The economics and finance of
pensions
Lecture 1
The book by Barr & Diamond is discussed in L1 & L2
The solution how much to consume, Ct = wT/D
T are the working years, D are the years you have to live. If you are 20 now and you will pass away at
80 and retire at 65: Ct = 45/60 consume 75% of wage, pension contribution 25%
Human wealth (T-t)w if t>T, this is zero when t>T
Financial wealth at retirement: Tw(D-T)/D =[45*w*15]/60 = 11.25w)
The example on the optimal contribution rate is far too simple, interest rates set at non-zero here.
Risks that are ignored:
- Longevity risk: you don’t know when you die, maybe you live very long
- Labor market risk: fulltime, wage?
- Risk in financial markets; risky investment opportunities available & potentially attractive?
Individual preferences and choices are ignored
- Retirement date does not have to be set in advance
- Partners, children
- Agents are irrational (few self-employed people pay adequate pension contributions)
Increase in retirement age in countries: actual retirement ages differ substantially
Current policy debates; how to set up a sustainable pension system given the demographics? UK no
longer mandatory insurance against longevity risk; free to consume pension capital as of age of 55
Dutch reform pension
- Age from 65 to 67 because life expectancy increases. It increases slower, 1 year longer life
expectancy means 8 months (2/3) later retirement age
- Phase out uniform contribution & uniform accrual system
- More transparent & personal
- Option to take partial lumpsum at retirement date (10%)
- Payment will be cut if there is too little money in the pot heavily discussed!
- Negative interest rates… is the funded system still attractive?
Key elements of the “new contract”
Take more risk when you’re young; life cycle investing of personal accounts
Limited choice options; collective decision making
Mandatory participation (to sector for firm, for employee)
Pension income lifelong (variable annuities rather than guaranteed benefits) apart from 10%
lumpsum (groot bedrag in 1x) option.
Collective assets with personal account as well as ‘solidarity reserve’ up to 10% of annual
contribution can be allocated to other through solidarity reserve
1
,The new contract should end the ongoing fights on level of funded rate and on adequate discount
rates. There are still many steps to be taken…! (2020-2026)
Chapter 1
Trends that require rethinking the design of pension systems
- Increasing life expectancy & declining mortality; unless retirement dates or contribution
levels are adjusted, systems become unsustainable
- Decline in fertility; the group that needs to pay the next bill way smaller…
- We start working at a later age
- Heterogeneity in participants; individuals can differ widely in characteristics (own house,
partner)
- Flexible labor careers; some self-employed and we don’t work 40 hours per week for the
same company our whole life. Pension products where one pays too little in some years and
too much in others under pressure – this is not fair
- Less stable relationships (more divorces)
- Preference for choice options: when to retire, in which stocks to invest, how much to
contribute
- Mature pension funds; companies do not guarantee income for their workers no more
Core purposes of pensions individual perspective: Chapter 2 Barr & Diamond
1. Consumption is smoothing over time (“Piggybank function”):
accumulation phase, saving – transform human capital into pension capital.
decumulation phase, dissaving – transform pension capital into consumption.
2. Provide insurance idiosyncratic shocks (risk pooling): long life in the unproductive phase and
short life/deprecation of human capital in the productive phase.
- Old age pensions are life-long payments; protect against longer than expected retired life
- Survivor benefits: protect against short life in active phase
- Disability insurance & pensions insure risk of too rapid depreciation of human capital.
3. Risk pooling vs desirable risk sharing/trading: we can only pool idiosyncratic risks.
Risk sharing (macro) = not idiosyncratic, this hits everyone. For example: better health care
means longer life expectancy for everyone. Or the stock market goes up, everybody profits.
Macro risks, shift risks to whom is able to bear this risk and give them a reward like equity
premium. Pension systems can potentially help to share risks where markets are incomplete
and/or distortions are sizeable.
4. Incorporate individual preferences and circumstances adequately
Individuals best know their own preferences and characteristics (e.g. retirement timing,
consumption plans, health status, external assets (house)): argument in favor of choice
Individuals often make ‘behavioral mistakes’, should one “nudge” them? E.g. not saving
enough, not insuring longevity risk, overconfidence in investment decisions
If choices are not actuarily fair (i.e. require same capital) offering choice options could be
costly for the pension provider or other participants
Public policy purposes of pensions
- Social purposes (Robin Hood): giving money from 1 person to another.
o Poverty relief (old people are less likely to be poor, because you receive a pension)
o Redistribute income and wealth across / within generations
o Sources of heterogeneity: age, number of children, health needs, families
2
, o Is this optimally achieved in a pension system or through the government by taxes
and subsidies?
