Short summary Economics & Finance
of Pensions
Lecture 1
How much to consume, Ct = wT /D
working years * wage divided by the years you have left to live
If you are 20 now, pass away at 80 (D=60) and retire at 65 (T=45) 45/60 = 75, consume 75% of
wage and pension contribution 25%
Human wealth (T-t)w if t < T 0 if t > T
Financial wealth at retirement: Tw(D-T)/D = 45w(15) / 60 = 11.25w
The above example to far too simple, risks (longevity, labor, financial markets) and preferences are
ignored
Key elements of new contract: shift from DBDC so variable lifelong annuities instead of guaranteed
benefits.
Trends that require rethinking the design of pension systems:
- Change in labor: flexible labor careers, work at later age
- Increasing life expectancy
- Declining fertility
- Individual heterogeneity causes preferences for choices
- More divorces
Core purposes of pensions from the perspective of the individual
- Consumption smoothing
- Risk pooling (idiosyncratic); long or short life risks
- Risk sharing (systemic); shifts these risks to whom is able to bear it and give risk premium
- Incorporate individual preferences
- Policy purpose: poverty alleviation
Lecture 2
Fully funded: the participant paid for this own pension income during the working years and the
funds are invested in financial markets > investment returns key risk factor. System can be run by
individual, employer, labor unions, commercial parties or government
PAYG: current working pays for current retirees. Demographic (longevity, fertility) and economic
growth (immigration, productivity) key risk drivers. Also poor property rights because income level
and first payment date can be adjusted by political decisions. It is run by the government
Pure DB: sponsor guarantees benefits and bears risks, benefits determined by years worked
adapt to shocks by sponsor and later future participants to maintain balance
design issues: US, Detroit had to cut on building the city because they had to pay some many
pensions. NL: benefits not indexed and even cut if value assets < liabilities
Pure DC: all risks for participant and all decisions taken by participant, there is no risk sharing across
generations and you might outlive your assets. Benefits determined by personal contributions &
investment returns.
,design issues: focused on max. capital at retirement but better to have stable capital > affects
investment strategy & requires sharing of longevity risks
Issues with design of balancing when shocks happen: which adjustment are preferable? (more
contributions, less benefits, late retirement?) / who shares these risks? (employer, participants,
future generation, tax payer) / build into contract in advance or some room for making adjustments?
Argument in favor of choice: individuals best know their own preferences (health, consumption,
when to retire)
Arguments against: 1. people make behavioral mistakes, should this be nudged?
2. If choices are not actuarially fair, offering choice options can be costly for pension provider / other
participants. It would be fair if you get more income if you retire later. It is actuarially fair if the value
of all options is the same.
Options to choose (6): how much to contribute, how to invest in pension wealth, which pension
provider, how much to spend per year once retired, when to retire and partner pension or not?
If you retire later, you should get more lifelong constant income stream (Yi) > we see this by 1+Delta
i pit Y i
Actuarial value for individual i of Y as of age 65: V 65= ∑
( 1+ R )t −65
t =65
i (1+δ s ) pit Y i
Actuarial value for individual i of Y as of age s>65: V s =∑ t−65
t= s ( 1+ R )
s = when you retire, if later than 65 you should get more
Problem: delta is dependent on current interest rates, use standard rate of 6% more per extra year
you work?
Pillar 1: public pensions, main objective: poverty alleviation. This is traditionally DB, but we shift to
DC elements. AOW age linked to life expectancy & in Sweden even linked to tax base to maintain
sustainability of the system.
Pillar 2: worker pensions, maintain the standard of living of the middle class (consumption
smoothing). Actuarially fair because you only receive it when you work.
Funded: promises backed by financial assets, protected against discontinuity of the firm / employer
Pillar 3: private personal pensions, tailor to individual preferences. No redistribution, risk protection
through capital markets and insurance companies. Voluntary, often tax favored.
Lecture 3
Changes that require rethinking of design: heterogeneity in participants & preference to have choices
Replacement ratio target 70%: average income after retirement / average income before retirement
net replacement rate: projected income after retirement as a ratio to the average income during
working life.
The lack of differentiation can cause welfare loss. Options for differentiation:
- Investment strategies: life cycle investing so more risk when young and look at risk tolerance
and risk capacity – what is the risk aversion?
- Partner pension yes or no? does the partner have income his/herself?
, - Level of required contributions dependent on existing pension wealth / housing wealth:
mandatory contributions avoided if one owns a house without a mortgage debt > the house
is the pension here, less costs in retirement for houseowners compared to renters.
