HomeMy WebLinkAbout07.a.1. Contra Costa County Employees' Retirement Association Update7. a.-~
Central Contra Costa Sanitary District
March 12, 2010
TO: HONORABLE BOARD OF DIRECTORS
FROM: JAMES M. KELLY, GENERAL MANAGER p
RANDALL M. MUSGRAVES, DIRECTOR AD INISTRATION /`~~~'
SUBJECT: CONTRA COSTA COUNTY EMPLOYEES' RETIREMENT
ASSOCIATION (CCCERA) UPDATE
UPDATE
The CCCERA Board met on March 10, 2010 and discussed de-pooling, update on five
year projection of employer contribution costs, economic assumptions to be used in the
Triennial Experience Study and the Valuation Study, final compensation for retired and
active members and final compensation for new members.
De-pooling (Item #6) - CCCERA's Actuary, Segal, discussed that they were having
problems with the Retirement System's data in order to develop assets and liabilities by
employer, see attachment. CCCERA is the only California 1937 Act County Retirement
system to group all retirees into one pool. Most others pool retirees by Tier. On October
14, 2009,, the CCCERA Board voted to de-pool the System's assets and liabilities by
employer. The Actuary found that the computer system used by CCCERA staff does not
collect the data in a manner to allow accurate accounting of employees and retirees to
accomplish distribution by employer. The system does not currently have the ability to
track the data going forward at a level to support accurate employer identification. After
much debate, the CCCERA Board created an Ad Hoc Committee to meet with staff and the
Actuary to evaluate the issues and alternatives. The Committee's meetings are to be open
to the public. District staff will be attending the meetings. Mr. John Bartel, an Actuary, has
been retained to assist staff in evaluating the Actuary's work and the issues as they impact
the District.
Update on 5 year projection of employer contribution costs (Item #7) - A presentation was
made by the Actuary, Segal, updating the Board as to the impact of the 21.9% rate of
return on investments for 2009, see attachment. In summary, while the news is good; it
appears the projected employer contribution rate for 2013 may be 39.5% instead of the
previous estimate of 46% for the entire pool. You will recall that CCCERA experienced a
27.9% loss in 2008. The District's current percent of payroll is 39.4% and using this
information would increase to 52.58% of payroll, or an additional $3,000,000 per year. The
difference in rate between the pool and the District is due to the County selling pension
obligation bonds to reduce their UAAL. The District's rate has a UAAL component. The
Actuary cautioned the use of theses numbers because they must complete the Triennial
Experience Study and the Valuation Study (May/June 2010) in orderto provide an accurate
updated employer contribution rate.
1
Economic assumptions to be used in the Triennial Experience Study and the Valuation
Study (Item #8) -The Actuary, Segal, reported their work, research and recommendations
to the CCCERA Board on a number of assumptions used to develop future funding
requirements, resulting in employer and employee contribution rates.
Investment Return -Reduced from 7.8% to 7.75%.
CaIPERS and CaISTRS uses 7.75%.
The formula used is Inflation+ Portfolio real rate of return -Expense adjustment
- Risk adjustment =Investment return.
Inflation -Reduced from 3.75% to 3.5%.
Data looked at the 15 year national CPI average and CCCERA's investment
consultant's short term projection for inflation at 2.75%.
Individual Salary Increases -Reduce increase assumption from 3.75% to 3.5%
and increase real "across the board" salary increase assumption from 0.5% to
0.75%. Promotional and merit increases will be reviewed as part of the triennial
experience study of non-economic assumptions.
Terminal Pay -Recommendation will be included in the triennial experience
study of non-economic assumptions..
Final compensation for retired and active members (Item #10) - The CCCERA Board voted
7-0, with one abstention, to maintain the current calculation of final annual salary for retired
and active members. They stated that their decision was based upon (1) fairness; (2)
recalculation would be impossible, extremely costly and an administrative nightmare; (3)
the benefits are legally protected due to vested rights and reliance; (4) when the Ventura
decision was finalized three legal firms and the Court confirmed the calculation of final
annual compensation; and (5) the Board was acting on legal advice from their attorneys.
Final compensation for new members (Item #11) -The CCCERA Board adopted the
attached addendum to the current policy for new employees, effective January 1, 2011.
