Update: SOA’s Retirement 20/20 Project CIA Annual Meeting, 25-26 June 2009 Speakers: Rob Brown, University of Waterloo Emily Kessler, Society of Actuaries

Update: SOA’s Retirement 20/20 Project

CIA Annual Meeting, 25-26 June 2009

Speakers:Rob Brown, University of WaterlooEmily Kessler, Society of ActuariesMartine Sohier, Watson Wyatt

2

Overview

Retirement 20/20 to date• Insurance vs. Investment• Choice vs. Default

Plan innovation in Canada Measurement Framework

• What employers think

Next steps: Retirement 20/20

3

About Retirement 20/20 Retirement 20/20 is our initiative to rethink

retirement systems What’s possible?

• Go beyond the limitations of today’s tax code How do we achieve the best outcome for

the most people? • Longer lifespan/fewer workers • Companies may not be able to shoulder the

(perceived) risk of traditional pension plans

4

Retirement 20/20 Systematically explore new retirement

systems (tier II) that go beyond DB/DC• Defining stakeholders: Society, Individuals,

Employers & Markets• Focusing on stakeholders’ needs, risks and

roles • How to improve plan function (e.g. self-

adjustments, retirement signals) What have we learned so far?

5

Current environment

Balance tipping toward choice & investment • Based on employer risks and preferences

Does this produce the optimal outcome for all stakeholders?

Insurance Investment

ChoiceDefault

DB

DB

DC

DC

6

Insurance vs. Investment

Society prefers a degree of insurance Individuals do better with insurance

• Need wealth cushion (health & long-term care)

Employers should be indifferent• Today, prefer DC (investment) model

because it poses less risk to employer

7

Insurance vs. investment

Markets are indifferent • Unresolved debate: Investment of

retirement assets in risk-free or risky assets

If we move away from the employer as insurance guarantor, will we need new market hedging instruments?

8

Choice vs. default

Caveats: • Any choice system needs strong

defaults to succeed • Role of signaling • Cost of choice

Insurance: anti-selection Investment: poor choices

9

Choice vs. default

Society likes defaults (offer protection)• Cost to society of choice

Individuals want choice but often make poor choices (behavioral economics) • Role of signals

10

Choice vs. default

Employers are indifferent to (individual) choice • But want more employer choice (models)

For markets, choice provides opportunity for innovation• But, for individuals, defaults (standardized

products) increases comparability, decreases costs

11

Conclusions Retirement system need a degree of insurance

• Protect society, individuals Consider (carefully) role of choice

• Choice brings cost • At a minimum, provide strong defaults

Rethink role of employers Market evolution

• Products, investments/hedging mechanisms Echoed by Canadian Provincial Reports …

12

Plan Innovation in Canada

13

• A Fine Balance: OECP PBA and ITA should allow and encourage plan

innovation (not true today)• Size matters• Lower admin costs per unit• Lower investment expense ratios• Can hire in-house investment expertise • Diverse investment opportunities (e.g., private placement)• Pooling of some risks (e.g., mortality)

$10 Billion in assets may be needed May require allowing smaller plans to commingle

assets

14

A Fine Balance: OECP Encourage MEPPs, JSPPs and Jointly

Governed Target Benefit pension plans (JGTBPPs)• automatic participation (but can opt out) • for worker, plan is target DB • for plan sponsor, it is DC • can be used for SEPPs • require a Pension Advisory Committee (with

some retirees) if not jointly governed

15

A Fine Balance: OECP Encourage commingling of small/medium size plans to

achieve the advantages of size Establish an Ontario Pension Agency for stranded

pension assets, or Allow existing MEPPs/JSPPs to service new, less-

connected, members Model = OMERS or TIAA-CREF Management could be private (so long as expenses <

0.5% of assets), or Expand the C/QPP (second tier) or create a comparable

provincial plan

16

Ontario Budget, March 26

Allows OTPP to take on new and different clients (Public plans only)

