Anaheim - Morbidity Improvement and Its Impact on … Morbidity Improvement and Its Impact on LTC...

1

Morbidity Improvement and Its Impact on LTC Insurance Pricing

and ValuationEric Stallard, A.S.A., M.A.A.A., F.C.A.

Research Professor of Demographic StudiesDuke University

[email protected]

Society of Actuaries Anaheim Spring Meeting, Session 107 PD

Anaheim, CAMay 21, 2004

Objectives For the entire session –

• To quantify the nature of temporal trends in functional disability rates

• To illustrate the potential impact of these temporal trends on LTCI premiums and reserves

• To provide guidance in properly reflecting this information in LTCI pricing and valuation

SOA 2004 Anaheim Spring Meeting - 107 PD, Morbidity Improvement and Its Impact on LTC Insurance Pricing and Valuation

2

ObjectivesFor my presentation –• To examine long-term trends in morbidity,

disability, and mortality rates among the elderly

• To examine short-term trends in disability and mortality rates among the elderly

• To examine the joint impact of disability and mortality rates in producing lifetime disability time (beyond age 65)

Objectives

For Ron Wolf’s presentation –• To present practical considerations

concerning the use of temporal trends in disability and mortality rates on– LTCI pricing

– LTCI valuation


3

Objectives

For Scott Weltz’s presentation –• To discuss the current methodologies and

assumptions used to estimate LTCI morbidity (with and without reflecting potential improvements)

• To illustrate the impact that morbidity improvement has on projected LTCI morbidity estimates

Four Fundamental Issues

1. What changes have been documented in morbidity and disability rates?

2. Why are these changes occurring?

3. Will they continue?

4. What do these changes imply for LTCI pricing and valuation?


4

Data for Long-Term Trends1900 & 1910 Medical Examinations from the Union

Army Pension Program– Covered 85% of UA veterans in 1900 and 90% in 1910

1985–1988 National Health Interview Survey (NHIS)– Random sample of noninstitutionalized male veterans

1988–1994 National Health and Nutritional Examination Survey (NHANES)– Random sample of noninstitutionalized white males

1960-2002: Various Reports from National Center for Health Statistics (NCHS)

Gender 1900 1910 1990 1999

Annual Rate of Decline

Male 10,612 6,154 0.55%Female 9,749 4,157 0.86%Total 10,079 4,898 0.73%

Male 10,444 6,526 0.59%Female 9,606 4,055 1.07%Total 9,937 4,986 0.86%

Source: Bell and Miller (2002, Table 1).

Table 1: Age-Adjusted Central Death Rates at Ages 65 Years and Older (per 100,000)

Year


5

ConditionUnion Army

1910

NHIS 1985-1988


Digestive (Hernia/Diarrhea) 84.0 8.0 3.03%Genito-Urinary 27.3 8.9 1.45%Circulatory 90.1 40.0 1.06%CNS, Endocrine, Metabolic, or 24.2 12.6 0.85%

Blood DisordersMusculoskeletal 67.7 42.5 0.61%Respiratory 42.2 26.5 0.61%Cancer 2.2 9.2 -1.89%

Source: Fogel and Costa (1997, Table 3).

Table 2: Prevalence of Chronic Conditions Among Elderly Male Veterans Aged 65 Years and Older (%)

ConditionUnion Army

1910

NHANES 1985-1988


Heart murmur 39.2 3.8 3.00%Irregular pulse 42.0 8.5 2.07%Decreased breath or adventitious 37.8 10.8 1.62%

soundsJoint pain/tenderness/swelling 55.0 35.2 0.58%

Source: Costa (2000, Table 1).

Table 3: Prevalence of Chronic Conditions Among Elderly Male Veterans Aged 60-74 Years (%)


6

Age ConditionUnion Army 1900/19101

NHANES 1988-1994


50-64 in 1900 or 1988-1994Paralysis 4.8 0.9 1.82%Difficulty bending 39.0 7.5 1.80%Deaf (either/both ears) 2.9 1.4 0.80%Difficulty walking 20.9 10.4 0.76%Blind (either/both eyes) 2.8 1.5 0.68%

60-74 in 1910 or 1988-1994Difficulty bending 49.7 16.1 1.38%Difficulty walking 30.9 13.8 0.99%Paralysis 6.0 2.7 0.98%Deaf (either/both ears) 3.8 2.7 0.42%Blind (either/both eyes) 3.8 3.1 0.25%

Note 1: Excludes wounded veterans, POW s, and disability discharges.Source: Costa (2002, Table 3).