Many studies argue in favor of libertarian paternalism: not too many choice options are offered and
trustees nudge individuals to take specific decisions. The nudged choice (default) can be well
dependent on observed / reported individual characteristics
Recall the 6 choice options, Henkens and Van Dalen (2016) considered 5 domains for additional
choice options (the only one not discussed here is when to retire)
They asked 2 questions: 1. Do you value choice? 2. Do you value that other take decisions on your
behalf? Paternalists score low on 1 and high on 2: want to have the option to chose but do not intend
to use it, libertarian paternalists: score high on both and libertarians score high on 1, low on 2.
nudging is appreciated by libertarian paternalists but not by libertarians.
Decision making is driven by 2 cognitive systems: automatic (irrational) and reflective (more rational).
people are overconfident, are loss averse, procrastinate, like status quo and framing works
Individual judgements are driven by: anchoring (compare to somewhat similar) / availability (certain
risks come to your mind directly but are not likely to happen, terrorism for example) &
representativeness (if you already had 3 times 6 with the dice, you think the chance it will happen
again is less than 1/6)
Lecture 4
Intergenerational distribution: between generations like PAYG: the young pay for the old and only if
choices are mandated, intergenerational transfers (solidarity) is possible
Intragenerational: within the same generation.
Backloading of benefits: you pay more at the beginning but eventually if you work there for a long
time you get the value of what you paid for. Advantage of backloading: pin down specific human
capital, administrative, avoids small pots etc.
Examples of backloading:
- Vesting: you only get pension rights after a longer time of working in that company. Creates
incentive to stay longer: pin down specific human capital, avoids ‘small pots’, good for economy
- Final pay vs. career average: good when your career is steep, but unfair for low skilled workers and
creates bad incentives like overtime in last years, promotions etc.
- Uniform premium & uniform accrual (doorsneesystematiek): all workers pay the same contribution
as % of wage, independent of age. Young pay too much, old get subsidized (backloading!) you only
get this subsidy back if you keep on working as an employee.
Disadvantages of backloading:
- Discontinuity risk; firm specific risk, what if the firm you would get benefits from goes
bankrupt? Political risk for PAYG component.
- Unfair for some; if your career is not steep.
- Impact on demand for labor; old workers more expensive and workers are reluctant to start
a second career as ZZP’er because they would lose the subsidy. Distortions in labor market.
The pension and benefit system can have a clear impact on the labor supply, people can be
encouraged to work depending on;
Child care subsidies
of Pensions
Lecture 1
How much to consume, Ct = wT /D
working years * wage divided by the years you have left to live
If you are 20 now, pass away at 80 (D=60) and retire at 65 (T=45) 45/60 = 75, consume 75% of
wage and pension contribution 25%
Human wealth (T-t)w if t < T 0 if t > T
Financial wealth at retirement: Tw(D-T)/D = 45w(15) / 60 = 11.25w
The above example to far too simple, risks (longevity, labor, financial markets) and preferences are
ignored
Key elements of new contract: shift from DBDC so variable lifelong annuities instead of guaranteed
benefits.
Trends that require rethinking the design of pension systems:
- Change in labor: flexible labor careers, work at later age
- Increasing life expectancy
- Declining fertility
- Individual heterogeneity causes preferences for choices
- More divorces
Core purposes of pensions from the perspective of the individual
- Consumption smoothing
- Risk pooling (idiosyncratic); long or short life risks
- Risk sharing (systemic); shifts these risks to whom is able to bear it and give risk premium
- Incorporate individual preferences
- Policy purpose: poverty alleviation
Lecture 2
Fully funded: the participant paid for this own pension income during the working years and the
funds are invested in financial markets > investment returns key risk factor. System can be run by
individual, employer, labor unions, commercial parties or government
PAYG: current working pays for current retirees. Demographic (longevity, fertility) and economic
growth (immigration, productivity) key risk drivers. Also poor property rights because income level
and first payment date can be adjusted by political decisions. It is run by the government
Pure DB: sponsor guarantees benefits and bears risks, benefits determined by years worked
adapt to shocks by sponsor and later future participants to maintain balance
design issues: US, Detroit had to cut on building the city because they had to pay some many
pensions. NL: benefits not indexed and even cut if value assets < liabilities
Pure DC: all risks for participant and all decisions taken by participant, there is no risk sharing across
generations and you might outlive your assets. Benefits determined by personal contributions &
investment returns.
,design issues: focused on max. capital at retirement but better to have stable capital > affects
investment strategy & requires sharing of longevity risks
Issues with design of balancing when shocks happen: which adjustment are preferable? (more
contributions, less benefits, late retirement?) / who shares these risks? (employer, participants,
future generation, tax payer) / build into contract in advance or some room for making adjustments?