January 1 was selected to provide sufficient time to clarify and address Board's concern
regarding deferred compensation payments as cash in the final year and reciprocity issues.
It also would provide time for the State legislature to adopt legislation addressing spiking
of terminal pay. The Board voiced concern about creating multiple tiers.
FOLLOWUP
The following is a list and status of Board requested information regarding the District's
retirement benefit and the Contra Costa County Employees' Retirement Association.
Please advise staff if other information is desired.
2
1. Legal opinion regarding CCCERA's and the District's fiduciary and legal
responsibilities. To be provided by District Counsel, Kenton Alm.
2. A review of the structure in which the retirement board functions and their role,
responsibilities, and how they are governed. To be provided by District
Counsel. Kenton Alm.
3. A review of District's ability to address spiking by modifying benefits and
MOUs. Submitted to the Human Resource Committee at the February 22
2010 meeting.
4. A comparison of District retirement benefits to other agencies, including Contra
Costa County. Submitted to the Human Resource Committee at the February
22, 2010 meeting.
5. A total compensation comparison to other agencies. Submitted to fhe Human
Resource Committee at the February 22, 2010 meeting.
6. A report of CCCSD employees with accrued vacation exceeding 160 hours
(four weeks). This report was provided to the Board.
7. A memo attaching past reports regarding moving the retirement program to
CaIPERS with updated information and market changes. Submitted to the
Human Resource Committee at the February 22, 2010 meeting.
The District Board may address the need for a workshop regarding the retirement benefit
and CCCERA at a later time.
Staff will be available at the Board meeting to address any questions the and receive
any additional request for information.
'~SEGAL
THE SEGAL COMPANY
100 Montgomery Streel. Suite 500 San Francisco. CA 94104-4308
T 415.263.8273 F 415.263.8290 www.segalco.cam
March 2, 2010
Ms. Marilyn Leedom
Chief Executive Officer
Contra Costa County Employees' Retirement Association
1355 Willow Way, Suite 221
Concord, CA 94520
Re: Contra Costa County Employees' Retirement Association
Data Issues in Depooling Study
Dear Marilyn:
MEETING DATE
MAR 10 2010
AGENDA ITEM
Paul Angelo, FSA, MAAA, FCA
Senior Vice president & Actuary
pangeloQsegalco.com
This letter_provides information related to the various data issues pertaining to the depooling
study that we have discussed with you. Based on those issues we also aze providing discussion
of some alternative applications of the depooling that we recommend should be considered by
the Board.
Background
At its October 14, 2009 meeting, the Boazd took action to depool CCCERA's assets, Actuarial
Accrued Liability (AAL) and Normal Cost both by tier and by employer for determining
employer contribution rates. However, the smaller employers (those with less than 50 active
members) will be pooled with the applicable County tier.
The Board action also included a retroactive application of the depooling back to December 31,
2002. This retroactive approach would not involve recalculation of employer rates prior to
December 31, 2008. However, it would involve reflecting the separate experience of the
employers in each individual cost.group back to December 31, 2002.
In the results provided in our January 6, 20101etter, we used an implementation date of
December 31, 2008 to illustrate the effect of this depooling as if it were implemented on that
date, on a retroactive basis. In practice, the implementation date would. be December 31, 2009,
so as to avoid changing the rates that were already adopted based on the already completed
December 31, 2008 valuation. The first actual employer rates affected by depooling would be
those in the December 31, 2009 valuation.
The January 6, 2010 letter contained results for General members only. This was due to the
unresolved data issues for Safety members as discussed below.
Benefits, Compensation and HR Coneulting ATUNTA BOSTON CALGARY CHIUGO CLEVEUND DENVER HARTFORD HOUSTON l05 ANGELES
MINNEAPOLIS NEW ORLEANS NEW YORK PHILADELPHIA P/1DENI% PRINCETON RALEIGH SAN FRANCISCO 70ROM0 WASHINGTON, OC
M G
MulGnatlonal Group of Aetuarle• and Consultants BARCELONA BRUSSELS DUBLIN GENEVA HAMBURG JOHANNESBURG LONDON MELBOURNE
A ~ C ME%ICO CITY OSLO PARIS
Ms. Marilyn Leedom
March 2, 2010
Page 2
Data Issues Encountered
In late January, we were provided with new data for Safety members that are currently retired.
