ARCH 2013.1 Proceedings August 1- 4, 2012 Louis Adam Actuarial - - PDF document

SLIDE 1

Article from:

ARCH 2013.1 Proceedings

August 1- 4, 2012

Louis Adam

SLIDE 2

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 1

Mortality Improvement for Canadian Pensioners: P d P j ti S l Proposed Projection Scales

Louis Adam, FSA, FCIA Université Laval, Quebec City, QC 47th Actuarial Research Conference Winnipeg, Manitoba, 2012‐08‐03

Outline

• 1. Introduction
• 2. Phases of Study
• 3. Methodology & Results, Phase II
• 4. Mortality Improvement Rate: Formulas
• 5. Projection Scales: Proposal
• 6. Conclusion

2 ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales

• 1. Introduction

Acknowledgements:

– Financial support from Canadian Institute of Actuaries, and formerly from Chaire d’actuariat, Université Laval and SOA – Data and support:

• Office of the Chief Actuary (CPP)
• Régie des rentes du Québec (QPP)

– CIA CCPME: Committee on Canadian Pensioners Mortality Experience – and many reviewers, colleagues, students

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 3

CCPME

• Created in 2008 by CIA
• Commissioned two studies in 2009
• Data collected in 2010
• Data analysis and review 2010 and 2011
• Reports drafted and reviewed:

– Registered Pension Plan Study in 2012 – CPM Study in 2011 and 2012

• CIA Annual Meeting: June 2012 Presentation

4 ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales

SLIDE 3

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 2

CPM

• Canadian Pensioners Mortality = CPM
• Pensioner data only, from CPP and QPP

administrators

– Separate and combined results Separate and combined results – Data comprehensive and high quality – Data segmented by pension income level – Almost 8 million exposed lives (86.9 M life‐years exposure from 1967 to 2008) – pensions payable since 1967

5 ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales

Messages, CIA June 2012 Session

• 1. It is time to change the Canadian standard for

mortality tables for pension plans: Canadian evidence for this 2 Income is important for pensioner mortality

• 2. Income is important for pensioner mortality

(with time, gender, age, source)

• 3. Mortality trend: Higher improvements rates
• bserved in recent past= troubling news
• 4. Impact might be material for many plans

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 6

• 2. CPM Study Phases

1: Get high quality data: CPP + QPP = CAN 2: Measure qx at recent point in time:

– 2005‐2007, centered in 2006 Ph II R M 31 t 2012 – Phase II Report: May 31st, 2012

3: Measure mortality trends: projection scales

– With recent experience over 15 years for short term scale: 1992‐2007 2006 to 2021 – Long term scale based on C/QPP Actuarial Reports – Phase III Report: Draft July 17th , 2012

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 7

• 3. Methodology & Results, Phase II
• Deaths & Exposure measured
• 5 variables: source, gender, age, income, year
• Exact age, constant force of mortality for

f i l fractional ages

• Exact age compares to “Nearest Birthday”
• Provides point estimate and confidence

intervals

• Graduation: Gompertz, modified at extreme

ages, values within bounds of 1 std dev.

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 8

SLIDE 4

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 3

Income

• 5 income classes
• Split in % of C/QPP Maximum Pension

– 1: <35%, 2: 35%‐94%, 3: >95%

• Remove lower pensions (Class 1) to get proxy

for mortality of pension plans members

• Class 4 = Class 2 (mid) + Class 3 (high)
• Class 5 = All income

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 9

Results shown here

• Ratios of q(x) : CPM‐CAN/ UP‐94 @ 2006 +
• 2006: No projection for CPM‐CAN
• 2012: 6‐Year Projection with short term scale
• UP‐94: Scale AA, static proj. to 2006 or 2012

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 10

CPM vs UP‐94 in 2006: ratios q(x) Male

90% 100% 110% 120%

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 11

50% 60% 70% 80% 60 65 70 75 80 85 90 95 Class 2= 35%‐94% Maximum Pension Class 3= 95%‐100% Maximum Pension Class 4= 35%‐100% Maximum Pension

