Living on a Risk Budget

1
Living on a (Risk) Budget
2005
Leo de Bever Executive Vice President MFC Global
Investment Management August 25, 2005
2
Main Ideas

• A pension replacing 50 of final income is 2 x as
  expensive as it was in 1950 because we live
  longer
• Taking more investment risk means higher risk of
  under-funding over long periods, making pensions
  less secure
• Asset-liability management tries to earn the best
  long-term return while controlling the risk of
  under-funding
• Pension governance is focused on short run, makes
  risk management ineffective, and adds to risk of
  under-funding

1
3
Why did DB Plans Become Under-funded in 1990s?
Funding Ratio
What Actuarial Funding Ratio Would Have Been If
No Plan Changes Had Been Made
Actuarial Funding ratio With asset smoothing,
benefit improvements, Contribution reductions
Actuarial Funding Ratio After Benefit
Improvements and Contribution Reductions
100Funding
Actuarial Real Liability Discount Rate
84
4 4 4 4 4
4 4 4.3 4.1 4.4 4.4
4.0 3.9 3. 1

Source Ontario Teachers Pension Plan annual
reports Assets at market liabilities discounted
at actuarial rate contribution rate 16

2
4
Annual Cost of a CPI Indexed Pension as of
Final Salary

• Real Pension as of final wage
• A 50 pension is twice as expensive as a 25
  pension
• x (years of benefits)/(years of saving for
  benefits)
• Being retired 20 yrs costs 2x as much as being
  retired 10 yrs
• Saving 20 years requires 2x the annual savings
  compared to saving 40 years
• x (final wage)/(average wage)
• If real wages grow 2/ year for 40 yrs, final
  income is 1.45 x average wage
• x .60 for every 2 of realized real return
• The higher the return, the larger the share of
  the pension that comes from investment income
• x .99 for every year retired that pension is not
  indexed
• A 30 year non-indexed pension is 30 less
  valuable than an indexed pension
• Areas in red have driven the cost of pensions
  in last 50 years

3
5
40 yrs Work, 10yrs Retired, 12.5 Savings _at_ 0
Real Return
Robinson Crusoe World Without Capital Markets
To draw 50 x 10 years, Robinson has to save
12.5 x 40 years
4
6
40 yrs Work, 20yrs Retired, 25.0 Savings _at_ 0
Real Return
Robinson Crusoe World Without Capital Markets
To draw 50 x 20 years, he has to double his
savings to 25 x 40 years
5
7
40 yrs Work, 20yrs Retired, 36.2 Savings _at_ 0
Real Return
No Capital Markets, 2 Real Wage Growth
Retiring on 50 of final salary requires 36 of
savings instead of 25
6
8
40 yrs Work, 20yrs Retired, 20.6 Savings _at_ 2
Real Return
Earning a 2 real return cuts contribution costs
from 36.2 to 20.6
7
9
40 yrs Work, 20yrs Retired, 11.5 Savings _at_ 4
Real Return
Earning a 4 real return cuts to 11.5, but is it
possible to get this without taking risk?
8
10
40 yrs Work, 20yrs Retired, 6.2 Savings _at_ 6
Real Return
6 real return is historically only possible with
100 equities
9
11
30 yrs Work, 30yrs Retired, 12.6 Savings _at_ 6
Real Return
About half of the capital comes from investment
earnings after the member is retired
10
12
30 yrs Work, 30yrs Retired, 9.8 Savings _at_ 6
Real
Pension Not Indexed to Inflation
Long retirement inflation erodes half of real
value over 30 years
11
13
Rising Pension Costs Masked by Taking More Risk
Impact of longevity has been masked by shifting
to risky asset classes and unsustainably high
return on taking market risk in 1990s
1 Robinson Crusoe Economy savings accumulate
but earn no capital market return 2 Marginal
cost cost of funding final income pension in 1
economy 3 Contributions compound at a real
return of 2 (100 year return on bonds) 4
Contributions compound at a real return of 4
(50-50 stock-bond real returns over last 100
years 5 Contributions compound at a real return
of 6 (real equity returns over last 100
years) 6 Pension not indexed erodes at 2.5
(CPI inflation) contributions drop by .01 x
(years of retirement)
12
14
Living on a Risk Budget

