Florida GFOA Webinar Series
June 19, 2014
Donald E Fuerst, Senior Pension Fellow
American Academy of Actuaries
James J. Rizzo, Senior Consultant and Actuary
Gabriel, Roeder, Smith & Company
1

DB Plans and DC Plans
• There has been a debate swirling around corporate and governmental employers for decades as to which is better:
– Defined benefit (DB) retirement plans, or
– Defined contribution (DC) retirement plans
• A hybrid plan is a good alternative resolution to the
DC - DB debate
• But let’s defined the terms first by examining their characteristics
2

DB vs. DC Debate
• Traditional Defined Benefit (DB) plans
– Benefits paid are defined by formulas and rules
– Contributions by the employer are actuarially determined
– Pension plans usually pay monthly pensions for life
– Like the pension part of FRS
– Like local police and fire pension plans
– Like Social Security
• Traditional Defined Contribution (DC) plans
– Employer contributions are defined by a formula
– Account balance plans that credit interest equal to the actual earnings of underlying investment assets
– Like the Investment Plan part of FRS
– Like 401(k) plans in the private sector
– Like so-called 401(a) plans and 457 plans in the government sector
3

Distinguishing Features
1. Individual Account Balances
2. Account Interest Credited
3. Investment Risk and Reward*
4. Predictability of Contributions*
5. Unfunded Actuarial Accrued Liability*
6. Retirement Planning*
7. Longevity and Other Risks*
8. Benefit Skew*
9. Form of Benefit*
10. Portability
11. Vesting
12. Funded Status
13. Operational Expenses*
14. Education and Communication
* Most important distinctions
4

DB vs. DC Comparison
Core Features
(Page 1 of 2)
Individual Account Balances
Traditional Defined Contribution (DC)
Retirement Plans
Traditional Defined Benefit (DB)
Retirement Plans
Yes
No individual account balances, just a defined benefit promise; pooled assets and pooled liabilities
Account Interest Credited Actual investment return of the underlying assets Not applicable
Investment Risk/Reward Borne 100% by the employee Borne 100% by the employer
Predictability of Employer
Contributions
Unfunded Actuarial Accrued
Liability
Retirement planning
Longevity and Other Risks
100% predictable
None
Unpredictable due to primary reliance on unpredictable investment earnings
Borne 100% by the employee
Much less predictable due to contribution being subject to investment performance and demographic experience
Yes, if less than 100% funded
Employees can plan on a predictable, stated (or easily calculated) percent of pre-retirement income; employers can design the plan toward a given replacement of pre-retirment income
Borne 100% by the employer
5

DB vs. DC Comparison
Core Features
(Page 2 of 2)
Traditional Defined Contribution (DC)
Retirement Plans
Traditional Defined Benefit (DB)
Retirement Plans
Benefit Skew
Benefits are skewed toward younger shorter service employees
Benefits are skewed toward older, longer service employees
Form of Benefit Almost always paid as a lump sum
Paid as a lifetime pension with options for survivorship
Portability
Vesting
Funded Status
Operational Expenses
Education and
Communications
Can be rolled over to an IRA; often not used for retirement income
Flexible in the design
Always 100% funded
Retained and paid as a lifetime pension upon retirement eligibility
Flexible in the design
Funding Policy determines funded status; rarely
100% funded
Higher investment-related expenses (in bps) than DB plans due to use of mutual funds, often subsidizing recordkeeping and communications expenses
Lower total operational expenses (in bps) than traditional DC plans; often soft dollar rebates from transaction costs reduce operational expenses further
Employee education meetings and materials required for effective investment elections; often name-brand Autopilot, little flair and fanfare; often unappreciated mutual funds, online information and investment choices; quarterly account balance statements
(until taken away)
6

Hybrid Plans
• Consider a hybrid plan as:
– An alternative to traditional DC and DB plans
– A creative solution
– Out-of-the-box thinking
7

