Topical issues in CI pricingDarshan Singh & Alex King

Issues Recent views on CI guarantees How different reinsurers viewpoints have changed Impact of new definitions and new diseases Recent trends in key diseases & future medical

advances Base level of morbidity including selection factors Changes to the ABI claims matrix

CI Guarantees…… a brief bit of history! Pre 2003, CI market was very competitive Most insurers offered both guaranteed & reviewable CI

policies Small margin between pricing of reviewable &

guaranteed CI – most sales were guaranteed Volumes were excellent – almost 1.2m policies sold in

2002, up almost 50% on prior year Competition over number of definitions Large proportion reinsured (90% common) often on

nil-premium structures. Mostly with 2 reinsurers (GE & Swiss)

CI guarantees – so what changed?!

Swiss Re withdrew from the market thus reducing capacity

Other reinsurers either increased prices or backed away from long-term guarantees – some offered 5 years!

Many insurers also withdrew guaranteed product (e.g. Zurich, AEGON, Pru)

The insurers that remained active increased prices by 50-60% in the first couple of months of 2003

Prompted “fire sale” as belief grew that the future for guaranteed CI was bleak

CI Guarantees – outcome? Insurers reinsured less Maximum benefits reduced – sometimes as low as

£250k Maximum term was capped at 25 years Stand-alone CI was priced same as Accelerated (or

withdrawn) Volumes fell to around half the 2002 peak – 550k

policies in 2006

Guaranteed & Reviewable CI PricingGuaranteed vs Reviewable Prices

1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 (?)

Year

Guarantee Loading

0

5

10

15

20

25

2002 2005 2008

GuaranteeLoad

Little difference pre-2003 between gtd & rev

Loadings increased due to reduced reinsurance capacity and uncertainty around prostate cancer, leukaemia, troponins & silent strokes

New ABI Definitions have helped to reduce uncertainty and lower the guarantee loadings

Lately very competitive reinsurance markets are driving down guarantee loadings

0

1

2

3

4

5

6

7

8

<2002 2003 - 4 2005 - 6 2007>

Number of reinsurers quoting

Guaranteed CI

Reviewable CI

Changing reinsurance landscape for guaranteed CI

Issues Recent views on CI guarantees How different reinsurers viewpoints have changed Impact of new definitions and new diseases Recent trends in key diseases & future medical

advances Base level of morbidity including selection factors Changes to the ABI claims matrix

Impact of new definitions and diseases Key changes to definition

Troponin hurdle for heart attacks Future proofing of cancer Introduction of permanent neurological deficit

New diseases added Traumatic head injury

Pricing issues around troponins Increase in heart attack incidence: in 2001/02 with

troponin testing starting to become widespread UK coverage of troponin testing is not complete as yet:

2006 chest pain survey1 showed 58% of hospitals having Troponin T capability and 44% having Troponin I (compared to 32% and 23% respectively in 2001)

Authors state: “Development of chest pain services in the UK is progressing in a disorganised way”

Only 90% of the heart attack claims we audited in 2007 had troponin measurements

Past experience needs to be adjusted for the impact of troponins

There may be future adverse trends as coverage becomes complete and claims practices bed down

Pricing issues around troponins

Troponins can also be released: During cardiac surgery In endurance events Septic shock Pulmonary embolism Scorpion venom

Potential for these to subsequently become claims – more so cardiac surgery

Depends on claims philosophy and enforcement of other pillars of the definition

Future proofing of cancerLargely future proofed except that some terms could become obsolete

in the future (shown in bold below);

All cancers which are histologically classified as any of the following: pre-malignant,; non-invasive; cancer in situ; having either borderline malignancy; or having low malignant potential

A small risk – but a risk nonetheless

Permanent neurological deficit with persisting clinical symptoms

Symptoms of dysfunction in the nervous system that are present on clinical examination and expected to last throughout the insured person's life.

Symptoms that are covered include…. [a big list that takes up two slides!].

The following are not covered: An abnormality seen on brain or other scans without definite related clinical

symptoms Neurological signs occurring without symptomatic abnormality, e.g. brisk

reflexes without other symptoms Symptoms of psychological or psychiatric origin

Where does the burden of proof lie on these? The Brain is a complex organ and there is no severity criterion in the definition

Permanent neurological deficit Offices have not had issues so far with the generic term Issue could arise in the future as medical science is

able to track more deficits back to injuries to the brain Could the following be causes of claim in the future?

Are we looking at the next TPD with many declined claims? Seeing flashing lights Vague cognitive impairment – no longer the same

person Inability to orgasm

My Jerry Springer Slide: Reduced libido - about half of people with traumatic

head injury experience a drop in sex drive2. The remainder experience increased libido, or no change at all.

Erectile problems - between 40 and 60 per cent of men have either temporary or permanent impotence following their injury2.

Inability to orgasm - up to 40 per cent of men and women report difficulties having an orgasm2

Is the insurance industry equipped to deal with

claims of this nature? Can you imagine a claims form

for this?!