This data was then used as the basis for updating our illustrations of the cost impact of
depooling.
Based on our further discussions with CCCERA, we believe that it is appropriate to not release
any results based on the updated data. This letter documents how the results critically depend
on the data used, data issues encountered and potential alternatives that the Board might want
to consider that would be less data intensive.
The illustrations of the cost impact of depooling depend critically on the employer by employer
contribution and benefit payment data reported for the six year period from December 31, 2002
through December 31, 2008, and on the employer by employer member data reported as of
December 31, 2002 and December 31, 2008. For example, some employer's contribution rates
show either a decrease or an increase depending on whether the original data or the revised data
is used.
Most of the data issues encountered so far were related to the prior inclusion of various fire
districts data in County payroll. CCCERA verified the accuracy of the employer and tier codes
for Safety retired (including survivors) and deferred vested members that were reported as ,
Safety members in the data originally provided to us for depooling and historical actuarial
valuations. As a result of this verification, many employer codes were changed for these
members. This new data was then used as the basis for updating our illustrations of the cost
impact of depooling. However, there is no guarantee that this data is fully consistent with the
actual employer these members worked for. For example, there could be Safety members that
are currently being classified as General members in the data originally provided to us for
depooling and historical actuarial valuations. This is not unexpected; this employer specific
historical data was never tracked because it was never expected that this level of employer
specific costs would be required.
The next data issue concerns members that worked for more than one employer and/or both
General and Safety tiers during their career. The illustrations of the cost impact of depooling
allocate each member's total liability to their last employer and tier. Therefore, for any
members that changed employers and/or were in both General and Safety tiers during their
career, the final employer and tier absorbs all of the liability associated with that member. For
example, if a County General member transfers to a Safety position late in their career at a fire
district, all of this member's County General liability will be allocated to the fire district along
with this member's Safety liability. Note that there are no offsetting adjustments or transfers of
assets made for such employees who worked for more than one employer and/or General and
Safety tiers during their careers.
5071166v1/05337.001
Ms. Marilyn Leedom
March 2, 2010
Page 3
The allocation of a member's liability amongst multiple employers and/or tiers presents
significant challenges. For tiers, we might be able to allocate the various portion of a member's
liability. However, we would have to.work with CCCEItA to determine the feasibility of
implementing this in the future. However, even if CCCEItA was able to provide us more
detailed data reflecting information for each emplover the member worked for, allocating
service among multiple employers would present significant complications relating to the
calculations involved in the actuarial valuation.
Alternative Considerations
Due to the data issues involved, we believe that it would be appropriate for the Board to
consider modifying the depooling decision in a way that is still generally consistent with the
intent of the action, but also more consistent with the data available and the fact that members
transfer between employers.
In particular, we believe that it is still reasonable and practical to depool UAAL across tiers.
This would generally be consistent with how the normal cost was pooled in the December 31,
2008 valuation prior to the depooling action. This would also include the key step of separating
and tracking CCCEItA's nonactive liabilities (including retirees) by tier.
Please note that this would not address the data issue related to members' transferring across
tiers. However, under this depooling structure, the liability associated with a member that
transfers across tiers would at least be pooled amongst all the employers in that tier. As
previously noted, we could work with CCCERA to determine if it would be feasible to
implement allocating liabilities across tiers.
Also, there is an exception in that the UAAL for the tiers that are generally closed to County
new~hires (General Tier 1 Enhanced and Safety Tier f~Enhanced) would continue to get pooled
with the corresponding open tier (as was also done in t e depooling study).
One reason for this recommendation involves what we understand was the leading rationale for
depooling by employer, which was to allocate the cost of terminal pay policies to the employers
that allowed such practices. Our results so far indicate that depooling may not accomplish this
intent because terminal pay results for specific members may not be consistent with the overall
experience of that employer. Also, we understand that some employers are already taking
action to control such practices.
If it is decided to depool by tiers, but not by employers, then it would also be reasonable to
apply depooling on a prospective basis from the December 31, 2008 valuation. This will
eliminate most, if not all of the historical data issues that were encountered.