Hybrid Plans
• A hybrid plan is a single plan that has some features of a DB plan and some features of a DC plan
• An arrangement with side-by-side DB and DC plans
– Two separate plans
– This arrangement is not really a hybrid plan
– Although some people use the term “hybrid” when describing a sideby-side DB and DC
– But consider a Toyota Prius – gasoline and electric in one car
• Two broad types of hybrid plans
1.
Cash Balance Plans
2.
Variable Annuity or Variable Benefit Plans
8

Hybrid Plans
2. Variable Annuity or Variable Benefit
– “Looks” a lot like a DB plan
• Lifetime pensions paid
• Rewards long-service employees
– Monthly benefit amount varies (up and down) depending on certain trigger points built into the plan design
• Investment Trigger -- Benefits change or vary (up and down) depending on level of investment returns, OR
• Employer Contribution Trigger – Benefits change or vary (up and down) depending on the level of employer contributions otherwise required
9

Hybrid Plans
2. Variable Annuity or Variable Benefit (continued)
– Investment Return Trigger –
• Some designs (e.g., Variable Annuity Plan) index the benefit earned for the year to retirement date using unit values based on investment returns
– higher indexing for higher returns; lower indexing for lower returns.
• Other designs can change the multiplier for the current year depending on investment returns for the year - higher multipliers for higher returns; lower multipliers or zero for lower returns.
• Still other plan designs pay additional “dividends” on benefits for higher returns.
10

Hybrids Plans
2. Variable Annuity or Variable Benefit (continued)
– Employer Contribution Trigger –
• The multiplier changes in order to keep the employer contribution predictably within a pre-set corridor – higher multiplier if employer contribution would be low; lower multiplier if employer contribution would be high
• Some designs can change the multiplier only for that year of trigger, while others change it retroactive to plan start date
• Some designs can retain the final average earnings concept instead of a career average (or career accumulation) earnings found in most investment trigger variable annuity hybrids or all cash balance hybrids
11

Hybrid Features
Core Features
(Page 1 of 4)
Traditional
Defined
Contribution
(DC) Retirement
Plans
Hybrid Plan Designs
Cash Balance Variable Benefit
Traditional Defined
Benefit (DB)
Retirement Plans
Individual Account Balances Yes Yes
Some designs have individual annuity balances; some do not
No individual account balances, just a defined benefit promise; pooled assets and pooled liabilities
Account Interest Credited
Older designs apply a fixed rate or tied
Actual investment to short-term yields. Alternatively, can return of the underlying assets be partially related to actual return on underlying assets (e.g., with smoothing,
Designs with annuity balances might index benefits to final average earnings, while others index to CPI; and still others can be partially related to actual floors and caps) return on underlying assets
Not applicable
Investment Risk/Reward
Borne 100% by the employee
Older designs put 100% on employers; recent designs share the risks/rewards
Some designs share the risks/rewards between employee and employer; while between employee and employer (e.g., fund return with floors and caps) other designs put 100% of the risk on employees.
Borne 100% by the employer
Predictability of Employer
Contributions
100% predictable
Older designs are as unpredictable as traditional DBs; recent designs that share risk have predictability in between traditional DBs and DCs
Less predictable than DC plans but more predictable than traditional DB plans
Much less predictable due to contribution being subject to investment performance and demographic experience
12

Hybrid Features
Core Features
(Page 2 of 4)
Traditional
Defined
Contribution
(DC) Retirement
Plans
Hybrid Plan Designs
Cash Balance Variable Benefit
Traditional Defined
Benefit (DB)
Retirement Plans
Unfunded Actuarial Accrued
Liability
None Yes, if less than 100% funded Yes, if less than 100% funded Yes, if less than 100% funded
Retirement planning
Longevity and Other Risks
Unpredictable due to primary reliance on unpredictable investment earnings
Some designs have ending cash balances that depend only on future salary and are just as predictable as DB plans; but most designs depend mostly or partially on unpredictable investment returns or yields
Some designs depend mostly or partially on unpredictable investment returns, while others have future benefits that fall within a predictable minimum and maximum
Employees can plan on a predictable, stated (or easily calculated) percent of preretirement income; employers can design the plan toward a given replacement of pre-retirment income
Borne 100% by the employee
Almost always borne 100% by the employee
Borne 100% by the employer Borne 100% by the employer
Benefit Skew
Benefits are skewed toward younger shorter service employees
Skewed toward younger, shorter service employees
Skewed toward older, longer service employees
Benefits are skewed toward older, longer service employees
13