• Funding with risky assets lowers pension security
• Risky returns do not become less so over longer
  horizons
• Financial regulators typically insist on a
  funding ratio that increases above 100 with
  asset/liability risk
• e.g. Canadian insurance firms hold 25 reserves
  against equities backing liabilities
• DB Pensions are managed at 100 asset/liability
  ratio
• Sponsors assume any amount above 100 can be
  withdrawn or used to increase benefits
• Creates a ratchet effect bias to under-funding
• Living on a risk budget requires the discipline
  to maintain a reserve against sustained periods
  of negative return on risk
• Sponsors now have reduced capacity to accept
  deficiency risk

13
15
Funding Pensions With 50 Stocks Works in Long Run
Starting at 130 in 1927, and without withdrawing
funds or increasing benefits, The funding ratio
oscillates (wildly) around 100 in the long run
14
16
Asset/Liability Mgt Incorporates the Joseph
Effect
Fat years make up for lean years
15
17
Actuarial Valuation Assumes Every Year is
Another Year
Actuarial valuations assume that at any time
Investment gains are as certain as a checking
account balance Assets in excess of liabilities
are surplus
Creates a long-term bias to being under funded
16
18
30 Year Scenarios with 50 Stocks Wide
Dispersion of Outcomes
17
No change in contributions or benefits
19
With Negative Bias When Surplus gt 110 is
Withdrawn
18
No change in contributions or benefits
20
Sponsors Have Lower Capacity To Make Up
Deficiencies

• DB plans have grown faster than sponsor balance
  sheet, e.g. for Ontario Teachers Pension Plan,
  since 1990
• Liabilities have quadrupled, assets are falling
  short 20
• Ontario Government revenue base has only doubled
  
• Risk has grown faster than pension liabilities
• Most pension plans increased exposure to equities
• OTPP had 0 equities in 1990, as much as 70 in
  1998, 50 now
• Incremental risk from pensions therefore has
  more than doubled in both private and public
  plans
• DB pension plans can add as much as 70 to risk
  of sponsor operations
• Capacity of employees to shoulder 50 of
  deficiency risk is debatable

19
21
DB Plans Will Require Better Risk Sharing

• Make part of benefit conditional on funding
  status
• Hard to do, particularly once members have
  retired
• The only real substitute for the sponsor
  guarantee
• Takes the Defined out of Defined Benefit
• Fund conservatively, invest aggressively
• Higher contributions now
• If markets deliver a lot of extra return, some
  of that can be amortized in the form of lower
  contributions
• Manage 50 Stocks to 110-120 funding ratio
• Amortize gains above that over many years to
  retain flexibility

20
22
Contributing Factors To Poor DB Pension Decisions

• Demographics Pensions have at least
  doubled in cost
• Employers Try to cut short-term
  pension costs Worry about long term
  deficiencies later
• Employees Do not accept that too
  good a pension deal can destroy a
  company
• Actuaries Discount liabilities at risky
  expected return
• Full funding deemed to be 100, no
  reserves
• Courts Asymmetric justice surpluses go
  to members, Deficiencies belong to
  Plan Sponsor
• Income Tax Act Sets maximum funding ratio
  limit Financial
  regulators typically insist on minimum

21
23
Have Lessons Been Learned?

• A pension plan has 85 billion in assets at
  market value
• The risk-free rate of return on the asset that
  matches the structure of the liabilities is
  CPI2
• Liabilities will be
• 85 billion at risk free return 2
  (funding ratio 100)
• 100 billion at risk free return 1
  (funding ratio 85)
• 110 billion at risk free return 0.5 (funding
  ratio 76)
• Pension contributions for the employer and the
  employees will have to rise dramatically in cases
  2 and 3
• Can this fund overcome our tendency to defer
  recognizing a loss if there is even the faintest
  chance of avoiding it?

22
24
Conclusion

• Living longer makes pensions inherently more
  expensive
• Pensions are long-term obligations and must be
  managed using long-term asset-liability
  risk/return mgt principles
• Short-term focus of regulation, accounting, and
  governance, impose conflicting short-term
  objectives
• Living within a risk budget will require some
  combination of saving more, working longer, and
  better risk sharing

23
25
2005
Leo de Bever Executive Vice President MFC Global
Investment Management August 25, 2005