Traumatic head injury – Not a complete overlap with stroke7

Traumatic Head Injury Cost The CI Trends Working Party will be commenting on

this in their final version of “The Critical Paper” paper Our view is that the cost will be higher for younger ages

and males where most THI occurs Thought needs to be given to whether it is included in

Children’s CI cover A rough estimate of the cost from HES data, taking into

account overlap with TPD, is significant single digits Companies have the option to not cover THI

Issues Recent views on CI guarantees How different reinsurers viewpoints have changed Impact of new definitions and new diseases Base level of morbidity including selection factors Recent trends in key diseases & future medical

advances Changes to the ABI claims matrix

Base morbidity - relationshipsThese are derived from our observations from quotes that we’ve done

Higher sums assured heavier experience => an amounts loading as opposed to a discount => too much NML drift?

Tied/Bancassurer business is on par with IFA business provided the same risk management practices apply

IFA experience more homogeneous than for mortality Reviewable business worse than guaranteed Experience is linked to sales volumes - better experience with

higher volume: stronger risk management as offices believe they can pick and choose?

Base morbidity - selection The key risk in interpreting experience is what table to use? Gen Re have produced a table incorporating a 3 year select effect:

but the difference between duration 2 and 3+ is only about 1% There is a clear selection effect: not only from CMI data but from

quote data we’ve seen However, data is not homogeneous:

Covers a variety of underwriting years With changing risk management practices And different definitions

The select effect may not therefore be as steep as derived from CMI or company data once adjustments are made for the above

Issues Recent views on CI guarantees How different reinsurers viewpoints have changed Impact of new definitions and new diseases Base level of morbidity including selection factors Recent trends in key diseases & future medical

advances Changes to the ABI claims matrix

Trends in major CI conditions Cancer Heart Attack Stroke

Cancer Trends - Males

Flat trend up to 2004. Melanoma increases balanced by others. 2005 jump in most cancers including melanoma and prostate

Male Cancer Incidence as % of 1998

80%

85%

90%

95%

100%

105%

110%

115%

1998 1999 2000 2001 2002 2003 2004 2005

20-29

30-39

40-49

50-60

Cancer Trends - Females

Flat trend in 1998-2003 with jump in 2004-05. Recent increases due to melanomas and ovarian cancer. Cervical cancer has shown improvements to counteract some of this.

Female Cancer Incidence as % of 1998

80%

85%

90%

95%

100%

105%

110%

115%

1998 1999 2000 2001 2002 2003 2004 2005

20-29

30-39

40-49

50-60

Cancer trends: what the experts say Many cancer registries are having a go at projecting future

trends using age-cohort and age-period models Scottish trend projected to be relatively flat with only a slight

deterioration in the next 5 years9

North West Cancer Intelligence Service projects a 1% p.a. deterioration for the next 15 years in the region

Thames Cancer Registry shows trends by individual cancer site10

Researchers at KCL predict little change in age-standardised incidence rates in England12

Irish trend extrapolated to be circa 0.9% p.a. deterioration11

Breast cancer scanning

In-situ breast cancers as % of non in-situ

0%

5%

10%

15%

20%

1998 1999 2000 2001 2002 2003 2004

50-54

55-59

60-64

Scanning appears to not have changed the distribution of cancer by stage. But it has picked up more carcinoma in situ, which is not covered

Breast Cancer Trends in Stage Distribution

Malignant breast cancer trends have been fairly flat over the last few years, so what impact has scanning had?

Melanoma and Cheap Flights

Female Melonoma Incidence as % of 1998 compared with EasyJet passenger numbers

80%

100%

120%

140%

160%

180%

1998 1999 2000 2001 2002 2003 2004 2005

% in

cre

as

e in

inc

ide

nc

e

0

5

10

15

20

25

30

35

pa

ss

en

ge

rs in

mill

ion

s

20-29 30-39 40-49 50-65 Passengers

I’ve been slightly misleading as melanoma trends have been bad for some time – sun exposure many years ago can do the damage

Heart attack trends HES data has shortcomings so trends have been

corroborated with Scottish data (which has different shortcomings!)

Scottish data shows continuing strong improvements at older ages… but a level trend at younger ages

English data shows a leveling off of rates at older ages and an increase in rates at younger ages

Heart attack trends

English MI incidence rates as % of 1998

50%60%

70%80%90%

100%

110%120%130%

140%150%

1997 1998 1999 2000 2001 2002 2003

Males 0-44

Females 0-44

Males 45-64

Females 45-64

Scotttish MI incidence rates as % of 1997

50%60%70%80%90%

100%110%120%130%140%150%

Males 0-44

Females 0-44

Males 45-64

Females 45-64

Heart attack trends Flattening of improvements for postulated to be due to:

reductions in smoking cessation4,6, increased obesity and diabetes4,6, higher resting heart rates in young adults3

Interestingly levels of physical activity have not changed much over the period suggesting that diet and lifestyle are more to blame4

Troponins are not mentioned in the literature as cause for the increase

BMI trends4

% of Males with normal BMI

0

10

20

30

40

50

60

70

80

%

Males 16-24

Males 25-34

Males 35-44

Males 45-54

Males 55-64

Worst trends for those aged 25-34 and 35-44

Emotional upset and heart attacks 30 June ’98 semi-finals of the World Cup:

England lost to Argentina. 25% more heart attacks on that day and in the 2 days following5

Increase in admissions suggests that MI can be triggered by emotional upset, such as watching your football team lose an important match

With England not in Euro 2008 it should be a good year for heart attacks!