Hybrid Features
Core Features
(Page 3 of 4)
Form of Benefit
Traditional
Defined
Contribution
(DC) Retirement
Plans
Hybrid Plan Designs
Cash Balance Variable Benefit
Traditional Defined
Benefit (DB)
Retirement Plans
Almost always paid as a lump sum
Almost always paid as a lump sum, unless prevented from doing so by plan design
Paid as a lifetime pension with options for survivorship
Paid as a lifetime pension with options for survivorship
Portability
Vesting
Funded Status
Operational Expenses
Can be rolled over to an
IRA; often not used for retirement income
Can be rolled over to an IRA; often not used for retirement income
Retained and paid as a lifetime pension upon retirement eligibility
Retained and paid as a lifetime pension upon retirement eligibility
Flexible in the design
Always 100% funded
Flexible in the design
Funding Policy determines funded status
Flexible in the design
Funding Policy determines funded status
Flexible in the design
Funding Policy determines funded status; rarely 100% funded
Higher investmentrelated expenses (in bps) than DB plans due to use of mutual funds, often subsidizing recordkeeping and communications expenses
Lower total operational expenses (in bps) than traditional DC plans; approximately same as traditional DB plans
Lower total operational expenses (in bps) than traditional DC plans; approximately same as traditional DB plans
Lower total operational expenses (in bps) than traditional DC plans; often soft dollar rebates from transaction costs reduce operational expenses further
14

Hybrid Features
Core Features
(Page 4 of 4)
Education and
Communications
Traditional
Defined
Contribution
(DC) Retirement
Plans
Hybrid Plan Designs
Cash Balance Variable Benefit
Traditional Defined
Benefit (DB)
Retirement Plans
Employee education meetings and materials required for effective investment elections; often name-brand mutual funds, online information and investment choices; quarterly account balance statements
Minimal employee communications, but depending on method of crediting interest; quarterly or annual account balance statements
Employee education is needed for how accrual multipliers might vary
Autopilot, little flair and fanfare; often unappreciated
(until taken away)
15

Hybrids Plans
• Primary motivations for moving from DB to DC plans
1.
“The corporate world has moved from DB to DC plans.”
2.
“ We got rid of our DB plan where I work(ed).”
3.
“I never had a DB plan; neither should they.”
4.
“ The conservative think-tanks and legislatively active organizations say we should move to a DC plan.”
5.
“My political party leaders say DB plans are bad.”
6.
“DB plans are more dangerous than DC plans in the hands of politicians.”
7.
“I don’t trust elected officials to stand up against the unions; they often grant retroactive benefit improvements in DBs.”
8.
“Employer contributions to our DB plans have become unbearably and unreasonably high.”
9.
“Employer contributions need to be more predictable.”
10.
“Employer (taxpayers) should not bear the investment risk.”
16

Positions vs. Interests
• Management and elected officials sometimes START with the
“position” of advocating DC plans over DB plans
• Consider moving the dialogue from “positions” to “interests” or “principles”
• Identify the underlying interests of the parties and resolve the matter with solutions and proposals:
– That appeal to their “interests”, rather than
– Fight their positions
17

• A pension plan where the annual pension benefit fluctuates (both before and after retirement) with the performance of investment funds
• Advantages for the employer:
– Financial predictability: stable cost, little or no unfunded liabilities
– Retention: allocates benefits to long-service employees
• Advantages for the employee:
– Provides lifetime income – employee can’t outlive assets
– Portable benefit, potential inflation protection
18

