Legislature(2005 - 2006)SENATE FINANCE 532
01/26/2005 09:30 AM Senate FINANCE
| Audio | Topic |
|---|---|
| Start | |
| Adjourn |
* first hearing in first committee of referral
+ teleconferenced
= bill was previously heard/scheduled
+ teleconferenced
= bill was previously heard/scheduled
| + | TELECONFERENCED | ||
MINUTES
SENATE FINANCE COMMITTEE
January 26, 2005
9:32 a.m.
CALL TO ORDER
Co-Chair Green convened the meeting at approximately 9:32:58 AM.
PRESENT
Senator Lyda Green, Co-Chair
Senator Gary Wilken, Co-Chair
Senator Con Bunde, Vice-Chair
Senator Fred Dyson
Senator Bert Stedman
Senator Donny Olson
Also Attending: SENATOR GARY STEVENS; REPRESENTATIVE MIKE KELLY;
MELANIE MILLHORN, Director, Division of Retirement and Benefits,
Department of Administration;
Attending via Teleconference: From an offnet location: BOB
REYNOLDS, Senior Actuarial Consultant, Mercer Human Resource
Consulting
SUMMARY INFORMATION
Overview By Department of Administration:
Public Employee's and Teachers Retirement Systems
AT EASE 9:35:42 AM
MELANIE MILLHORN, Director, Division of Retirement and Benefits,
Department of Administration, directed her testimony to a handout
titled, "State of Alaska, PERS & TRS, Presentation to the Alaska
Legislature" [copy on file.]
Ms. Millhorn indicated the three funding policies that govern the
Public Employee's Retirement System (PERS) and the Teacher's
Retirement System (TRS), as demonstrated on page 2 of the overview.
The Alaska State Pension Investment Board governs the investment
policy. The Funding policies are administered by the Public
Employee's Retirement Board, which also has the authority to set
the employer contribution rate. The Teacher's Retirement Board has
the authority to make a recommendation to the commissioner of the
Department of Administration for the employer contribution rate.
Ms. Millhorn noted that over the past several years, the Teacher's
Retirement Board had actually adopted and implemented the employer
contribution rate. This came to the attention of Commissioner Ray
Matiashowski, who then reminded the TRS Board that it did not have
that authority and the practice has ceased.
Ms. Millhorn informed that some funding policies governed by the
PERS Board, the TRS Board and the commissioner include "experience
studies" conducted on the boards every four to five years.
Actuarial audits are done approximately every five years. This
information is available on the Internet.
Ms. Millhorn stressed that the legislature has the "unfettered
right" to determine the benefits policy for the retirement systems.
In 1978 the State attorney general issued an opinion finding that
retirement benefits could not be negotiated through the collective
bargaining process. This decision has not been challenged.
Senator Dyson clarified that the legislature has the authority to
determine benefit policy for the future, but could not make changes
retroactively.
Ms. Millhorn affirmed the legislature could only establish new tier
levels.
Ms. Millhorn characterized page 3, "Alaska Public Employees' and
Teachers' Retirement System, Earnings-Actuarial Rate-Health Cost-
Employer Rates-Funding Ratios" as a summary sheet. The PERS and TRS
systems are defined benefit plans in that the benefits paid are
based on a formula established in statute. Members receive a
lifetime benefit that is based on years of service, salary and age.
Actuarial evaluation reports are prepared annually on the PERS and
TRS programs and measure the assets and the liabilities of the
plan. One primary function of an actuarial evaluation is to
determine the annual contribution amount that is expected to
adequate provide for future benefit payouts. There are over 20
separate assumptions that divided into demographic assumptions and
economic assumptions. Two of the major economic assumptions are the
investment return and the health cost trend. An actuarial
evaluation is essentially a "snapshot in time" to demonstrate what
is necessary in the event all benefits must be paid out at one
point in time. She qualified this would never occur, but the
information is necessary to determine the solvency of the system.
Ms. Millhorn noted the measurement years of FY 01 through FY 03,
and employer rate years of FY 04 through FY 06, listed on page 3.
The employer contribution rate is set two years ahead. The actual
investment return for FY 01 and FY 02 were negative and the return
for FY 03 was a positive rate of 3.69 percent. The system is
projected over a 25-year period at a targeted rate of 8.25 percent.
The shortfall over this three-year period is represented on the
summary as a percentage and as a dollar amount.
