the breakdown of society… …by past and projected mortality

03/05/2017

1

The breakdown of society…

…by past and projected mortality

experience

Steven Baxter, Hymans Robertson LLP

Alice Woolley, RMS

03 May 2017

Colour palette for PowerPoint presentations

Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184

Secondary colour palette

Primary colour palette

Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

The next 30 minutes…

03 May 2017 2

Historical perspective… Building projections…

03/05/2017

2

A social history of improvements Why socio-economics matter

03 May 2017


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

68

70

72

74

76

78

80

82

84

86

88

1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010

Period life expectancy from age 65

Men Women

170 years of longevity..

03 May 2017 4

Infectious diseases (Respiratory &) Circulatory

Source: Hymans Robertson using data from ONS and Human Mortality Database

03/05/2017

3


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

A long history of socio-economic differences

03 May 2017 5

Source: BBC via YouTube


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Why socio-economics matter

03 May 2017 6

Smoking

Over half the difference in

premature death between SECs

(Sharma et al, Wanless et al; death before age 70)

Smoking cessation

Higher SECs nearly three times

more likely to successfully quit

(Hiscock et al)

03/05/2017

4


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Why socio-economics matter

03 May 2017 7

£

£

Typical annuitant

portfolios

Finance /

Exec

portfolio

Heavy

industrial

portfolio

Source: Club Vita. Coloured areas relate to Club Vita based socio-economic classes (VitaSegments) which are subtler than the

Cleese / Barker / Corbett socio groupings.


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Key features

1. Calibrated to over 20

years of BPA back history

2. Based on 2m+ lives / 60k

annual deaths

3. Use widely available

variables

4. Statistically credible

groups capturing

differences in historical

mortality improvements

Hard-Pressed Making-Do Comfortable

Hard-Pressed Making-Do /

Comfortable

Source: Club Vita

03 May 2017 8

03/05/2017

5


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Latest emerging evidence

03 May 2017 9

18.0

19.8 20.0 20.3

14.4

16.717.0 17.0

12

13

14

15

16

17

18

19

20

21

22

23

24

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015

Life Expectancy at age 65 (men)

Not slowing down?

Notes to chart:

(i) Points relate to life expectancy for the calendar year (using crude mortality smoothed over the age range via a Gompertz curve)

(ii) Dashed lines apply three year smoothing i.e. calendar year and year prior and after)

(iii) Labels relate to the values under three year smoothing

Convergence – health cascade? Slow down

Been going up slower…

Source: Club Vita / Hymans Robertson

Hard-Pressed

Making-Do

Comfortable


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

• Lower socio-economic groups had been catching-up (convergence, 2005-2010)…

…but recently slowed down (divergence, 2010-2015)

Underlying improvements

• Annuitant improvements mirror national data on a lives basis (including recent

slowdown)

03 May 2017 10

Group Annualised mortality improvement (age-standardised)

2000-2005 2005-2010 2010-2015

England & Wales 2.8% (±0.1%) 2.8% (±0.1%) 1.1% (±0.1%)

Club Vita 2.4% (±0.5%) 2.8% (±0.3%) 1.3% (±0.4%)

Comfortable 2.4% (±1.1%) 2.1% (±0.8%) 2.1% (±0.7%)

Making-do 2.2% (±0.8%) 3.2% (±0.5%) 0.9% (±0.6%)

Hard-pressed 2.5% (±0.7%) 2.9% (±0.5%) 1.0% (±0.6%)

1

2

3

1

• Comfortable annuitants who dominate liabilities seen stabler improvements and no

recent slow down 2

3

Source: Club Vita / Hymans Robertson. Confidence intervals calculated consistently with the approach set out in CMI WP97

03/05/2017

6


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Tail winds, head winds and side winds…

03 May 2017 11


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Unpacking the VitaSegments

03 May 2017 12

Source: Club Vita / Hymans Robertson. Statistics quoted during seminar using Club Vita data and data from English

Longitudinal Study of Ageing (ELSA) All comments relate to those in retirement with DB pensions (typically aged 65+)

Hard-pressed Comfortable

03/05/2017

7

Does it make sense to segment

projections by socio-economic

class?

03 May 2017


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Cancer mortality varies between SECs

Percentage Deprivation Gap in

Age-Standardised Mortality

Rates

(Males, England 2007-2011)

03 May 2017 14

03/05/2017

8


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Does one trend fit all?

Industry Assumption:

New developments in the real world will trigger

similar responses across SECs.