CPM vs UP‐94 in 2012: ratios q(x) Male

90% 100% 110% 120%

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 12

50% 60% 70% 80% 60 65 70 75 80 85 90 95 Class 2= 35%‐94% Maximum Pension Class 3= 95%‐100% Maximum Pension Class 4= 35%‐100% Maximum Pension

SLIDE 5

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 4 CPM vs UP‐94 in 2006: ratios q(x) Female

90% 100% 110% 120%

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 13

50% 60% 70% 80% 60 65 70 75 80 85 90 95 Class 2= 35%‐94% Maximum Pension Class 3= 95%‐100% Maximum Pension Class 4= 35%‐100% Maximum Pension

CPM vs UP‐94 in 2012: ratios q(x) Female

90% 100% 110% 120%

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 14

50% 60% 70% 80% 60 65 70 75 80 85 90 95 Class 2= 35%‐94% Maximum Pension Class 3= 95%‐100% Maximum Pension Class 4= 35%‐100% Maximum Pension

CPM vs UP‐94: 2006 and beyond

• Male: depends on income class, but lower

than UP‐94 in 2012 until age 84

– Class 4 ages 74‐77: under 102%; Wide gap between income classes – Wide gap between income classes,

• For Female: CPM mortality is lower (age<87)
• Projected to 2015, 2020: lower ratios
• Next Charts: compare Classes 2, 3 & 4 only
• Also: compare Sources CPP, QPP, CAN

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 15

CPM‐CAN 2005‐2007 Male: Classes 2, 3 & 4

110% 120% 130% Class 2‐Low Class 2‐High Class 2‐Grad. Class 3‐Low Class 3‐High Class 3‐Grad. ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 16 70% 80% 90% 100% 60 65 70 75 80 85 90 Class 4‐Low Class 4‐High Class 4 ‐Grad.

SLIDE 6

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 5 Source 2005‐07 Male: CAN, CPP, QPP

102% 104%

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 17

96% 98% 100% 60 65 70 75 80 85 90 95 Canada Canada Pension Plan Quebec Pension Plan

• 4. Mortality Improvement Rate,

Formulas

Next Slide show formulas for reference Charts follow and illustrate trends

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 18

Excerpts from Phase III Draft Report (2012‐07)

Formulas: force, prob., weights

Deaths and exposure (exact): Probability of death, from force: Variance of force, and of prob.:

x x x

D ˆ μ = E

x

• μ

x

q =1-eˆ ˆ

D ˆ Weight:

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 19

( )

x

x x x μ 2 x x

μ D Var μ = = E E

2 ˆ

ˆ ˆ σ =

( )

( )

( )

x

2 μ x x

Var q = e × Var μ

−ˆ

ˆ ˆ

( ) ( ) ( )

2 x 2 x x x

E 1 w = = ˆ Var q ˆ 1-q ×D

t

Formulas: Regression

Improvement rate: Regression on ln(qx): linear form

( )

_ _

= 1

t init year t init year x x x

q q IR

+

× −

( ) ( )

( )

_ _

ln = ln ln 1

init year t init year x x x

q q t IR

+

+ × − Weighted Linear Regression (Min W, find slope):

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 20

( ) ( )

( )

x x x

q q

1

=

i i

y x β β + × ( )

2 1 1 n i i i i

W w y x β β

=

= × − − ×

∑

SLIDE 7

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 6

Formulas: slope, IRx

Slope factor, weighted linear regression:

1 1 1 n n i i i i n i i i i i n i i

w x w y w x y w

= = =

⎛ ⎞ ⎛ ⎞ × × × ⎜ ⎟ ⎜ ⎟ ⎝ ⎠ ⎝ ⎠ × × − ∑

∑ ∑ ∑

Improvement rate:

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 21

1 1( ) 2 2 1 1 1

ˆ

i w n i i n i i i n i i i

w x w x w β

= = = =

= ⎛ ⎞ × ⎜ ⎟ ⎝ ⎠ × −

∑ ∑ ∑ ∑

1

1

x

IR e β = −

Formulas: bounds, R2

Confidence interval: upper and lower bounds

1

ˆ 1( ) /2, 2 ˆ 1( ) /2, 2

ˆ ˆ

w n w n

LB t s UB t s

α β α β

β β

− −

= − × = + ×

( ) 1 ( ) 1

LB UB

rate LB e rate UB e = − = −

Worth of regression: R2

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 22 1

1( ) /2, 2 w n α β

( )

( ) ( )

2 2 1 2 1

ˆ

n i i i n i i i

w y y R w y y

= =

× − = × −

∑ ∑

15‐Year Regression Ending in 2007, CAN‐4‐M, Age 70

3 2% 3.4%

CAN-4-M, Age 70, Prob. of Death q70 Weighted Linear Regression, 15-Year Ending in 2007 q(70,1992+t)=q(70,1992)×(1-IR70)t, IR=2.78%, 95% C.I.= 2.46%-3.11%, R2=96.0%

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 23

2.0% 2.2% 2.4% 2.6% 2.8% 3.0% 3.2% 1992 1997 2002 2007 Probability of Death Observed Regression

15‐Year Regression Ending in 1992, CAN‐4‐M, Age 70

3 70% 3.80% 3.90%

CAN-4-M, Age 70, Prob. of Death q70 Weighted Linear Regression, 15-Year Ending in 1992 q(70,1977+t)=q(70,1977)×(1-IR70)t, IR=1.0747%, 95% C.I.= 0.736%-1.412%, R2=76.7%

Observed Regression

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 24

3.10% 3.20% 3.30% 3.40% 3.50% 3.60% 3.70% 1977 1982 1987 1992 q(70) Regression

SLIDE 8

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 7

15‐Year Regression Ending in 2007, CAN‐4‐M, Age 95

30.0% 31.0%

CAN-4-M, Age 95, Prob. of Death q95 Weighted Linear Regression, 15-Year Ending in 2007 q(95,1992+t)=q(95,1992)×(1-IR95)t IR= -0.14%, 95% C.I.= -0.74%-0.46%, R2=1.6%

Observed

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 25

23.0% 24.0% 25.0% 26.0% 27.0% 28.0% 29.0% 1992 1997 2002 2007 Probability of Death Regression

CAN‐4‐M: 15 years in 2007

80% 100% 3.0% 4.0% nt Rate

CAN-4-M:15-year Period Ending in 2007

Mean, 95 % Confidence Interval and R2 (Right Axis)

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 26

0% 20% 40% 60%

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 100 Mortality Improvemen Mean Lower Upper R2

Age 70: 2.78% C.I.: 2.46% - 3.11% R2: 96%

CAN‐4‐F: 15 years in 2007

80% 100% 3.0% 4.0% ent Rate

CAN-4-F:15-year Period Ending in 2007

Mean, 95 % Confidence Interval and R2 (Right Axis)

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 27

0% 20% 40% 60%

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 100 Mortality Improveme Mean Lower Upper R2

Results vary by length of regression

3.0% 4.0% nt Rate

CAN-4-M Mortality Improvement Rate

Various Lengths of Regression Period Ending in 2007

Maximum 30 years 25 years 20 years 15 years 10 years 5 years ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 28

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 100 Mortality Improvemen y

SLIDE 9

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 8

Results of Last 30 Years (Males)

3.0% 4.0% Rate

CAN-4-M Mortality Improvement Rate

15-year Regression Periods 1992-2007 and 1977-1992

30 years 1977-2007 15 years 1992-2007 15 years 1977-1992 1992-2007 Lower Bound 1992-2007 Upper Bound 1977-1992 Lower Bound