Ms. Millhorn reported that historically employer and employee
contributions comprise of 25 percent of the total contributions and
that 75 percent is derived from investment returns on the Fund. The
health care cost is another primary economic assumption and was 7.5
percent in FY 01. This assumption was reset beginning in FY 02 to
twelve percent. The consequence of this was "that it took in
considerable liability to the System by changing that health trend
cost upward." However it more accurately reflects the actual
experience that has occurred for the systems over "some period of
time."
Ms. Millhorn stated that the total accrued liability from the
"health care percent" was 30 percent for FY 01 and 40 percent in FY
03. This is fairly significant to the PERS systems. The same
liability to the TRS system is 28 percent.
Ms. Millhorn explained that the average calculated rate is
determined by the consultant, Mercer Human Resources Consulting,
and presented to the boards with a recommendation of what the
contribution rate must be to fund future benefits for members. The
average calculated rate for the PERS system was 25.63 percent. The
PERS Board is allowed by regulation AAC 35.900 to adopt up to a
five percent increase or decrease in a one-year period. The PERS
Board adopted this regulation in 1991 at the request of the
Municipality of Anchorage for budgetary purposes.
9:45:17 AM
Ms. Millhorn reminded that, unlike other pension systems, the PERS
and TRS plans pre-fund for medical benefits. The funding ratio of
the plans calculated without medical costs is 121.4 percent;
calculated with the medical liability component, the funding ratio
is 72.8 percent. The states of Ohio, Michigan and Kentucky are the
only other states that pre-fund for medical benefits.
Senator Stedman asked if the calculation of the funding ratio
excluding the medical liability included all the assets in the
plan.
Ms. Millhorn answered yes.
Senator Stedman noted that a portion of the contributions are
"embedded" for health care costs.
Ms. Millhorn deferred to Mercer Human Resource Consulting.
BOB REYNOLDS, Senior Actuarial Consultant, Mercer Human Resource
Consulting, testified via teleconference from an offnet location to
reaffirm that the 121 percent funding ratio for non-medical
benefits includes all the assets of the PERS system. He also
affirmed Senator Stedman's observation that the ratio is as high as
it is because the systems have been funding to target both medical
and non-medical benefits over time. Mr. Reynolds qualified that if
the contributions had only been targeted to fund non-medical
benefits, the funding ratio would be considerably lower.
Senator Stedman clarified that if the systems had never been funded
to pre-fund medical costs, contribution rates would have been much
lower and the 121 percent funding ratio would also be lower.
Mr. Reynolds affirmed.
Senator Stedman surmised that Alaska's systems are therefore not
readily comparable to the systems of other states that do not pre-
fund for medical benefits.
Mr. Reynolds agreed; although in terms of comparing the liabilities
and the assets of the system this is a good comparison. The funding
ratio is determined by dividing the assets by the non-medical
liabilities.
Senator Stedman clarified that the Board-adopted contribution rate
for FY01 was 6.77 percent and was 16.77 percent for FY 03. He asked
if this increase was "substantially broader" compared to other
states. He asked if the contribution rates for other states had
been "extremely low".
Mr. Reynolds did not have specific survey data on the matter, but
estimated Senator Stedman's assessment was correct. The standard
actuarial practice is to calculate a contribution rate that takes
into account the funded status of the system. At a time when the
funded status of a system is in excess of 100 percent, the surplus
could be drawn upon to lower contribution rates. Many, if not most,
systems had a perceived surplus during the later 1990s prior to
market losses and changes in assumptions to reflect health care
experience.
Senator Stedman characterized the over-funding of the Alaska
pension systems that occurred in 1999 and 2000 as "slight" at
approximately one to five-percent, rather than a more significant
20 percent.
Mr. Reynolds cited the consultant's evaluation reports finding that
the funding statuses not significantly in excess of 100 percent.
One reason is the use of "smoothing practices" in valuing the
assets of the system to mitigate some of the volatility inherent in
the investment market. Actual funding status relative to the
market value of the assets would have demonstrated a higher over-
funding of between 115 and 120 percent. The adjusted funding status
factoring the smoothing practices calculates the funding status at
between 101 and 106 percent.
Senator Stedman asked for explanation of how the smoothing practice
operates: its benefits as well as the possibility of "skewing"
results.
Mr. Reynolds remarked that smoothing is a standard actuarial tool.