Consistent with goal of commingling of assets

Already allowed for OMERS

17

Province of Quebec Member-Funded Pension Plans

• A DB target benefit plan• For employer/plan sponsor, it is a DC plan• A commingled asset plan, not individual accounts • Indexation of benefits, both before and after retirement

is contingent on the funding health of the plan• Can be a MEPP or a SEPP• Is meant to be fully funded at all times

18

Promises to Keep: The Nova Scotia Pension Review Panel States that existing rules inhibit innovation Notes that special MEPPs and JSPPs are

DB but with some contingent benefits These plans have joint governance The panel likes this model, says it should

be utilized more broadly

19

Promises to Keep: The Nova Scotia Pension Review Panel Finally suggests a new province-wide plan

• Anyone could join (but voluntary)• Administered by independent provincial

agency• Could accept commuted value of orphans • Could move poorly managed plans in • A DC target benefit plan• Joint trusteeship

20

Getting our Acts TogetherAlberta / British Columbia (ABC) Start a new provincial ABC plan

• Available to anyone• Commingled assets ( = size)• Pooling of risks• Total expenses < 0.4% of plan assets • A DC plan (specified contribution)• But also a target DB plan

21

Getting our Acts TogetherAlberta / British Columbia (ABC) No benefit improvement unless plan

funding is healthy Auto default is “you are in” Can opt out (i.e. not mandatory) Board of Governors decides

investment direction at arm’s length from government)

22

Conclusion

Strongly similar recommendations (at least at big picture level)

Four panels / five provinces Wise and independent Must be content worthy of further

discussion

23

What employers think

24

24

Trade-offsWilling Unwilling

Already Required

Restricting plan improvements for poorly funded plans

85% 7% 8%

Requiring a minimum funding cushion/security margin

65% 27% 9%

Providing immediate vesting for members

35% 40% 25%

Requiring the employer to fully fund any wind-up deficit

31% 41% 28%

Giving plan member representatives greater involvement in plan governance

18% 64% 18%

2009 Watson Wyatt Pension Risk Survey Results from 156 respondents across Canada Copyright © Watson Wyatt Worldwide. All rights reserved.

25

25

Future Plan DesignsYes No Unsure

Not Familiar

Province-wide pension plan on (opt-in or opt-out basis)

43% 25% 23% 8%

Adjusted cost benefit / target benefit plan

34% 13% 27% 27%

Member funded plan 32% 22% 24% 21%

Cash balance plan 20% 32% 34% 14%

Employer managed investment for all DC contributions

13% 70% 14% 2%


26

2626

DC Conversions 2009 2008

No conversion made in the past and not considering any changes in the future

41% 42%

Made conversion more than 2 years ago 26% 21%

Made conversion in the last 24 months 11% 9%

Implementing conversion in next 12 months 3% 3%

Considering changes for the future 29% 33%

- DC for new hires only

9% 15%

- DC for new hires/for all future services

(no changes to past service)8% 5%

- Conversion includes some past service/full plan wind up 12% 14%

26


27

Pension Plan Risk Spectrum

Plan Design DC, ABC, CPP2

Member-funded

Target Benefit, Negotiated

Cost, MEPP

Participating Jointly Sponsored

Single Employer

Governance Employer or Independent

Pension Committee

Joint Joint or Employer

Joint Employer

Contributions Employer

Fixed Fixed Fixed Mostly fixed Variable Variable

Contributions Employee

Fixed Variable None or Fixed Fixed Variable Fixed

Benefits Variable Variable Variable Minimum Fixed (except on wind up)

Fixed

Risk Pooling

Member Shared Employer

Copyright © Watson Wyatt Worldwide. All rights reserved.

28

Measurement Framework

29


Retirement 20/20 tool to assess how well particular designs meet stakeholders needs and risks • Role alignment

Objectively measure new plan designs Also consider effects of “moral hazard”

30


To date reviewed DB, DC, Ontario Teachers Pension Plan, Dutch industry wide schemes, United Methodist Church Plan

More to evaluate Preliminary lessons learned

31

Society’s Needs & Risks (Composite rating )

Plan being evaluated: Ontario Teachers Pension Plan

Criteria Description Rating Evaluation Moral hazard (& adjusted rating)

Sustainable Sustainable across and within generations. Equitable across and within generations.