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 29

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 Mortality Improvement

1977-1992 Upper Bound

Improvement by Classes, CAN‐M:2,3,4,5

3.0% 4.0% Rate

CAN-M Mortality Improvement Rate

by Income Classes 15-year Regression Period Ending in 2007

Class 2 Class 3 Class 4 Class 5

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 30

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 Mortality Improvement

Class 5

Improvement by Source, 4‐M:CAN, CPP, QPP

3.0% 4.0% t Rate

4-M Mortality Improvement Rate

by Data Source (CAN, CPP, QPP) 15-year Regression Period Ending in 2007

Canada CPP QPP ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 31

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 Mortality Improvement

• 5. Projection Scales, Proposal
• 15‐year: 1992 to 2007 mirrored to 2006‐2021
• CPM‐CAN Experience
• Income class 4
• Long term: blend of CPP and QPP

g

– 2060 assumptions in December 2009 Report

• Mid term: transition from 2021 to 2030
• Based on blended C/QPP 2020 assumptions,

adjusted

• Impact: q(x) decreases faster, higher e(x)

higher ä(x), higher actuarial liabilities

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 32

SLIDE 10

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 9

Proposal: 3 Projection Scales

Short term scale: 2006 to 2021, 15 years Mid term scale: 2021 to 2030, 9 years

( )

2006 2006 Short term

1 ; 1 15

k Male k Male Male x x x

q q IR k

+ =

× − ≤ ≤

( )

15 2006 2006 Short Term

1

Male k Male Male x x x

q q IR

+ =

× −

Long term scale: 2030 and after

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 33

( ) ( )

15 Mid Term

1 ; 16 24

x x x k Male x

q q IR k

−

× − ≤ ≤

( ) ( ) ( )

15 9 2006 2006 Short Term Mid Term 24 Long Term

1 1 1 ; k 25

Male k Male Male Male x x x x k Male x

q q IR IR IR

+ −

= × − × − × − ≥

Projection Scale: Male

2.5% 3.0% 3.5% ent Rate

Mortality Projection Scales, Male

1992-2007 1977-1992 Scale AA

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 34

0.0% 0.5% 1.0% 1.5% 2.0% 60 65 70 75 80 85 90 95 Mortality Improvem Scale AA Short Term Long Term C/QPP 2020 C/QPP 2030 C/QPP 2040 Mid Term

Projection Scale: Female

2.5% 3.0% 3.5% ment Rate

Mortality Projection Scales, Female

1992-2007 1977-1992 S l AA

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 35

0.0% 0.5% 1.0% 1.5% 2.0% 60 65 70 75 80 85 90 95 Mortality Improvem Scale AA Short Term Long Term C/QPP 2020 C/QPP 2030 C/QPP 2040 Mid Term

Results

• Charts show impact on a generational basis
• Change from UP‐94 G to CPM‐CAN‐4 2005‐

2007 U 94 G AA j i S l

• UP‐94 G : AA Projection Scale
• CPM‐CAN‐4 2005‐2007: Short/Mid/Long

Projection Scales

• Effect on complete life expectancy and PV of

life annuity‐due

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 36

& $1,000

x x

e a ×

SLIDE 11

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 10 PV, i=3%, Valuation in 2012: Male

Age UP‐94G CPM‐CAN‐4‐M % Increase

60 17,032 17,506 2.78% 65 14,729 15,097 2.50%

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 37

, , 70 12,434 12,648 1.72% 75 10,099 10,169 0.68% 80 7,876 7,811 ‐0.83% 85 6,032 5,727 ‐5.06%

Impact on PV at 3%: Male (generational)

• % Increase in

ä(x)