It is done in an attempt to help the State and the boards to budget
for the contributions to the plans. The contribution requirements
are a function of the funded status of the systems, which is the
difference between the liabilities and the assets of the systems.
The assets are diversely invested in the marketplace by the Alaska
State Pension Investment Board, according to investment policy, and
include stocks, bonds and real estate. Those investments measured
over short time periods of one year can be volatile. The theory of
smoothing methods is to help mitigate that volatility to allow for
"a more orderly" funding process. Over longer time periods, this
theory is successful. However, the situation was different in
recent years because of several consecutive years of significantly
higher than expected rates of returns followed by three successive
years with significantly lower rates of returns than expectations.
Senator Bunde noted that the average calculated rate for the PERS
and TRS plans were significantly higher than the contribution rates
adopted by the boards.
Ms. Millhorn defined the average calculated rate as the
contribution rate recommended by the consultant to the boards as
necessary to fully fund the plans. She explained that the adopted
rates are lower because the PERS board is limited by regulation to
a five percent annual increase or decrease. The TRS Board is not
limited by such a cap, and could have adopted a higher percentage
rate increase.
Senator Bunde understood that the PERS Board adopted the maximum
increase allowed, and the TRS Board could have adopted a higher
rate increase, but chose not to.
Ms. Millhorn affirmed.
Senator Bunde commented to this action of "not facing reality".
Ms. Millhorn indicated this would be discussed.
Ms. Millhorn spoke of the consequences of the boards' adoption of
rates lower than the recommended average calculated rates. The
systems incur additional liability for the next 25-year period.
This represents an additional liability of 1.4 percent for PERS and
1.1 percent to the employer contribution rate.
Senator Bunde was interested in why the boards failed to adopt the
recommended rate. He surmised this "makes the problem worse not
better."
Ms. Millhorn agreed. She acknowledged the "flip side" from the
employers' perspective of the amount of increases an employer could
budget for in any one year.
Ms. Millhorn continued with her presentation noting the information
listed for the PERS plan is also listed for the TRS plan on page 3.
The situation of TRS is similar to that for PERS. The projected
investment return is 8.25 percent, although the actual returns were
3.68 percent for FY 03, compared to 3.67 percent for the PERS plan.
The cumulative shortfall for TRS during the three-year period of FY
01 through FY 03 is 34 percent and represents a shortfall to the
system of $2.1 billion.
Ms. Millhorn noted the cumulative health care trend was changed
during this time period and represents 28 percent in accrued
liabilities to the TRS plan and 40 percent to the PERS plan. The
benefit is higher for TRS members at approximately $2,500 per month
per member compared to the average benefit of approximately $1,500
for PERS members.
Ms. Millhorn reported that the TRS system is funded at 89.5 percent
excluding the medical liability. Including the medical component,
the system is funded at 64.3 percent.
Senator Stedman requested further information be provided to him
regarding the funding ratio, specifically the impact of the
smoothing practices.
Ms. Millhorn would provide this information.
Co-Chair Wilken noted the funding ratio is calculated as a total
and with the medical benefits excluded. He assumed that the medical
benefits and non-medical benefits funding ratios would calculate to
the total. He asked the funding ratio of the medical benefits
excluding the non-medical benefits.
Ms. Millhorn deferred to Mr. Reynolds.
Mr. Reynolds replied that because all assets have been applied to
the calculation of non-medical benefits, the medical benefits would
have zero assets.
Senator Stedman commented that the data could be calculated to show
all the assets for either medical or non-medical benefits, and
therefore the analysis is meaningless because the assets are
"double-counted".
Ms. Millhorn next directed attention to page 4, a press release
from Gallagher Benefit Services, Inc. titled, "New GASB Rules for
Public Sector Plans Finalized". This announces that in November
2004, the Government Accounting Standards Board (GASB) released a
Technical Bulletin mandating that by the year 2007, employers
subject to GASB rules account for retiree health benefits in
defined benefit pension plans. Most pension systems currently make
annual appropriations for medical benefits expenses. This will have
no affect on the Alaska systems because they already pre-fund for
medical benefits. It will have a significant impact on other
pension systems.
Senator Stedman emphasized that the position of Alaska's pension
plan is different than that of the state of California or other
states that currently do not pre-fund medical benefits. Alaska's
unfunded liability is considerably "smaller and easier to handle"
than other states.
