Dr. Andrew Coburn
Senior Vice President, RMS LifeRisks
The Long-Term Rate Setting mortality improvement assumptions
for longevity liabilities
Investigating Future Mortality: Blending Medical & Actuarial Science for Life & Longevity Risk Management
RMS LifeRisks Seminar l May 24, 2012
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
In this presentation I‟m going to focus on:
The 10-year Average Annual Improvement Rate
Prevailing at 2032 – twenty years‟ time
80 Year Old Males
– Modal pension in payment
England and Wales Population
Ignore any Cohort Effect
Equivalent to what is known as the CMI Long-Term Rate
– Technical definition of LTR is 2048, but with cohort effects exhausted this is the same as the 2032 rate in CMI model
What Will the Rate of Mortality Improvement be at 2032?
2
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
A Century of Change and then…
40 Years of Unrelenting Improvement
3
Mortality Improvement Rates for 75-85 year old UK Males 1841-2009
(10-Year Average Annual MI plotted at 5-year mid-point)
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020
10-y
ea
r A
ve
rag
e A
nn
ua
l M
ort
ality
Im
pro
ve
me
nt
Dickensian
Dystopia
The Sanitary
Reformation
The Conquest of
Infectious Disease
The Smoking
Epidemic
The Health
Consciousness
Revolution
Targeting
CVD Risk
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Lifestyle Trends
– Smoking reduction
– Health awareness
Medical Intervention
– Prevention
– Response
– Post-event treatment
Health Environment
– Increasing expenditure on
healthcare (5% a year)
– Reduction in passive smoking
– Internet spreading health
information
The past 40 years has been…
A Success Story in Reducing Premature CVD Deaths
4
0
10
20
30
40
50
60
70
1980 1990 2000 2010
An
nu
al
Death
s p
er
1000
CVD
Cancer
Other Causes
Mortality by Cause
Males 75-84 years, E&W
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Lifestyle Trends
5
Smoking rates halved since 1970s
– Public smoking bans led to 15% decrease in
heart attacks
– Bans on TV advertising, warnings on cigarette
packets, screening sales
Growing health awareness in diet
– e.g. „Look after your heart‟ low fat substitutes
introduced in 1980s reduced butter sales
– Calorie labeling now common
Health awareness in exercise
– e.g. Gym membership increased from 4% in
1998 to 14% by 2006
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Medical Intervention - Prevention
Antihypertensive drugs reduce blood pressure and reduce heart death risk
– BP drugs became widely available from 1980s
– 20% of UK adults now take BP medication
– Treatment accounts for 12% of all doctor activity
– NHS spends £1 Bn on BP drugs
Statins reduce cholesterol and prevent clogging of arteries
– 4S trial in 1994 proved that statins reduced CVD mortality risk by 30%
– Rapid uptake of statins in population
– 33% of adults over 45 on statins by 2008
– Pfizer‟s Lipitor statin became world‟s best-selling drug in 2008 with sales of $12.4 Bn
6
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Medical Intervention - Response
7
Past two decades has seen major
advances in responding to a CVD event
When a heart attack occurs, speed of
attention is critical
– Medical response times reduced from
around 45 mins to current target of 8 mins
– Better organization of emergency services
– Fast-response paramedics
Faster notification of emergency services
– Mobile phones have become ubiquitous
– 15% of population had phones in 1990
– >80% of population has phones today
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Medical Intervention – Treatment
8
Treatment of someone who has had a heart attack has been revolutionized
Heart surgery has become safer and more common
– New techniques: stents; angioplasties; bypass surgery; revascularization
New classes of drugs introduced
– Blood thinning, clot dissolving (thrombolysis), resistance to platelet formation
Pacemaker technology advances
– Smaller and cheaper
Survival rates doubled
– Mortality within a year 45% in 1995 reduced to 25% today
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
-1%
0%
1%
2%
3%
4%
1980 1990 2000 2010 2020 2030 2040
2%
1%
1.5%
0 years
per
decade
1 year
per
decade
2 years
per
decade
An
nu
al
Mo
rtali
ty I
mp
rove
men
t The Long-Term Rate
9
80-year-old male,
without cohort effects
Actual 3-year average annual
mortality improvement, 75-85 yr old male
Smoothed annual mortality improvement,
75-85 yr old male
2032 R
efe
rence D
ate
for
Level of LT
R
1.25% 1.5%
>1.5%
Pension Funds
Insurers
Investors
Cairns-Dowd
M3 Model
KPMG Survey May 2011 of pension fund clients using the CMI 2010 model
“A large majority of clients still use an earlier method of
cohort projection predominantly with a Long-Term rate of
1% for both men and women”
<1%
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
What Might Cause an LTR of 1.5% in 20 years?
10
Could an Extreme Trend Reversal be caused by:
Rapid exhaustion of reducing premature CVD
deaths?
