The Macro-Economics
of Pensions
Adair Turner
September 2nd, 2003
Pensioner Income as a % of GDP
Today
2030
If we assume
13.5
• Average retirement
age unchanged
9.3
Private
3.8
State
5.5
8.0
+4.5
5.5
1
• Average pensioner
income a constant
proportion of
average income
Pension Claims Against Future Output –
Three Current Models
PAYG STATE
SYSTEM
Nature of Claim
 Against totality of GDP
 Enforced via
contributions/taxes
Risks
 Political
 Indirectly – Demographic and
Macro-economic
FUNDED DC
 Against profits
 Volatility of:
o Profit slice of GDP
o Market price of equities
and bonds
 Via ownership of
equities and bonds
FUNDED DB
 Against profits
 Via ownership of
equities and bonds in the
fund
 But also via bond-like
promise of the
individual company
2
 Volatility of:
o Profit slice of GDP
o Bond and equity prices
 Bankruptcy of the individual
firm
Two Near Equivalences
Compulsory DC scheme
invested in Government
index linked bonds
and
Tax financed PAYG
DC scheme invested in
corporate bonds
If the rate of return on indexlinked government bonds is
equal to rate of growth.
Still different in respect to
and
DB average salary scheme
•Life expectancy risks
•Diversification of bond
portfolio
But similar since DB schemes
are bond-like liabilities
3
Key Question: How Best To Structure Pension Claims?
Against Future GDP
Against Future Profit
Slice of GDP
PAYG
DC in equities
DB
Funded in government
bonds
DC in corporate bonds
4
Two Possible Routes To Increased Pensioner Income
• Investment and future output unchanged, but
some other group of savers relinquishes a
matching claim on future assets and profits
• Savings and investments increase, and as a result,
so does future output
5
Real Return On Large Stocks: US 1926 - 2000
No. of 20 Year Periods with Return in Range (57 Periods)
8
Median 7.9
7
Mean 7.2
7
7
5
4
3
3
3
2
2
2
3
1
0-1
1-2
2-3
____________________
Source: Ibbotson Associates: 2002 Yearbook
3-4
4-5
5-6
6-7
7-8
8-9
9-10 10-11 11-12 12-13 13-14
% Cumulative Average Return Over 20 Year Periods
6
Real Return On Large Stocks: US 1926 - 2000
Mean 7.64
Median 8.55
No. of 5 Year Periods with Return in Range (72 Periods)
16
10
8
4
4
5
6
5
6
3
3
1
1
-12
-9
-6
____________________
Source: Ibbotson Associates: 2002 Yearbook
-3
0
3
9
6
12
15
18
21
24
0
27
% Cumulative Average Return Over 5 Year Periods
7
Real Returns on Large Company Stocks: US 1926 – 2001
Median 7.04
10
Mean 6.86
No. of 30 Year Periods with Return in Range (47 Periods)
9
8
7
6
4
3
2-3
3-4
4-5
____________________
Source: Ibbotson Associates: 2002 Yearbook
5-6
6-7
7-8
8-9
9-10
10-11
11-12
12-13
% Cumulative Average Return Over 30 Year Periods
8
Real Returns on Large Company Stocks: US 1926 – 2001
9
Median 6.99
No. of 50 Year Periods with Return in Range (27 Periods)
7
4
4
2
1
2-3
3-4
____________________
Source: Ibbotson Associates: 2002 Yearbook
4-5
5-6
7-8
6-7
8-9
9-10
10-11
11-12
% Cumulative Average Return Over 50 Year Periods
9
Claims And Risks
PAYG
OR FUNDED IN
Claim Against
Inherent Risks
 Totality of future GDP
Wages and profits
 Unanticipated
variations in future
GDP
 Profit slice of
future GDP
 Unanticipated
variations in future
profits
—
GOVERNMENT BONDS
FUNDED IN CORPORATE
CAPITAL
10
Dividend Valuation Model: Inherent Uncertainty
PV = Rational value of equities
D0 = Current year dividends
r = risk free rate
D0 (1 + g)
PV = 1 + r + p
PV =
+
p = equity risk premium
g = rate of growth
D0 (1 + g)2
(1 + r +
D0
r+p-g
11
p)2
+
D0 (1 + g)3
(1 + r + p)3
Risk Management And Risk Sharing In Funded Schemes
1. DC schemes – Managing the crystallisation risk
2. The DB to DC Shift
 Is it concerning or desirable?
 Is it inevitable?
3. Should DB schemes invest in equity?
12
Risks In Final Salary DB Schemes
CATEGORY OF
RISK
ABSORBED
ABSORBED
BY
EMPLOYER
BY
EMPLOYEE
CONSEQUENCE
Investment
return

x
Changes in life
expectancy

x
No incentives for
later retirement
Individual salary
progression

x
Complex accrued
rights and
transfer values –
to disbenefit of
early leavers
13
Overall
risk borne
by
employer
is high
DB / DC Hybrids –Risk Sharing Alternatives
CATEGORY OF
RISK
ABSORBED BY
EMPLOYER
ABSORBED BY
EMPLOYEE

x
Changes in life
expectancy
x

Pensionable ages
increasing over time in
line with life
expectancy estimates
Individual
salary
progression
x