6071186v1l06337.001
Ms. Marilyn Leedom
March 2, 2010
Page 4
Finally, we would continue to recommend the elimination of the two specific cost-sharing
adjustments that were for General Tier 1 and Safety Tier A. The previous Board action on
depooling did eliminate those adjustments.
If there is any additional information you require, please let us know
Sincerely,
~'-`~ ~~
Paul Angelo
JZM/kek
5071186v1/05337.001
MEETING DATE
~SEGAL
THE SEGAL COMPANY
100 Montgomery Street. Suile 500 San Francisco. CA 94104-4308
T 415.26J.8200 F 415.263.8290 wwMr.segalco.com
March 2, 2010
Ms. Marilyn Leedom
Chief Executive Officer
Contra Costa County Employees' Retirement Association
1355 Willow Way, Suite 221
Concord, CA 94520
Re: Contra Costa County Employees' Retirement Association (CCCERA)
Projection otRetirementPlan Costs
Dear Mazilyn:
Maa 1 o zoo
AGEN
As requested by the Association, we have projected the employer contribution rates for the
CCCERA Retirement Plan over the next five years, based on results from the December 31,
2008 actuarial valuation. "The results of our projections are shown in the enclosed exhibit.
Description of Projections
We understand that the market value rate of return was about 21.9% for 2009. In comparison,
CCCERA's current investment rate of return assumption for valuation purposes is 7.8% per
year. For the projections, we have utilized the actual market value return of 21.9% for 2009 and
have assumed that the fund will earn a market value rate of return of 7.8% for each year after
December 31, 2009, consistent with our current valuation assumption.
'The projections aze based on the actuarial assumptions and census data used in our
December 31, 2008 valuation for the Retirement Plan. Future experience is expected to follow
all of those assumptions, except as noted above.
Note that we are in the process of performing a triennial experience study for CCCERA for the
three-year period ending December 31, 2009, along with a review of the economic
assumptions. It is our understanding that any actuarial assumptions (both economic and non-
economic) that will be adopted from those two studies will be implemented in the upcoming
December 31, 2009 valuation. As our recommended assumptions are not yet known, these
projections are based on the Plan's current actuarial assumptions, with the exception of the
actual market return for 2009, as we mentioned above.
aanefih, Compenaatlon and HR Conaulting ATUNTA BOSTON CALGARY CHICAGO CLEVEUNO OENVER HARTFORD HOUSTON LOS ANGELES
MINNEAPOLIS NEW ORLEANS NEW YORK PHIUOELFTIIA PHOENIX PRINCETON RALEIGH SAN FRANCISCO TORONTO WASHINGTON, D.C.
M °
MulUnatlonal Group of Actuarlea and Conaultinb BARCELONA BRUSSELS DUBLIN GENEVA HAMBURG JOHANNESBURG CONDOR MELBOURNE
A ~ ~ MEXICO CITY OSLO PARIS
Ms. Marilyn Leedom
March 2, 2010
Page 2
Consistent with CCCERA's current funding policy, any actuarial gains or losses in the UAAL
have been amortized over a separate t 8-yeaz period effective with the valuation date the gain or
loss first appeared.
Results
The results of our projections are shown in the enclosed Exhibit 1, which provides a graph of
the projected (aggregate) employer contribution rates, along with the incremental and
cumulative changes in the rates. Please note that the projections reflect the 18-month lag
between the valuation date and the date the contribution rates are implemented. In addition, the
employer rates do not reflect any employer subvention of the member rates or any member
subvention of the employer rates.
As described in our December 31, 2008 valuation report for the Retirement Plan, the total
unrecognized investment losses as of December 31, 2008 under the actuarial value of assets
smoothing method are $1.5 billion. These investment losses will be recognized in the
determination of the actuarial value of assets for funding purposes over the next few years. This
means that earning the assumed rate of investment return of 7.8% per year (net of expenses) on
a market value basis will still result in investment losses on the actuazial value of assets in the
next few years.
Even though the mazket value of assets earned 21.9% for the 2009 Plan Year and are assumed
to earn 7.8% per year beginning January 1, 2010 (and all other actuazial assumptions are
assumed to be met), the employer contribution rate increases over the next few years following
the December 31, 2008 valuation date as these deferred losses (partially offset by the 2009
gain) aze recognized. This is illustrated in Exhibit 1, which shows increases in the projected
employer rates for the December 31, 2009 through December 31, 2013 valuations.