Argument in favor of choice: individuals best know their own preferences (health, consumption,
when to retire)
Arguments against: 1. people make behavioral mistakes, should this be nudged?
2. If choices are not actuarially fair, offering choice options can be costly for pension provider / other
participants. It would be fair if you get more income if you retire later. It is actuarially fair if the value
of all options is the same.
Options to choose (6): how much to contribute, how to invest in pension wealth, which pension
provider, how much to spend per year once retired, when to retire and partner pension or not?
If you retire later, you should get more lifelong constant income stream (Yi) > we see this by 1+Delta
i pit Y i
Actuarial value for individual i of Y as of age 65: V 65= ∑
( 1+ R )t −65
t =65
i (1+δ s ) pit Y i
Actuarial value for individual i of Y as of age s>65: V s =∑ t−65
t= s ( 1+ R )
s = when you retire, if later than 65 you should get more
Problem: delta is dependent on current interest rates, use standard rate of 6% more per extra year
you work?
Pillar 1: public pensions, main objective: poverty alleviation. This is traditionally DB, but we shift to
DC elements. AOW age linked to life expectancy & in Sweden even linked to tax base to maintain
sustainability of the system.
Pillar 2: worker pensions, maintain the standard of living of the middle class (consumption
smoothing). Actuarially fair because you only receive it when you work.
Funded: promises backed by financial assets, protected against discontinuity of the firm / employer
Pillar 3: private personal pensions, tailor to individual preferences. No redistribution, risk protection
through capital markets and insurance companies. Voluntary, often tax favored.
Lecture 3
Changes that require rethinking of design: heterogeneity in participants & preference to have choices
Replacement ratio target 70%: average income after retirement / average income before retirement
net replacement rate: projected income after retirement as a ratio to the average income during
working life.
The lack of differentiation can cause welfare loss. Options for differentiation:
- Investment strategies: life cycle investing so more risk when young and look at risk tolerance
and risk capacity – what is the risk aversion?
- Partner pension yes or no? does the partner have income his/herself?
, - Level of required contributions dependent on existing pension wealth / housing wealth:
mandatory contributions avoided if one owns a house without a mortgage debt > the house
is the pension here, less costs in retirement for houseowners compared to renters.
Many studies argue in favor of libertarian paternalism: not too many choice options are offered and
trustees nudge individuals to take specific decisions. The nudged choice (default) can be well
dependent on observed / reported individual characteristics
Recall the 6 choice options, Henkens and Van Dalen (2016) considered 5 domains for additional
choice options (the only one not discussed here is when to retire)
They asked 2 questions: 1. Do you value choice? 2. Do you value that other take decisions on your
behalf? Paternalists score low on 1 and high on 2: want to have the option to chose but do not intend
to use it, libertarian paternalists: score high on both and libertarians score high on 1, low on 2.
nudging is appreciated by libertarian paternalists but not by libertarians.
Decision making is driven by 2 cognitive systems: automatic (irrational) and reflective (more rational).
people are overconfident, are loss averse, procrastinate, like status quo and framing works
Individual judgements are driven by: anchoring (compare to somewhat similar) / availability (certain
risks come to your mind directly but are not likely to happen, terrorism for example) &
representativeness (if you already had 3 times 6 with the dice, you think the chance it will happen
again is less than 1/6)
Lecture 4
Intergenerational distribution: between generations like PAYG: the young pay for the old and only if
choices are mandated, intergenerational transfers (solidarity) is possible
Intragenerational: within the same generation.
Backloading of benefits: you pay more at the beginning but eventually if you work there for a long
time you get the value of what you paid for. Advantage of backloading: pin down specific human
capital, administrative, avoids small pots etc.
Examples of backloading:
- Vesting: you only get pension rights after a longer time of working in that company. Creates
incentive to stay longer: pin down specific human capital, avoids ‘small pots’, good for economy
- Final pay vs. career average: good when your career is steep, but unfair for low skilled workers and
creates bad incentives like overtime in last years, promotions etc.
- Uniform premium & uniform accrual (doorsneesystematiek): all workers pay the same contribution
as % of wage, independent of age. Young pay too much, old get subsidized (backloading!) you only
get this subsidy back if you keep on working as an employee.
Disadvantages of backloading:
- Discontinuity risk; firm specific risk, what if the firm you would get benefits from goes
bankrupt? Political risk for PAYG component.
- Unfair for some; if your career is not steep.
- Impact on demand for labor; old workers more expensive and workers are reluctant to start
a second career as ZZP’er because they would lose the subsidy. Distortions in labor market.
The pension and benefit system can have a clear impact on the labor supply, people can be
encouraged to work depending on;
Child care subsidies