Failure to reduce cancer mortality?
Obesity Epidemic?
General slowdown in the economy?
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
CVD Mortality Improvement Could Slow if…
11
Most of the preventable CVD deaths have already been
prevented, and remaining deaths are a lot harder to reduce
No further improvements will be made in the prevention,
response, or treatments of CVD events
All of the people who could ever benefit from hypertensives
and statins are already taking them
Lifestyle changes will stop spreading through the population
– Smoking rates, diet and exercise levels stop improving
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Modeling the Scenario of Rapid CVD Exhaustion
12
-2%
0%
2%
4%
6%
1980 1990 2000 2010 2020 2030
Mo
rtality
im
pro
vem
en
t (p
er
an
nu
m),
ag
e 7
5-8
4
All
CVD
Cancer
Other
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
But CVD Exhaustion in this Timeframe is Unlikely
13
Because
There is still very significant potential for further improvement
– Limits aren‟t approached until 4.5 deaths per 1000
– Continuation of current rate would see 8 per 1000 by 2032
Lifestyle changes will continue, smoking rates will continue to diminish (question is how fast)
– Today‟s smokers don‟t need to quit: the 60 year olds who will be 80 in 2032
smoke less than the 60 year olds did in 1992
Many more people can benefit from taking preventative hypertensive or statin medication
– Today‟s takers are only those who self-refer or are identified through other checks
– Adherence to drug regimes is still sub-optimal
– Discussion of putting everyone over 50 on statins or „polypill‟
Medical treatments are continuing to improve
New drugs for CVD are currently in clinical trials
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Scenarios for Extreme Trend Reversal
14
0%
1%
2%
3%
4%
5%
1980 1990 2000 2010 2020 2030
Mo
rtality
im
pro
vem
en
t (p
er
annum
), a
ge 7
5-8
4
Failure to make progress in
reducing cancer mortality
Obesity „Shock‟ 60% of 80 yr olds
Obese by 2032
Rapid CVD
Exhaustion
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
No Obvious Candidate to Cause Extreme Trend Reversal
15
There is no plausible mechanism of sufficient magnitude to
halve the current rate mortality improvement in the next 20
years
– „Rapid CVD Exhaustion‟ could do it, but there is still room for
improvement at close to the current rate for many years
– „Failure to Cure Cancer‟ can‟t do it as progress is expected to be
slow anyway
– „Acceleration of Obesity Epidemic‟ can‟t do it, even in extreme
scenarios which are implausible
In the absence of an obvious candidate for driving trend
reversal, what would happen if trends continue?
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
-2%
0%
2%
4%
6%
1980 1990 2000 2010 2020 2030
Mo
rtali
ty i
mp
rove
me
nt
(pe
r a
nn
um
), a
ge
75
-84
All
CVD
Cancer
Other
0%
2%
4%
6%
8%
10%
12%
1980 1990 2000 2010 2020 2030
Mo
rta
lity
ra
te,
ag
e 7
5-8
4
All
CVD
Cancer
Other
A „Trend Neutral‟ Assumption
16
Mortality Rate Mortality Improvement
There is a natural diminution of trend rate as CVD deaths fall and
cancer and other causes of death begin to dominate the overall
mortality contribution
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Unchanged or Accelerated
17
0%
1%
2%
3%
4%
5%
1980 1990 2000 2010 2020 2030
Mo
rtality
im
pro
vem
en
t (p
er
annum
), a
ge 7
5-8
4
Further
Acceleration
of Trends 3X CVD
5X Cancer
2X Other Causes
Trend Neutral
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
How might the next 20 years be different to the last 20?