• Career accumulation plan
• Benefit credit is converted to variable annuity units at yearend purchase price of the units
• Units accumulate throughout an employee’s career
–
• At NRD employee receives an annual retirement income based on number of “variable units” accumulated
• Annual income in retirement for each unit is the unit value at the end of the previous year
19

•
– Example: 1% of pay
– Formula is applied to actual pay each year
– Benefit is annual annuity at normal retirement age
• Participant earns $50,000
• Participant accrues a benefit = $500/year payable at 65
•
– Example: Variable unit value = $10
– $500 benefit buys 50 variable units
20

A variable unit is the right to an annual pension benefit equal to the unit value beginning at NRD
• Participant’s 50 units would generate $500/year at retirement if the unit value remains $10
– If unit value increases to $11, 50 units would generate $550/year at retirement
– If unit value decreases to $9, 50 units would generate $450/year at retirement
• If unit value increases by average 2%/ year for 25 years
– Unit value is $16.41 when Participant retires at age 65
– 50 variable units earned when age 40 then generate $820.50 annual pension income
• Unit value can continue to fluctuate during retirement
– If the value falls 1% to $16.25 during first year of retirement, benefit in year
2 will decline to $812.50
21

$2,541
$2,139
$1,502
$1,045
Units earned
Accumulated units
End-of-year unit value
Year-end pension income
$500
50
50
$10.00
$500
45
95
$11.00
$1,045
48
143
$10.50
$1,502
43
186
$11.50
$2,139
45
231
$11.00
$2,541
22

• Initial unit value can be any value
• Unit value fluctuates annually based on the actual return of the pension fund (for the previous year) relative to a benchmark rate called the hurdle rate
• If assets earn more than hurdle rate, unit values are increased
• If assets earn less than hurdle rate, unit values are decreased
• For example, if hurdle is 5% and assets earn 8% (or 2%), unit values increase (or decrease) approximately 3%
• VAP is equivalent to the plan sponsor funding a fixed annuity determined at an interest rate equal to the hurdle rate
– Investment experience above or below the hurdle rate is passed through to the participant through a change in the unit value
23

Formula for Unit Value to pass all investment gains or losses to participant is:
UV t
= UV t-1 x (1 + i t
) / (1 + h) where i t rate is the actual return during the period t-1 to t and h is the hurdle
24

• The hurdle rate is a plan provision, not an actuarial assumption, although sometimes referred to as assumed interest rate or AIR
• Hurdle rate is a critical component of plan design
– Higher rate lowers plan costs, but increases the likelihood that participants’ benefits will decrease
– Lower rate increases plan costs, but increases the likelihood that participants’ benefits will increase
– Setting the hurdle rate at (or a little above) a reasonable risk-free rate of return enhances certain features
• Minimum permissible hurdle rate is 3%
• Plans with hurdle rate < 5% are subject to statutory hybrid rules (private sector concern)
25

26

•
•
•
27

Back to back declines three times in 50 years:
1973-74: 32% Recovered 92% in 2 years, 100% in 6
2000-01: 18% Recovered 100% in 2 years
2008-09: 17% Recovered fully in 2013
28

Back to back declines three times in 50 years:
1973-74: 34% Recovered 88% in 2 years, 100% in 8
2000-01: 20% Recovered 100% in 3 years
2008-09: 19% Still 4% under peak value
29

VAPs compare favorably to traditional DB plans in several respects:
• Benefits are comparable to Final Average Pay plans
• Benefits are more portable than Final Average Pay plans
• Benefits provide potential inflation protection
• Variations of VAP offer participant ability to manage risk exposure
30

FAP Benefit Formula = 1% of Final Average Pay for each year of service
RSP Benefit Formula = 1% of Annual Pay, Assumed Interest Rate = 4%
Average pay increases over career = 3%
Retire at 65 with 30 years Service
Benefit expressed as Percent of Final Average Pay
Average
Investment Return
5%
6%
7%
8%
FAP Benefit
30%
30%
30%
30%
Variable Benefit
24.3%
27.7%
31.8%
36.7%
31