Ms. Millhorn agreed that because the Alaskan plan accounts for
liabilities on pension and medical benefits on an accrual basis, it
is better prepared than other systems. In 2007 those other systems
would be required to account for medical liabilities. This is
significant and could affect the bond ratings for those states.
Officials are attempting to "cut those costs any way that they
can."
Mr. Reynolds had only anecdotal data on the size of the other
states' pension plan medical liabilities. The new GASB rules would
not require those systems to pre-fund for medical benefits, only to
account for the liability. However the discount rate on the assumed
rate of return that governmental entities utilize to calculate
liabilities is supposed to be based on the assets held toward those
liabilities. The PERS and TRS systems utilize a return assumption
of 8.25 percent, based on the investment policy of the ASPIB board
for pre-funding medical benefits. According to the new GASB rule,
states that do not pre-fund for liabilities must utilize the
internal rate of return, essentially a cash rate, to value those
liabilities. This could be approximately three-percent. Not only
would these plans "be booking a liability for the first time" but
would also be required to state a liability that is considerably
higher than if the medical benefits had been pre-funded. This is
because a "very low" discount rate is used to value those
liabilities.
Mr. Reynolds next addressed pages 5 through 7, which indicate
investment returns and funded status for the PERS and TRS plans.
This information "deals with question: 'How did we get here?'" in
listing data for the previous years and analyzing changes in the
funded status of the systems categorized by source. He outlined the
information on the graphs of pages 5 and 6. He noted that the
funding status increased in some years and decreased in other
years.
Mr. Reynolds explained the sources as follows: Non-Health
Assumption Changes, includes mortality improvements or changes in
retirement rates based on the consultant's analysis of the
experience of the system; Health Assumption Changes, the primary
instance being the expectation in which medical costs would
increase in the future; Demographic Experience, determined by
analyzing how the population has changed relative to the
assumptions used; Plan Changes, as improvements in plan benefits;
Medical Experience, how the medical costs changed relative to the
expectations regarding increases; and Investment Experience, how
investments performed relative to the 8.25 percent assumptions. The
charts demonstrate the magnitude of the different sources on
increases and decreases for various years.
Senator Bunde referenced the inset listing "Declines in funding
status were caused by (in descending order of magnitude)" noting
that demographic experience is the sixth-most contributing factor
for PERS and the second most contributing factor for TRS. He asked
the reason for the discrepancy.
Mr. Reynolds explained the demographic experience represents
changes in the population that differ from the assumptions. This
includes the rate and ages in which members retire, termination
rates from active membership prior to retirement age, mortality
experience and salary increases. Ordinarily demographic experience
is a relatively minor factor because the system experience is
specifically analyzed in setting assumptions. The assumptions tend
to be good predictors of actual occurrences. Demographic experience
is a larger contributor for changes to the TRS plan as a result of
a computer system change in the method of reporting teacher
membership. The evaluation date for the systems is June 30 or July
1. The prior systems reported members as terminated at the end of
the school year in the spring then re-hired in the fall for the
following school year. These members were therefore not included in
the data submitted to the consultant. Those members were then
included in the data after the computer system change. This
resulted in a loss to the TRS' funding status.
Co-Chair Green asked the year of this computer system change.
Mr. Reynolds replied the year was 2001.
Senator Stedman also noted that health assumptions was the second-
most contributing factor to the PERS system and the fifth-most
contributing factor to the TRS system and asked if different
assumptions were used to calculate funding status for PERS and TRS.
Mr. Reynolds responded that health experience is not necessarily
different between the two systems. The health assumptions are based
on the rate that health care costs are expected to increase and are
affected by "more national-type concerns". Therefore the same
assumptions are used for both plans. The health assumptions are a
larger factor in the PERS funding status than the TRS because the
non-medical benefits are lower for PERS members than TRS members on
average. As a result the medical liabilities are a larger
contributor to the unfunded liability portion of the PERS plan and
changes to the health assumptions make a greater difference to its
funding status.
Senator Stedman asked for an explanation of the unbalance and
whether it relates to the previous funding ratio and the weight
given to the health assumptions.
Mr. Reynolds replied that the unbalance had nothing to do with
funding, rather the legislated benefit levels of the two systems.
The TRS system has somewhat higher non-medical benefit levels than
the PERS system. As a result, a larger portion of the benefits is
medical for PERS members.
Senator Stedman then pointed out that currently investment
performance is calculated utilizing fair market value. In some
years, many plans held assets at cost or "book value". He asked
when, if ever, the Alaskan plan held assets at cost, or if the
assets have always been held at fair market value.