Seriously…. With the credit crunch this is something to

watch… A Cambridge study suggests that a system-

wide banking crisis increases population heart disease mortality rates by 6.4% (95% CI: 2.5% to 10.2%, p < 0.01) in high income countries8

The effect could be 4 times worse in lower income countries

Stroke Trends

English stroke incidence rates as % of 1998

50%

60%

70%

80%

90%

100%

110%

120%

130%

140%

150%

1997 1998 1999 2000 2001 2002 2003

Males 20-44

Females 20-44

Males 45-64

Females 45-64

Scottish stroke incidence rates as % of 1997

50%60%70%80%90%

100%110%120%130%140%150%

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

p

Males 0-44

Females 0-44

Males 45-64

Females 45-64

Stroke trends commentary

Scottish data excludes incidence where a patient has had a stroke in the last 10 years. English data includes all strokes

Both show younger ages having lower improvement rates than older ages

Reasons are as per MI – obesity, smoking, general health

Is this cohort more unhealthy?

Trends summary and future outlook Trends for cancer relatively flat Trends for MI increasing for younger ages and

flat or reducing slightly for older ages Stroke trends indeterminate Off the shelf testing a big risk for underwriting

and claims in the future Tests for cancer Family history Genetics Not just a problem for new policies but also anti-

selective lapsation and earlier claims identification

Issues Recent views on CI guarantees How different reinsurers viewpoints have changed Impact of new definitions and new diseases Base level of morbidity including selection factors Recent trends in key diseases & future medical

advances Changes to the ABI claims matrix

ABI TCF changes

Non-linked non-disclosure will now be paid in full

All ratings up to and including +50 will not be classed as deliberate or without care and will attract a proportionate or full payment

Requests for medical records need to be more fully justified

ABI TCF– impact on claims costs

Circa 10% of claims are declined for non-disclosure in the first 5 years. Declinature rates for non-disclosure thereafter are minimal

30-40% of these might no longer be investigated because of the need for more justification of medical evidence requests (+3-4%)

10% might have resulted in exclusions not linked to claim so now paid in full (+1%)

There will be more proportionate payments as below a +50 rating (+1-2%)

Total impact on claims paid will be to increase claims in the first 5 years by 5-7%

ABI TCF– impact on claims costs

Impact will depend on pre-changes claims philosophy and expected future philosophy

Will also depend on other risk management tools, specifically: GPR sampling Tele-underwriting App form and online submission design Expert underwriting Channel/distribution management

Other Topical Issues

PS 06/14: not much changes to %’s reinsured or structure

Additional illnesses: Mastectomy, CJD etc Kiddie CIC: An increasingly common claim

cause Solvency II

References1. Development of acute chest pain services in the UK, Elizabeth Cross, Steven How, Steve

Goodacre, Emerg Med J 2007;24:100–1022. http://www.disability.vic.gov.au/dsonline/dsarticles.nsf/pages/Traumatic_brain_injury_and_se

xual_issues?opendocument/

3. Secular trends in heart rate in young adults, 1949 to 2004: analyses of cross sectional studies, Black, Murray, Cardwell, Davey, Smith, McCarron, Heart 2006;92:468-473

4. Health Survey for England, Department of Health5. Admissions for myocardial infarction and World Cup football: database survey, Carrol et al,

BMJ 2002;325:1439-14426. Coronary heart disease trends in England and Wales from 1984 to 2004: concealed levelling

of mortality rates among young adults, O’Flaherty et al, Heart 2008;94:178-1817. http://discovermagazine.com/2004/dec/lights-out/8. Can a bank crisis break your heart? David Stuckler et al, Globalization and Health 2008 9. Cancer in Scotland: Sustaining Change, Cancer Incidence Projections for Scotland (2001-2020) , The

Scottish Government Statistics 10. Cancer in South East England 2005, Thames Cancer Registry11. Trends in Irish cancer incidence 1994-2002 with predictions to 2020, National Cancer Registry

12. The future burden of cancer in England: incidence and numbers of new patients in 2020. Møller et al, British Journal of Cancer 2007

Contact

Darshan Singh, Head of Marketing Actuarial

darshan.singh@scottishre.com

Alex King, Head of Protection Marketing

alex.king@scottishre.com

Thanks to: Matthew Smith, Warren Copp, Paul Reddick, Dave Heeney, Paul Lewis, Steve Nuttall, Scott Reid, Ian Rowe