Variable units continue to appreciate for terminated vested participants
• Benefits continue to grow with investment experience, unlike traditional pension benefits
• Participants are not harmed by changing employment as in a final pay plan
• Benefits are not paid in lump sum or rollover, thus preventing leakage
• EBRI study indicates as much as 60% of early distributions are not rolled over
32

60
50
40
30
20
10
0
Final Pay
Plan - One
Job
Final Pay
Plan -
Multiple
Jobs
Variable
Annuity Plan
One Job
Variable
Annuity Plan
Multiple
Jobs
Job 5
Job 4
Job 3
Job 2
Job 1
33

34

Variable Benefit versus Purchasing Power
(1961-2012)
$12.00
$10.00
$8.00
$6.00
$4.00
$2.00
$-
1961 1966 1971 1976 1981 1986
Year
Variable Benefit with 4% Hurdle
Purchasing power based on CPI-U from BLS
1991 1996 2001
Constant Purchasing Power
2006 2011
35

•
–
–
36

•
•
•
•
•
37

Each unit represents a sub-account of the pension trust
• Stable Units: High-quality shortto-intermediate term fixed income securities.
• Diversified Units: Generally, an actively managed fund where plan sponsor selects the investment managers and asset allocation. Higher potential return than Stable Shares but more volatile.
38

Each unit fund represents a sub-account of the pension trust
• Stable units: High-quality short-to-intermediate term fixed income securities.
• Equity units: 100% equity investment, actively managed or an index fund.
• Diversified units: Generally, an actively managed fund where plan sponsor selects the investment managers and asset allocation. Less risky than
Equity units.

Pension fund allocation
Stable
Equity
Diversified
Other variations are possible :
- More/less classes of shares
- Different classes of shares
(e.g. Lifecycle shares)
39

• Employees can structure the risk/reward of their retirement portfolio based on their own circumstances and tolerance levels
• Participants elect once per year which class(es) of shares they wish to purchase at year-end
• Could purchase just one type of share or all three share classes
• Also can exchange or reallocate any shares previously earned
• Statement sent in first quarter of following year shows the results of the share purchase and/or exchanges
40

• Conversion from existing DB Plan
– Simplest approach is to freeze current plan benefits
• VAP unit balance begins at zero and grows
• Ultimate retirement benefit is: frozen old-plan benefit plus benefit from RSP
• Risk due to the old frozen plan benefits still exists under this approach
– Other approaches, including converting accrued benefits, are possible
• Can significantly reduce the risk associated with prior accrued benefits quickly and cost effectively
• Integrating VAP into existing DC Plans as payment form
41

42

•
•
•
•
•
•
– Arise from demographic experience
– Usually smaller in magnitude
43

Case Study: City of Ocala GE
• Driving principles/interests from City Council
– Roll back future benefits to bend the cost curve soon
– Future benefit levels should be adequate and competitive
– Share risks between employees and employer
– Make employer contributions more predictable
44

Case Study
• Bend the expected cost curve
– Changing benefits for new hires alone will take a very long time to bend the expected cost curve
• So putting in a DC plan for new hires alone won’t do much soon enough
• Putting in a new DB formula for new hires alone won’t do much either
– Must change benefits for all employees (current and new) in order to bend the expected cost curve in a reasonably short time
• Either all in a DC or
• All in a less generous benefit structure for future service
45

Case Study
• Bend the expected cost curve (continued)
– Not permitted to roll back future benefits earned by current active employees eligible for normal retirement
– Not permitted to roll back benefits retroactively, i.e.:
• Not permitted to cut benefits for current retirees and beneficiaries
• Not permitted to cut accrued benefits below what active employee have earned so far
46

Case Study
• Bending the expected cost curve (continued)
– Freeze the benefits for current actives at what they earned as of the transition date; call it Part A
– Start the new and less generous benefit structure for future service; call it part B
– Final benefit is Part A plus Part B
– Grandfather current active employees within five years of Normal
Retirement Date
47