03 May 2017 15


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Cancer survival rates vary

Early detection is

strongly linked with

higher survival rates

03 May 2017 16

0%

20%

40%

60%

80%

100%

Breast Cancer Bowel Cancer Ovarian Cancer

5-year female relative survival, adults aged 15-99

Stage I Stage II Stage III Stage IV

03/05/2017

9


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Cancer screening

Screening is the

most effective

tool for catching

cancer sooner

03 May 2017 17


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Responses to screening programs vary by SEC

Bowel Cancer screening

• 2006: screening using blood tests can reduce mortality by up to 25%.

• Screening uptake showed marked differences by SEC.

• 2015: more invasive screening test (tiny camera) can reduce mortality by up to 43%.

• “People from poorer neighborhoods were less likely to take up the offer than those in

affluent areas”

03 May 2017 18

Richard F A Logan et al. Gut doi:10.1136/gutjnl-2011-300843

IMD quintile Uptake of faecal occult blood test screening (%)

5 (most deprived) 41.7

4 50.1

3 55.5

2 59.6

1 (least deprived) 61.4

03/05/2017

10


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Does one trend fit all?

Industry Assumption:

New developments in the real world will trigger

similar responses across SECs.

03 May 2017 19

Building a segmented longevity

trend model

03 May 2017

03/05/2017

11


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

What is driving the changes in mortality rates?

03 May 2017 21

Lifestyle

Health

Environment

Medical

Intervention

Regenerative

Medicine

Anti-Aging

Processes

Now Later

Much

Later


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Medically informed projection constraints

03 May 2017 22

How

much?

How fast?

Mo

rta

lity

lowest mortality achievable

03/05/2017

12


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Quantifying lowest mortality: Lifestyle

03 May 2017 23

Potential

mortality

reduction

Cu

rre

nt M

ort

alit

y

Ide

al M

ort

alit

y


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Parameterizing a segmented model

03 May 2017 24

How can we derive the

mortality improvements by

Cause of Death for each

SEC?

Segmented Longevity

Trend Model

Where can we find a

credible dataset for

parameterization

purposes?

What would be a credible

SEC segmentation which,

in addition, can be easily

applied?

What additional insights

from medical science can

inform the projections by

SEC? How can we specify a

correlation framework which

reflects the relationships in

SEC improvements?

1

2

3

4

5

03/05/2017

13


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

How can we specify a

correlation framework which

reflects the relationships in

SEC improvements?

M C M MD M HP F MDC F HP

M C

M MD

M HP

F MDC

F HP

What additional insights

from medical science can

inform the projections by

SEC?

How can we derive the

mortality improvements by

Cause of Death for each

SEC?

Where can we find a

credible dataset for

parameterization

purposes? “Tracking over 1 in 7 of the

UK’s pensioners”

Segmented Longevity

Trend Model

Parameterizing a segmented model

03 May 2017 25

What would be a credible

SEC segmentation which,

in addition, can be easily

applied?

1

2

3

4

5

What could trend segmentation

mean for my business?

03 May 2017

03/05/2017

14


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

940 960 980 1000 1020 1040 1060 1080 1100

Liability Distribution (£m)

SegmentedModel

Scenario: heterogeneous pool of

annuity business, size £1bn

Reduced best estimate:

• Better reflecting socio-economic

composition of business

Reduced risk distribution:

• Greater certainty in trends for

some SECs

• Reduced basis risk

• Diversification across SECs

Impact on Illustrative Book

03 May 2017 27

940 960 980 1000 1020 1040 1060 1080 1100

Liability Distribution (£m)

PopulationModel

50th:

-0.7%

(95th-50th):

-24%

1

2


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Implications for New Business

28

0.6%

-0.3%

-0.7%

-0.6%

IV

III

II

I

Scenario: 4 identically-sized

pension schemes

Segmented approach refines

liability estimates:

• Better pricing

• Improved risk selection

• Enhanced competitiveness

Be more

competitive?

Selection

risk?

Change in liabilities due

to segmentation

03/05/2017

15


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

Summary

03 May 2017 29

One size does not fit all

Socio-economic differentiation offers

commercial opportunities


Dark blue

R17 G52 B88

Gold

R217 G171 B22

Mid blue

R64 G150 B184



Light grey

R63 G69 B72

Pea green

R121 G163 B42

Forest green

R0 G132 B82

Bottle green

R17 G179 B162

Cyan

R0 G156 B200

Light blue

R124 G179 B225

Violet

R128 G118 B207

Purple

R143 G70 B147

Fuscia

R233 G69 B140

Red

R200 G30 B69

Orange

R238 G116 29

Dark grey

R63 G69 B72

03 May 2017 30

Expressions of individual views by members of the Institute and Faculty

of Actuaries and its staff are encouraged.

The views expressed in this presentation are those of the presenters.

Questions Comments

the breakdown of society… …by past and projected mortality

Documents