Mr. Reynolds responded that the assets of the system are always
accounted at fair market value when reported to the consultant on
June 30 of each year. He qualified that assets may have been held
at cost, but if so, it was many years prior. A smoothing method is
used; however, cost is not a factor in that method.
Senator Stedman clarified that the analysis of all the assets,
including bond packages, during the last decade have been market
value and not cost.
Mr. Reynolds understood this to be so. The data submitted to Mercer
Consulting lists the market value of the portfolio as of June 30 of
each year and reflects the amount the portfolio could be sold for
at that date.
Senator Stedman requested the June 30 information provided to
Mercer for the past ten years.
10:27:30
Senator Stedman informed that he served on a municipal assembly and
that ten years ago he recalled learning of an unfunded pension
liability at that time. He wanted to assure that was an anomaly and
not embedded in the analysis.
Senator Bunde expressed confusion about the previous computer
system that did not include all teachers in the TRS system. He
questioned the practice of reporting members as terminated then
rehired for each school year. He asked how the plan could be
managed without proper accounting of the members. He asked how this
could have been allowed to occur, as it involved 90 to 95 percent
of the membership.
Mr. Reynolds clarified that only approximately five percent of the
members were reported in this manner. The consultants do not audit
the accuracy of the data provided to it from the Division of
Retirement and Benefits, although the information is reviewed to
"reasonableness". Because of the minimal number of members
affected, the discrepancies were not apparent. The conversion to
the new computer system identified the discrepancy, which resulted
in additional liability.
Senator Bunde was assured the amount was five-percent rather than
95 percent.
Mr. Reynolds overviewed the information contained on pages 7 and 8,
which is a letter dated January 3, 2005, from himself to Ms.
Millhorn, that details changes in the funded percentages for PERS
and TRS from July 1, 1992 through June 30, 2003. This information
is based solely at fair market value of assets and involves no
smoothing methods. The data could be different if smoothing methods
were employed. The most significant factors affecting funding
status for the past three to four years are investment experience
and health care costs. However, over a longer time period,
investment losses are a small contributor to funding status because
the increases in market performance from the 1990s is factored in.
Mr. Reynolds pointed out that losses due to liabilities of 18.5
percent for PERS and 26.6 percent for TRS are significantly higher
than losses due to assets of 3.8 percent for PERS and 2.7 percent
for TRS. During the analyzed time period, investments performed
below the 8.25 percent anticipated rate, although "not dramatically
less". The majority of the changes are caused by the liability
"side".
Mr. Reynolds reiterated the breakdown of the liabilities for both
plans as: health experience, health assumption changes, plan
changes, demographic experience, and non-health assumption changes.
The aforementioned letter contains a chart listing the percentage
impact on funded status of each factor in addition to the dollar
impact on the unfunded liability of each factor over the 11-year
time period.
Ms. Millhorn continued the presentation with page 9, a spreadsheet
titled, "Public Employees'/Teachers' Retirement System Information
Briefing". This delineates PERS and TRS members in the categories
of active, deferred vested, deferred and retired, and into tiers.
The total PERS population is 71,554 members, and the TRS total
population is 22,098 members. The spreadsheet also describes the
medical plans of the different tiers. As of December 2004,
approximately 27,000 members received medical benefits. This
combined with members' dependent coverage amounts to 52,000 people
with medical benefits. This number is expected to increase in the
future.
Ms. Millhorn referenced page 10, "Retiree Medical Insurance", a
table listing the monthly premium per retiree for health coverage
for each year between 1977 and 2003, the annual percentage change,
and the average compounded annual increases. The health trend
calculation was reassessed to 12 percent for approximately three
years beginning in 2002, then decreases by one-half percent
thereafter. She demonstrated the volatility of the actual rates by
pointing out that in 1993, the rate decreased by seven percent,
then increased by 37 percent in 1994 and in 1995, increased nine
percent. Because of this volatility, it is difficult to trend
increases in health care costs. Data for the year 2005 is not
included in the spreadsheet, although the actuarial benefit
consultant has indicated the premium increase would be $850, a 5.5
percent increase. The primary factor for this increase is "that it
represents less than the projected claims costs" and fewer claims
were submitted. This is beneficial for the year 2005; however, the
impacts for 2006 are unknown.