Other Considerations
We emphasize that projections, by their nature, are not a guarantee of future results. The
modeling projections aze intended to serve as illustrations of future financial outcomes that are
based on the information available to us at the time the modeling is undertaken and completed,
and the agreed-upon assumptions and methodologies described herein. Emerging results may
differ significantly if the actual experience proves to be different from these assumptions or if
alternative methodologies are used. Actual experience may differ due to such variables as
demographic experience, the economy, stock market performance and the regulatory
environment.
Note that the results shown herein will be different than results based on any potential
assumption changes that the Board might adopt as a result of our upcoming triennial experience
study.
5068869x2/05337.103
Ms. Marilyn Leedom
March 2, 2010
Page 3
These projections were completed under the supervision of John Monroe, ASA, MAAA,
Enrolled Actuary.
Please let us know if you have any question regarding this letter and/or the enclosures.
Sincerely,
~~ ~ ~
~{~.1~1~oc"
Paul Angelo, FSA, MAAA, FCA
Senior Vice President & Actuary
DNA/bgb
Enclosure
John Monroe, ASA, MAAA, EA
Vice President & Associate Actuary
5068869v2/05J37.103
Exhibit 1: Projected Employer Rates
Assuming Investments Earn a Market Return of 21.9% for 2009 and 7.8% Per Year Thereafter
45°/.
40%
0 35°~
a.
m
a
w
0
M
C
m
V
`m 30%
n.
25°~
20°/.
2008 2009 2010 2011 2012 2013
Valuation Date (12/31)
Valuation Date 12/31 2008 2009 2010 2011 2012 2013
Pro acted Em to er Rate 26.0% 29.6% 32.6% 35.8% 38.9% 39.5%
Incremental Rate Chan a 3.6% 3.0% 3.2% 3.1% 0.6%
CumulaWe Rate Chan a 3.6% 6.6°~ 9.8% 12.9% 13.5%
5068869x2/05737.103 3 S E G A L
CONTRA COSTA COUNTY EMPLOYEES' RETIREMENT ASSOCIATION
ADDENDUM TO POLICY FOR DETERMINING WHICH PAY ITE~1~G DATE
"COMPENSATION" FOR RETIREMENT PURPOSES
[BAR 10 2010
PURPOSE:
This Addendum to the Policy for Determining Which Pay Items are "Compensation" or
Retirement Purposes, adopted December 5, 1997, and amended January 13, 1998 ("Policy"),
reflects the Retirement Board's re-examination of the Policy since its adoption and the
Retirement Board's decision to revise certain aspects of the Policy with respect to persons who
become members of the retirement system on and after July 1, 2010. This Addendum is made as
of March _, 2010 and shall be referenced in communications to affected members regarding the
Policy following its adoption.
POLICY:
Notwithstanding anything to the contrary in the Policy, the following policies and procedures
shall be effective as to the determination of "compensation," "compensation earnable" and "final
compensation," as defined in the County Employees Retirement Law of 1937, Government Code
sections 31450, et seq.
1. Remuneration paid in cash for time earned is considered "final compensation" and is
limited by the following:
a. Annual "cash out"
The value of accrued time, such as vacation, holiday, sick or administrative leave, that is
both earned and sold back to the employer during the final compensation period under a
"cash-out" agreement, is includable in compensation eamable.
b. Lump sum at termination
Only the portion of accrued time (such as vacation, holiday, sick or administrative leave)
that is paid in the form of a lump sum at termination, and that represents time both earned
and cashable during the final compensation period, is includable in compensation
earnable.
2. Additional examples of amounts that are not included in "compensation earnable"
a. For each year of the final compensation period, leave amounts sold back during any
twelve-month period that were accrued over two or more fiscal or calendar years, and
that exceed the amount that_was both earned and cashable during service in that twelve-
month period.
b. Incentives, bonuses and other payments to the extent they may not be received in cash
during service; but only upon termination or retirement.
c. Conversion of in-kind benefits and other advantages to cash during the final
compensation period.
3. Members affected
This Addendum shall be effective as to the calculation of retirement allowances for those
members of CCCERA whose original date of membership in the retirement system is on or after
July 1; 2010.
-2-