“It‟s the Economy, Stupid”
18
0.0
0.5
1.0
1.5
2.0
2.5
3.0
1970 1975 1980 1985 1990 1995 2000 2005 2010
UK GDP
GD
P (
$T
rilli
on
s 2
01
0)
-6.0
-4.0
-2.0
0.0
2.0
4.0
6.0
8.0
1970 1975 1980 1985 1990 1995 2000 2005 2010
UK GDP Growth
An
nu
al
% C
han
ge
Average GDP Growth
1979-2007: 3.5%
Bank of England
GDP Forecast
February 2012
Average GDP Growth
2009-2015: <1%
If we return back to 3.5% growth
after 2015, then
2009-2032: <2.5%
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
NHS Funding Will be Frozen for at Least 3 Years
19
Source: Harker, R., 2011, „NHS funding and expenditure‟; UK Government House of Commons Library, Social and
General Statistics, Standard Note SN/SG/724; 14 September 2011. http://www.nhshistory.net/parlymoney.pdf
1999-2008 >5%
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Consequences of NHS Funding Freeze
20
Reduced capability to expand healthcare, do new initiatives or
extend prevention programmes to lower risk individuals
– Rate of expansion of antihypertensives & statins will slow
– Ability to be more proactive in preventative care is reduced
New facilities, such as specialist NHS cancer diagnostic
centres, originally planned for 2015, postponed
Acquisition of expensive new drugs is limited
– Which in turn has implications for the pharmaceutical industry
Has an impact on mortality improvements resulting from
medical interventions and the health environment
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Crisis in the Pharmaceutical Industry
21
Economic crisis has accelerated and coincided with a major readjustment in bio-tech research & development
Big Pharma accuses European governments of cost cutting:
– “5% cut-back in drug expenditure and deliberate delay in introducing new cancer drugs” – Sir
Andrew Witty, GSK CEO
Pharmaceutical companies radically restructuring their business models as
– Biggest selling drugs come off patent
– No new „blockbuster‟ drugs have been discovered
– Drug development pipelines shrink
R&D returns have halved over the past decade
Pfizer, GSK, Novartis
– Big layoffs in US and Europe & disbanding research operations
Slow down in the rate of medical advance
– “What would have taken 10 years of research in good economic times will take 12-15 years
during austerity” – Sir Andrew Witty, GSK CEO
Business model of Pharmaceutical companies is shifting away from drugs targeted on mass populations towards niche applications
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Scenarios for Lower Economic Growth
22
0%
1%
2%
3%
4%
5%
1980 1990 2000 2010 2020 2030
Mo
rtality
im
pro
vem
en
t (p
er
annum
), a
ge 7
5-8
4
Trends Moderated
by Lower Growth
(<2.5%) Trends Moderated by
Economic Stagnation (Growth <1%)
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
0%
1%
2%
3%
4%
5%
1980 1990 2000 2010 2020 2030
Mo
rtali
ty i
mp
rovem
en
t (p
er
annum
), a
ge 7
5-8
4
Historical
1 CVD Exhaustion
2 Cancer Failure
3a Obesity Burden on 3%
3c Accelerated Obesity + Trend Neutral
4 Trend Neutral
5 Accelerated
6 Trend + Lower Economic Growth
7 Trend + Economic Stagnation
Summary – Covering the range of LTR Options
23
1
2
3a
4
5
7
6
3c
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Descriptions of Long-Term Rates of Mortality Improvement
24
Longevity
Risk
Category
LTR Range LTR
Representative
Value
Summary Description
1 Below 0.75% 0.25% Social collapse, decline in national healthcare performance and population health
metrics, consistent with lengthy economic recession. Smoking rate remains at or
above 25%. Minimal medical advances.
2 0.75% to 1.25% 1.0% Trends in mortality reduction slow to below half of their recent levels, consistent with
long-term economic stagnation. 20% smokers by 2032. Adoption of healthier lifestyle
ceases.
3 1.25% to 1.75% 1.5% Trends in mortality reduction slow to two-thirds of recent trends. 18% smokers by
2032. Obesity epidemic accelerates. Healthcare budget reductions result in slower
mortality improvements in CVD and cancer. Biotech advances curtailed.
4 1.75% to 2.25% 2.0% Recent trends continue, with CVD improvement moderating as other causes of death
become more significant. 15% smokers in 2032. Improvements may be further
moderated by reduced outlook for economic growth that perpetuates current NHS
funding freeze and slows private sector biotech investment.
5 2.25% to 2.75% 2.5% Mortality improvement levels of the past 20 years are continued for the next 20
years. 12% smokers in 2032. Slowing of CVD improvement offset by increased
improvements from other causes of death. Obesity epidemic slows. Cancer
outcomes in U.K. rapidly achieve those of the European average.
6 2.75% to 3.5% 3.0% Mortality improvement level of the past 10 years perpetuated for next 20 years,
enabled by economic growth that improves the funding and performance of a
prevention-focussed UK healthcare system. 8% smokers by 2032. Cancer
breakthroughs. Regenerative medicine treatments for previously terminal conditions.
7 Above 3.5% 4.5% U.K. healthcare comparable with U.S. privatized healthcare, enabled by a doubling
of per capita funding. 4% smokers by 2032. Premature death from CVD almost
eliminated. Cancer mortality halved. Regenerative treatments impact mortality.
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
It is important that each company owns its own view of LTR
– Our analysis is intended to provide context for that decision
All projections combine some component of historical experience with
expectations of future differences
– Those using lower rates (<1.5%) may be giving too much weight to the pre-
1970 historical past
Projections of mortality improvement should explicitly incorporate the
assumptions being made on future medical progress, lifestyle trends,
and healthcare funding levels
The underlying processes that are driving higher levels of mortality
improvement than in the historical past are now endemic to our modern
society
Cause of improvement modeling can help the managers of longevity risk
set realistic longevity estimates that provide financial stability for all
parties
Conclusions
25
CONFIDENTIAL © 2012 Risk Management Solutions, Inc.
Currently under going final
review
Release expected Summer 2012
RMS LifeRisks Whitepaper on Long Term Rates
26