Senator Stedman requested additional information on the structure
of the medical insurance benefit. He asked if the State of Alaska
is "stand alone" with the premiums set by the experiences of just
the State, similar to a self-insured plan, or whether experience
models from outside the State system are utilized.
Ms. Millhorn replied that the actuarial benefit consultant makes
recommendations for premium increases to the Division based on
experiences of the State as well as national trends. She would
provide additional detailed information on this matter.
Ms. Millhorn informed of actions of the Division to address the
increasing medical costs given the restraints of the benefits
plans. These include positive enrollment requirements for retiree
and active populations. This is an industry standard and involves
members providing proof that claimed dependants are qualified as
such through marriage licenses, birth certificates, etc. This
practice would begin in the current year.
Ms. Millhorn told of another other efforts to utilize generic
prescription drugs in place of brand name drugs where possible.
Membership use of generic drugs has increased from 37 percent to 42
percent in the few years this has been facilitated. For each one-
percent change from brand name to generic drugs, the plan saves $1
million. There is misunderstanding regarding differences between
generic and brand name drugs and the Division is providing
education to the membership to clarify the matter.
Ms. Millhorn expressed that to the Division's "dismay" it was
discovered the plan was not in compliance in verifying coverage for
dependants between the ages of 18 and 23. The plan did not include
a requirement that these dependants provide verification of college
attendance. This omission had occurred for approximately 20 years.
The Division has taken immediate action to remedy the situation in
mailing letters to all members claiming such dependents notifying
them of the new requirement to submit school transcripts proving
full-time attendance. Of the 2,600 letters sent, the Division has
received 1,300 replies. As a result the 1,300 dependents who did
not respond have been un-enrolled in the system.
Ms. Millhorn also spoke of the annual renegotiations of third party
administrative contract with Aetna. At the latest session, after
Aetna had submitted a contract for increased fees, the Division was
successful in negotiating a reduced fee schedule. This has resulted
in an immediate savings of $350,000. The Division also negotiated a
"network guarantee" savings of approximately $26 million in which
the fee schedule for active and retiree members would be "flat"
with no increase in fees "until and unless they can guarantee to
the State of Alaska through a reconciliation process, that we will
have a realized savings of $30 million." If this occurs, Aetna
would be able to recover some of the fees, up to $470,000.
Ms. Millhorn continued to page 11 a spreadsheet titled "Public
Employees' Retirement System, System Membership by Status" and a
bar graph titled "Public Employees' Retirement System, 10-Year
Comparison of Active and Retired Members." This information is
garnered directly from the comprehensive annual financial report.
She noted that in 1993, the system had approximately 9,100 members
compared to 18,400 members in 2003, representing almost double the
number of members in a ten-year time period. The year 2004 had a
record retirement of 20,25 members. This has not occurred since the
previous Retirement Incentive Plan (RIP) of 1995.
Ms. Millhorn noted page 12 provides information on a spreadsheet
titled, "Public Employees' Retirement System, Expenses by Type
(000's omitted)" and a bar graph titled "Public Employees'
Retirement System, 10-Year Comparison of Expenses by Type". In
1995, health care benefits cost $40 million compared to
$167,360,000 in the year 2004.
Ms. Millhorn noted identical spreadsheets and bar graphs on pages
13 and 14 utilizing data for the TRS plan. Again the amount of
retirees has almost doubled in ten years.
Senator Stedman surmised from the trend that in a few years, the
number of retirees would be greater than the number of workers. He
asked if this is the forecast.
Ms. Millhorn responded that after reviewing workforce demographic
information she concluded that a larger population would be
eligible for retirement. She has not undertaken any analysis of
this, but offered to calculate the percentages of members who would
become eligible for retirement in different years.
Senator Bunde informed that the Department of Labor and Workforce
Development has demographic information indicating that the largest
segment of the population in Alaska is currently between the ages
of 30 and 50. In ten years, this age group would represent the
smallest percentage of the population with the largest segments
being senior citizens and children under the age of five.
Ms. Millhorn commented on the increase of health care benefit
expenses for TRS from $18,264,000 in 1995 to $75,601,000 in 2004.
Ms. Millhorn continued that page 15 is a spreadsheet titled "Alaska
Public Employees' and Teacher's Retirement System, Investment
Return/Medical Costs - Assumption v. Actual & Mortality, Employer
Actuarial Computed Rates and Board Adopted Rates, Rate Year FY 90 -
FY 06". This summarizes the investment return assumptions, the
recommendations of Mercer Human Resources Consulting and the actual
adopted rates of the PERS and TRS boards.