• From UP-94 G

to CPM-CAN- 4-M

0% 1% 2% 3% 4% 60 6 80 8 2006

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 38

• Generational:

with impact of projection scales

• 3 Valuation

Years

‐7% ‐6% ‐5% ‐4% ‐3% ‐2% ‐1% 60 65 70 75 80 85 2012 2016

PV, i=3%, Valuation in 2012: Female

Age UP‐94G CPM‐CAN‐4‐M % Increase

60 18,201 18,857 3.61% 65 16,033 16,572 3.36%

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 39

, , 70 13,812 14,198 2.80% 75 11,472 11,733 2.27% 80 9,146 9,255 1.19% 85 6,983 6,903 ‐1.14%

2% 3% 4% 2006

Impact on PV at 3%: Female (generational)

• % Increase in

ä(x)

• From UP-94 G

to CPM-CAN- 4-F

‐2% ‐1% 0% 1% 60 65 70 75 80 85 2006 2012 2016

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 40

• Generational:

with impact of projection scales

• 3 Valuation

Years

SLIDE 12

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 11

• 6. Conclusion
• Canadian pattern of mortality known
• Cost of pensions using UP‐94 and AA may be

underestimated d i li

• Recent trend in mortality

– faster decrease than thought with previous scales – not known when it will trail off – No crystal ball: use consensus for long term

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 41

Next Steps

• CIA: decision to release Phase II and Phase III

reports

• CIA: may provide additional comments

i i CIA ll 2012 i S i

• Discussion at CIA Fall 2012 Pension Seminar
• Actuarial Standards Board (Canada): decide

future recommendations for mortality tables for pensions plans purposes

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 42

Thank you !

Length of Regression: CAN‐4‐M in 2007

3.0% 4.0% ate

CAN-4-M Mortality Improvement Rate

Various Lengths of Regression Period Ending in 2007

Maximum 30 years 25 years 20 years 15 years 10 ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 67

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 100 Mortality Improvement Ra 10 years 5 years

SLIDE 13

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 12 Length of Regression: CAN‐4‐F in 2007

3.0% 4.0% ate

CAN-4-F Mortality Improvement Rate

Various Lengths of Regression Period Ending in 2007

Maximum 30 years 25 years 20 years 15 years 10 years ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 68

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 100 Mortality Improvement Ra 10 years 5 years

Improvement by Classes, CAN‐F:2,3,4,5

3.0% 4.0% t Rate

CAN-F Mortality Improvement Rate

by Income Classes 15-year Regression Period Ending in 2007

Class 2 Class 3 Class 4 Class 5

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 69

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 Mortality Improvement

Improvement by Source, 4‐F:CAN, CPP, QPP

3.0% 4.0% Rate

4-F Mortality Improvement Rate

by Data Source (CAN, CPP, QPP) 15-year Regression Period Ending in 2007

Canada CPP QPP ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 70

• 1.0%

0.0% 1.0% 2.0% 60 65 70 75 80 85 90 95 Mortality Improvement R QPP

Impact on Life Expectancy: Male (generational)

• % Increase in

complete e(x)

• From UP-94 G

to CPM-CAN- 4-M

0% 1% 2% 3% 4% 2006

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 71

• Generational:

with impact of projection scales

• 3 Valuation

Years

‐7% ‐6% ‐5% ‐4% ‐3% ‐2% ‐1% 60 65 70 75 80 85 2006 2012 2016

SLIDE 14

Actuarial Research Conference, Winnipeg, Session P4A, Projection Scales 2012-08-03 Louis Adam, FSA, FCIA, Université Laval 13 Impact on Life Expectancy: Female (generational)

• % Increase in

complete e(x)

• From UP-94 G

to CPM-CAN- 4-F

2% 3% 4% 5% 2006

ARC, Winnipeg. Mortality Improvement for Canadian Pensioners: Proposed Projection Scales 72

4 F

• Generational:

with impact of projection scale

• 3 Valuation

Years

‐3% ‐2% ‐1% 0% 1% 60 65 70 75 80 85 2006 2012 2016