Average salary based
not final
Investment
return risk
14
Who Killed DB?
• Unanticipated increases in life expectancy
 Slow response to 1980s and 1990s mortality declines
• Increased life expectancy not matched by later retirement ages
or increased contributions
• Irrational exuberance, apparent “surpluses” and contribution
holidays
• Tax and residual ownership disincentives to very large
surpluses
• “Costless” tax increase of 1997
15
How To Increase Resources For Future Pensioners
With aggregate
savings,
investment
and output
unchanged
OPTION 1: Future pensioners claim
higher share of future
capital and profits ….
At expense of other
savers
OPTION 2: Profits rise as % of GDP… At expense of future
workers
Via increased
savings,
investment and
output
OPTION 3: Domestic investment rises
and thus GDP (and GNP)
OPTION 4: Overseas investment rises
and thus GNP (but not
GDP)
16
Option 1: Pension Saving Up, Some Other Saving Down
Either
• Increased pension saving, but offsetting dissaving by the
same people
• Increased pension saving but offsetting dissaving by
“others”
17
1963
1964
1965
1966
1967
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2
2002 001
(est)
Percent
UK Pension Fund Assets As a % Of GDP
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
____________________
Note: Occupational schemes only
18
% of UK Equities owned by individuals &
UK pension funds 1963-93
70
60
Percent
50
40
30
20
10
0
1963
1969
1975
1981
1989
1990
1991
1992
UK Pension funds
____________________
Note: After 1993 the % owned by individuals continued to fall, but so too did the % owned by UK pension funds. But the growing category in 1993-2002 was
overseas investors (1993: 16%, 2002: 32%), many of whom are pension funds; and conversely UK pension funds have invested more overseas. The overall pension
fund role in the market has therefore probably continued to grow
19
1993
Individuals
Capital Investment And The Marginal Product
Of Capital
Savings Schedule 1
Savings Schedule 2
Rate of
return
Quantity of capital invested
20
Fishing Boats, Fish And A Falling Population
Steady State
Demographic Change
•1,000 population
•500 per generation
•500 working, 500 retired
•Workers work
Half time fishing
Half time building a boat
•Boats wear out over one working life
• 500 generation of retirees
followed by 250
generation of workers
• Retirees would like to sell
500 boats they have built
• Workers only need to buy
250 boats
Relative price: 1 boat = ½ of catch for
one working life
21
Price of boats, relative to
fish, will fall
Funded Pensions in Corporate Capital: Two Dimensions
Claim on future
Cash Profits
Inter-Generation Funds
Flow
• Selling price of
accumulated assets
must equal the
discounted value of
future cash profits
thereafter
• Value of Generation 1’s
asset decumulation must
equal Generation 2’s
asset accumulation
• Future pensioners
consumption is financed
from future workers’
consumption deferral
• Funded pensions are
a claim on future
profits
22
Rising Longevity, Constant Fertility
Same trade-off in
each generation
 No change in retirement age
Different trade-off
• G 1 saves more
• G 1 saves more
• Capital stock increases,
returns decline
• Capital stock increases,
returns decline
• Asset prices unchanged since
• Asset prices fall since
 G 2 chooses later retirement
 G 2 has same target capital
stock as G 1
 Required returns fall in line
with actual
23
 G 2 has lower target capital
stock than G 1
 G 2’s required returns have
not fallen (relative to the no
rise in longevity base case)
Falling Fertility, Constant Longevity
• Generation 1 keeps saving as before (since longevity
unchanged)
• In Generation 2, unchanged capital stock but less labour
 Lower returns to capital
 Higher real wages
• Value of assets falls because
 Returns down, discount rate unchanged
 G 2 has lower target capital accumulation than G 1
24
Demographic Impacts on Returns to Capital: Model
Results
Garry Young:
Baby-boom generation
Increased longevity
Falling fertility
David Miles:
Given future actual trends in UK
demographics, returns fall:
• 4.56% (1990) to 4.22% (2030)
• 4.56% (1990) to 3.97% (2060) if PAYG
phased out
25
-0.1%
-0.1%
-0.3%
Financial Asset Prices – Transitional Effects
Increase in
Aggregate Savings
Rate
Fall in Anticipated
Future Rates of
Return
Long-Term Impact
– Higher Capital Stock
– Lower Long-Term Returns
Transitional Impact
– Increased demand for initially sticky supply
– Rising financial asset prices
Long-Term Impact
– Fall in long-term equilibrium returns
– e.g. 4.6%  4.2%
Transitional Impact
– From PV=
4.6
r+p
4.2
– To PV= r+p
– Fall of 10% if r and p unchanged
26
Overseas Investment as the Solution –
Conditions Required
Other countries to invest in which have
Increasing numbers of
productive and prosperous
future workers able to:
•Different demographics –
dependency ratios not rising
• Provide future
output available for
consumption
• Buy capital assets
from decumulating
savers in the
investor countries
•Economic success – growing
per capita GDP
+
27
Small numbers of domestic
decumulating pensioners
also trying to sell assets
Demographic Change in UK and China –
UN Medium Variant
% Population by Age Band
UK
China
2000
2050
21
30
60
54
19
16
10
30
65
54
25
16
60+
15-59
28
0-15 Years
26
The Fundamental Choice
Some mix of:
Increasing
longevity
plus falling
fertility
• Future pensioners poorer relative to
average income than today
• Future workers must give up more
consumption in savings,
contributions or taxes
• Retirement ages must rise
29