Senator Bunde requested future explanation of why the PERS plan is
governed by a regulation limiting annual increases in employer
contributions to five percent and the TRS plan has no such
regulation. He asked the Division to provide a recommendation of
whether the regulation should be changed.
Ms. Millhorn would provide this information.
Ms. Millhorn pointed out that the information on page 16, a
spreadsheet titled "Alaska Public Employees' Retirement System -
Teachers' Retirement System, Composite Employer Contribution -
Increase Amount and Total Contributions, FY 06 - 07 - 08 - 09" was
also provided to the Committee the previous legislation.
Ms. Millhorn stated that the spreadsheet on pages 17 through 20,
titled "Public Employees' Retirement System (PERS) & Teachers'
Retirement System (TRS), Composite Employer Contribution Rate -
Active Employers, Estimated FY 06 - 07 - 08 - 09 - Change in
Employer Contribution" provides a detailed outline of individual
employers. These calculations are based on five-percent increases.
Mr. Reynolds proceeded with pages 21 through 24, noting that the
information demonstrates for each of the systems, Mercer's
expectations of future actuarial calculated rates would be for the
next 25 years under various assumptions. Most of the assumptions
are cataloged in the valuation reports, although a few are shown in
more detail on page 21 for PERS and page 23 for TRS. The
assumptions for the two systems are similar.
10:49:55
Mr. Reynolds directed attention to page 21, a spreadsheet titled,
"1.5(c) Actuarial Projections - Effect of Economic Scenarios". This
lists future calculated contribution rates assuming that the
population of the active membership grows at a rate of one percent.
This rate is reflective of the analysis of population growth of the
past five years. He qualified this "appears reasonable" to the
consultant, but is "not our opinion based on your demographics as
to how your population will actually change in the future." The
data on the spreadsheet assumes that the PERS and TRS boards would
adopt the actuarial rate in each year. However, the five-percent
annual increase limitation of the PERS board and the likelihood
that the TRS board would practice similar restraint has been
factored. Future investment returns have also been factored showing
the 8.25 percent assumptions as well as a growth or optimistic
scenario of higher investment returns over the next few years, and
a pessimistic scenario of a prolonged recession in which investment
returns are "poor" for the next three years then revert to a
somewhat higher level but still less than 8.25 percent.
Mr. Reynolds noted that the chart on page 22 demonstrates the
information listed on the spreadsheet on the previous page with the
three investment return scenarios. Page 23 and 24 contain a similar
spreadsheet and graph for the TRS system. The expectation is that
the PERS rates would increase five percent per year for the next
several years until the contribution rate reaches 27 to 28 percent.
The rates are then expected to continue to gradually increase over
the next 20 years until they reach slightly more than 30 percent.
At that point, the intention of the funding method is that the
unfunded liability would be eliminated and rates would begin to
decline. The rates would vary in the event of higher investment
performance or lower investment performance.
Mr. Reynolds listed the contribution rates for the TRS plan. Rates
are expected to continue to increase by five percent per year for
the next several years until rates reach 45 percent or higher.
Rates would continue to more gradually increase to a level between
50 and 55 percent for the longer term. The rates would not be
expected to decline until after a 25-year period of paying off the
unfunded liability.
Mr. Reynolds pointed out that for each scenario the known
investment returns of 15.08 percent for FY 04 have been factored.
Senator Stedman asked about asset allocation.
Ms. Millhorn indicated this subject would be addressed at the next
meeting.
Senator Stedman noted the targeted rate of 8.25 percent, which
"varies your allocation, in other words you set your target rate
and you've got to set your allocation to attain it or visa versa."
He expressed curiosity on the probability of attaining that target
rate. He also asked the impact of a lower rate or loss. These
issues would be discussed at the next hearing.
Ms. Millhorn stated that the timelines for the PERS plan on page 25
and the TRS plan on page 26 show the impact of significant
legislation to systems.
Senator Bunde requested future discussion on the impact of
Retirement Incentive Programs (RIP) on the systems.
Co-Chair Green announced that the discussion would continue at the
following meeting.
ADJOURNMENT
Co-Chair Green adjourned the meeting at 10:57 AM
| Document Name | Date/Time | Subjects |
|---|