Can be funded to directly allocate costs back to the current generation of taxpayers/teachers.

Can be tempting to push costs to future generations of teachers/ taxpayers. Fund is facing significant demographic issues with maturation of plan and retirement of baby boom. Provisions that allow teachers to return to work and collect a pension could affect fund stability.

Robust Fair, covers great majority, creates shared economic growth, avoids adverse incentives

Uses markets efficiently through group investing and expertise of well trained advisors. All teachers are covered by equivalent plan, avoiding cost and expense of “competing benefits” in various school districts.

Board control equally by union and government minimizes likelihood of significant increase or decrease in future benefits. Taxpayers may not have same sentiment about cost increases as do government representatives.

Does not promote economic risk

Efficiently allocates resources and encourages labor force participation.

Plan contains generous early retirement features introduced decades earlier to encourage retirement which are now part of the accrued benefit and may not work to retain teachers in the future.

Teachers can return to work and collect full pension.

Green

Yellow

Green

32

Society (composite rating)

Meets society’s needs and risks.

Individual criteria ratings:

Structure of plan, with costs and control equally shared by teachers and government means plan does well meeting social goals without putting future taxpayers at risk. Plan also covers all teachers in Ontario, ensuring continuity of coverage and that individual school boards spend time educating students not running a pension plan.

Ratings after adverse incentives:

Adjusted composite rating:

The strength of the plan is only as good as the governance structure. To date, government and union have been committed to appointing professional trustees. If that commitment changed, and Board appointments were politicized, plan experience could be different. An all-Ontario teachers strike poses significant political risk.

Green

The average rating of -green is a composite of the nine society subcategories.

The ratings for the nine-subcategories are shown in the color line. They are color grouped so you can see how many of each rating were received.

Ratings for each subcategory were adjusted for the effects of moral hazard. This shows the new color line after moral hazard is considered as well as a new composite rating.


33

Summary – Ontario Teachers Pension Plan Criteria Description Rating Evaluation Effect of moral hazard Self-adjusting Automatically adjusts to

changing demographic and economic conditions.

Rating after moral hazard:

Aligns roles with skills

Aligns stakeholders’ roles well with skills.


New norms for work and retirement

Could support flexible work arrangements, e.g. phased retirement, return to work, etc.


Alignment with markets

Could use market mechanisms effectively to hedge risks.


Society (composite rating)

Meets society’s needs and risks.


Ratings after moral hazard:


Individuals (composite rating)

Meets individuals’ needs and risks.




Employers (composite rating)

Meets employers’ needs and risks.




Markets (composite rating)

Meets markets’ needs and risks.




Green

Yellow

Green

Yellow-Green

Green

Green

Yellow-Green

Green

34

Initial lessons learned

Strong governance Aligning roles with skills Self-adjusting mechanisms Member solidarity Independence from employer Use of groups Default free discount rates

35

How does OTPP do? Displays all 7 characteristics, to some degree Strengths: governance, solidarity, independence Weaknesses

• Generous benefit formula (including rule of 85) limits ability to adjust benefits

• High benefit to contribution ratio may push plan to take too many investment risks

• Requires politicians, unions to keep hand out of the cookie jar

36

Next Steps

37

Call for models Call for models

• Out now: Based on the Retirement 20/20 ideas, what

would YOU design as the next tier II retirement system?• Principles outlined in Wharton paper (these slides)• Measurement framework

Seminar(s) (2010), prizes to winners• Hopefully, US and Canadian seminars with results

38

For more information

www.retirement2020.soa.org If you are interested in the call for models

or any of the Retirement 20/20 initiatives • Emily Kessler, [email protected] or 847/706-

3530• Andy Peterson, [email protected] or

847/706-3591

39

Thank you!

Documents

Update: SOA’s Retirement 20/20 Project CIA Annual Meeting, 25-26 June 2009 Speakers: Rob Brown, University of Waterloo Emily Kessler, Society of Actuaries