.585 billion. Mr. Kershner answered in the affirmative. Senator Micciche asked whether the unfunded liability was $8.83 billion. Mr. Kershner stated that that unfunded liability would be the difference between the actuarial accrued liability and the asset values, which were not on the slide. Co-Chair Stedman remarked that the presentation had been nearly two hours long, and he could not say that he felt better informed on the matter. He expressed the desire for a frank conversation on the money available and how the problem was going to be resolved. He was concerned about the increase in unfunded liability under the new calculation. Mr. Kershner stated that the 2018 valuation would reflect the 6/30/18 liability, assets, and unfunded amounts. He admitted that the slide did not reflect the unfunded liability. 10:46:51 AM Co-Chair Stedman queried the assets and liability and what was expected in the report to the board. Mr. Worley stated that actuarial valuation process was still underway, the results of which still needed review by actuaries. Co-Chair Stedman asked for market value of the portfolio on June 30, 2018. Mr. Worley agreed to provide to provide the information. Co-Chair Stedman expected to see a $1 billion increase. He thought it was clear that there was a potential increase of $1 billion in liability to the state. He stressed that the committee was very concerned about the impact on the next budget cycle. He lamented that the committee struggled with the delay of information. 10:49:49 AM Commissioner Tangeman recalled a presentation at the January ARM Board meeting, at which time the liability prior to recent action was $6.9 billion, or 78 percent funded. The estimate after the experience study (without EGWP) was about $2 billion. With the EGWP the total was $7.2 billion. The numbers would be trued up in the coming month. Co-Chair Stedman understood that without EGWP and other modifications the numbers would be higher. Commissioner Tangeman answered in the affirmative. Co-Chair Stedman summarized that the state had a $7.2 billion liability. He reiterated his request for a bugle chart of the analysis over the last decade. He noted that a similar chart had been done in the past and believed that it would give the committee a better idea of the state's exposure to risk. He felt that the committee needed more information to understand what needed to be done going forward. 10:53:50 AM Senator Micciche questioned the sources for the numbers being used by Commissioner Tangeman. Co-Chair Stedman said that the information would be made available to the committee. 10:54:40 AM Mr. Kershner discussed Slide 26, "Percentage Impact of New Assumptions/Methods on Actuarial Accrued Liability as of June 30, 2017," which showed a data table showing the impact of new assumptions on the 6/30/17 liabilities. He noted there was four categories of changes: demographic assumptions, salary increase rates, inflation rate (impact on COLA-related benefits only), and investment return. He discussed the two types of COLAs - the residency adjustment and benefits tied to inflation. He highlighted the dropping of the investment return from 8 percent to 7.38 percent and the impact of PERS and TRS pension and healthcare. 10:57:01 AM Mr. Kershner turned to Slide 27, which showed a data table. He emphasized that the numbers were hypothetical and illustrated how the 25-year layering would affect numbers over time. Layer one was the unfunded liability that would continue to be amortized over what was left of the 25-year period, or through FY 39. The columns represented subsequent layering hypotheticals through 2046. Co-Chair Stedman thanked the presenters. He noted that the subject matter could be frustrating. He believed that the working with presenters to gather information for future discussions could be fruitful. He highlighted the magnitude of the issue and the importance that the state meets its obligations. 11:01:48 AM Co-Chair Stedman stated the co-chairs would work with the presenters and the Commissioner of Revenue on the issue. ADJOURNMENT 11:03:12 AM The meeting was adjourned at 11:03 a.m.