- Other objectives
o Well-functioning labor markets: (adequate pension provision makes it easier to retire
timely or part-time; vesting stimulates investment in company specific human
capital)
o Pension arrangements can stimulate savings and thereby economic growth and
development of capital markets
Polls on social purposes and solidarity
- we have a couple, both young but one passes away. What should be pension fund do?
70% for limited amount of years (5 years) - I think this is best but the system now gives you
70% income lifelong.
- Should self-employed workers be forced to insure for income risks like disability? We think
so, but many ZZP’ers don’t want this.
- NL: AOW live in the country for 50 years and you don’t have to contribute anything.
Public policy and pensions: relevant in case of imperfect people and imperfect markets
- Limits of man; decision making under uncertainty, lack of info: information processing
capabilities & moral limitations
- Imperfect and incomplete markets
o Financial markets are incomplete (macro risks) > these macro risks are untraded,
inflation, wage, longevity risk are examples
o Insurance markets fail (micro risk); adverse selection & moral hazard
Main controversies about pensions
- Individual behavior & quality of institutions; how rational do people deal with time and risk
and how heterogeneous are preferences & circumstances? trade off individual ‘failures’
vs. tailor-made decisions
- Do people trade freely on complete markets or do they need help from institutions to
complete markets?
- Social preferences; how much inequality aversion and redistribution?
Different positions on these controversies as well as history of the pension system explain the wide variety of pension
systems across and within countries. Different emphasis over time drive pension reforms (including Dutch reform).
Lecture 2a
Chapter 3 Barr & Diamond – basic features of pension systems
Key design issue DB/DC: whether benefits/contributions are dependent on risk factors such as (in
funded) investment returns, interest rates, inflation or changes in life expectancy and (in PAYG):
fertility, immigration and changes in productivity.
Pension systems differ around the world, key dimensions on which actual pension schemes differ:
(1) PAYG (Pay as you go) vs. funded system
(2) Risk sharing among who? DB / DC?
(3) Mandatory vs. voluntary; role of government
(4) Actuarially fair vs. (intra- intergenerational) transfers
3
, Funded vs. PAYG (1)
Fully funded: assets of individual/fund always exactly enough to pay accrued benefits. The
participant has paid for his own pension income during working years.
The funds are often invested in risky assets in financial markets. Investment returns are key risk
factor. The system can be run by individual, by employer, by employer and labor unions jointly
(Dutch pension funds), by commercial parties (e.g. exchange listed insurers) or by government
Pay as you go: current working generations pay for current retirees. Think about AOW, you now pay
for your grandparents. Projected same income level for all (but sometimes depends on past income
& contributions)
The first generation gets ‘free gift’ – they never paid for this. The working generation at that time had
to pay for the retirees.
- Run by government;
- Projected income level and first payment date can be adjusted by political decisions (poor
individual property rights)
- Demographic risk (longevity; fertility) and economic growth key risk driver
- Notional DC (Sweden, Eastern Europe): PAYG component where benefit linked in advance to
these risk factors; risks not with future tax payers; risk factors that are not traded on financial
markets can be incorporated
Some countries implement PAYG and some are funded heterogeneity
Risk sharing among who: DB, DC? (2)
Pure DB: sponsor guarantees benefits & bears risks. Benefits determined by number of years
worked, wage, parttime etc. Example: 40years of accrual of 1.75% lifelong of final pay 70%
Pure DC: All risks for the participants, all decisions taken by the individuals. You might outlive your
assets when you live longer. (no risk sharing across generations). Purely individual accounts, no
redistribution or risk sharing within/across generations. The benefits are determined by personal
contributions and investment returns.
Pension systems need mechanisms to adapt to unexpected shocks (investment returns, inflation, size
of labor force, life expectancy) otherwise they’ll become unsustainable.
- In pure DB: recovery contributions of injected funds (by sponsor; later: future participants) to
maintain the balance
- In pure DC: benefits are adjusted to maintain balance
- In reality pension funds are not purely DC or DB, in many systems both contributions and
benefits will be adjusted.
Issues with the design of balancing mechanisms
- Which adjustments preferable (contributions, benefits, both, retirements date etc.)
- Who shares in the risks (employer, current participants, future participants, tax payer)
- Built into the contract or into legislation in advance: write it all down in advance or leave
some freedom to a board of trustees to make adjustments when life expectancy increases /
returns were good or bad (rules vs. discretion debate)
Design issues DB in real life (sponsor guarantees benefits & bears risks) – never pure DB
4