Case Study
• Make the employer contribution more predictable . . .
• By sharing the risk with employees.
– Moving to a DC plan for new hires alone will take a very long time to share the risk
– Not permitted to share the risk with current retirees or with current active employees eligible for normal retirement
– Not permitted to share the risk on benefits earned at transition
• Part B benefit structure is the hybrid plan design
– Part B monthly projected benefit can go up or down, depending on the trigger mechanism (total contributions), thereby sharing the risk and reward
– Part A monthly benefit is fixed and frozen at the transition date
48

Case Study
Plan Asset Growth Over Time
$350
$300
$250
$200
$150
$100
$50
$0
2012 2017 2022 2027 2032
Closed DB Plan
2037 2042 2047
DC Plan for New Hires
2052 2057
$200
$150
$100
$50
2062
$0
$350
$300
$250
49

Case Study
• Without a VBH feature in the legacy DB plan -- It will take over 40 years to reach a 50-50 risk-sharing with employees; in 20 years
City/taxpayers still bear 85% of investment risk
Total Investment Risk-sharing Comparison
Closed DB plus DC for New Hires
With and Without VBH Feature
100% 100%
Employees Bear
100% Risk
50% 50%
City/taxpayers
Bear 100% Risk
0%
2012 2022 2032
Closed DB w/ VBH, plus DC for New Hires
2042 2052 2062
Closed DB w/o VBH, plus DC for New Hires
2072
0%
50

Case Study
• Consider DC, cash balance, variable annuity plans
– Implemented for benefits earned in the future
– Even if covering current employees’ future service and new hires
• Most do nothing about the massive legacy obligations of Part A
– No risk sharing
– Costs still unpredictable
• Ocala’s variable benefit hybrid design did
– The employer and employee share the risks (for the legacy Part A)
– The total costs (for Part A and Part B combined) are more predictable
51

Case Study
• Part B starts out with a 1.6% multiplier, like FRS Regular Class
– Multiplier can go up, but not above a cap of 2.55% and
– Multiplier can go down, but not below a floor of 1.0% --
– Depending on the level of the actuarially required contribution.
– As long as the actuarially determined contribution (ADC) stays within a pre-set employer contribution budget, the multiplier remains unchanged; improves predictability
– Makes the employer contribution more predictable
52

Case Study
• If the ADC were to go above the top of the budget corridor:
– The multiplier is reduced in order to keep the ADC inside the corridor
– Employee bears the risk above the corridor.
• If the ADC were to go below the bottom of the budget corridor:
– The multiplier is increased in order to keep the ADC inside the corridor
– Employee reaps the reward below the corridor
• Makes the employer contribution more predictable by sharing the risk (and reward) with the employee
53

With a VBH
Feature:
Budget
Corridor
60%
50%
40%
30%
20%
10%
0%
Case Study
Total Projected City Contributions
Closed DB Plan with Future Benefits Rolled Back to 1.6%
With 8% DC Plan for New Hires
Positive and Negative Stresses at Year 11 Outside the VBH Corridor
Corridor
Stress Test on City Alternative 2
Fiscal Year Ending
City Alternative 2
Better than Expected Asset Returns on City Alternative 2
54

Disclaimers
• Circular 230 Notice: Pursuant to regulations issued by the IRS, to the extent this presentation concerns tax matters, it is not intended or written to be used, and cannot be used, for the purpose of (i) avoiding tax-related penalties under the
Internal Revenue Code or (ii) marketing or recommending to another party any tax-related matter addressed within. Each taxpayer should seek advice based on the individual’s circumstances from an independent tax advisor.
• This presentation shall not be construed to provide tax advice, legal advice or investment advice.
• Readers are cautioned to examine original source materials and to consult with subject matter experts before making decisions related to the subject matter of this presentation.
• This presentation is not an expression of the views of FGFOA, or Gabriel, Roeder,
Smith & Company, or the American Academy of Actuaries and may not even express the views of the speakers.
55

56

57