Table 4: Prevalence of Functional Limitations Among Elderly Males (%)

Long-Term Trends• Except for cancer, chronic disease

prevalence declined at least as fast as mortality rates (~0.6% per year) for older males during 1910-1990 (Fogel and Costa, 1997)

• Except for sensory losses, functional limitation prevalence declined at least as fast as mortality rates for older males during 1910-1990 (Costa 2000, 2002)


7

Quantitative Estimates 1900-1990

• 50% of decline in mortality predicted by changes in height and weight (Fogeland Costa, 1997)

• 29% of decline in chronic diseases due to occupational shifts (Costa, 2000)

• 18% of decline in chronic diseases due to declines in infectious diseases (Costa, 2000)

Quantitative Estimates 1900-1990

• 37% of decline in functional limitation due to reduced prevalence of chronic diseases (Costa, 2002)

• 24% of decline in functional limitation due to reduced debilitating effects of chronic diseases (Costa, 2002)


8

Year Unisex Males Females

1960 14.3 12.8 15.81970 15.2 13.1 17.01980 16.4 14.1 18.31990 17.2 15.1 18.92000 17.9 16.3 19.22002 18.2 16.6 19.5

Rate (% per yr.; 42 yr.) 0.58% 0.62% 0.50%Rate (% per yr.; 22 yr.) 0.47% 0.74% 0.29%

1980 10.4 8.8 11.51990 10.9 9.4 12.02000 11.3 10.1 12.12002 11.6 10.4 12.5

Rate (% per yr.; 22 yr.) 0.50% 0.76% 0.38%

Source: NCHS (2003, Table 27); Kochanek et al. (2004, Table 6).

At Age 65

At Age 75

Table 5: Life Expectancy at Ages 65 and 75, United States, Select Years

Age 1960 2000


65-74 3822 2399 1.16%75-84 8745 5667 1.08%85+ 19,858 15,524 0.61%

Source: NCHS (2003, Table 35).

Table 6: Unisex Death Rates (per 100,000) for All Causes, United States


9

AgeDeath

Rate

Percent of All

DeathsDeath

Rate

Percent of All

Deaths

Annual Rate of Decline in Death Rate

65-74 1741 46% 666 28% 2.37%75-84 4089 47% 1780 31% 2.06%85+ 9318 47% 5926 38% 1.13%


Table 7: Unisex Death Rates (per 100,000) for Diseases of Heart, United States

1960 2000

AgeDeath

Rate

Percent of All

DeathsDeath

Rate

Percent of All

Deaths


65-74 714 19% 816 34% -0.34%75-84 1127 13% 1336 24% -0.42%85+ 1450 7% 1819 12% -0.57%


Table 8: Unisex Death Rates (per 100,000) for Cancer, United States

1960 2000


10

AgeDeath

Rate

Percent of All

DeathsDeath

Rate

Percent of All

Deaths


65-74 469 12% 129 5% 3.18%75-84 1491 17% 461 8% 2.89%85+ 3681 19% 1589 10% 2.08%


Table 9: Unisex Death Rates (per 100,000) for Cerebrovascular Diseases, United States

1960 2000

AgeDeath

Rate

Percent of All

DeathsDeath

Rate

Percent of All

Deaths


65-74 2924 76% 1611 67% 1.48%75-84 6708 77% 3577 63% 1.56%85+ 14,448 73% 9335 60% 1.09%

Source: NCHS (2003, Tables 36-38).

Table 10: Unisex Death Rates (per 100,000) for Heart, Cancer, and Cerebrovascular Diseases, United States

1960 2000


11

AgeDeath

Rate

Percent of All

DeathsDeath

Rate

Percent of All

Deaths


65-74 899 24% 789 33% 0.33%75-84 2037 23% 2089 37% -0.06%85+ 5409 27% 6190 40% -0.34%

Source: NCHS (2003, Tables 33-38).

Table 11: Unisex Death Rates (per 100,000) for Residual Causes of Death, United States

1960 2000

Mortality Changes 1960-2000

• Increases in life expectancy at age 65 due to decreases in death rates from heart and cerebrovascular diseases

• Cancer death rates increased, but not enough to offset heart and cerebrovascular decreases


12

Mortality Changes 1960-2000

• Cancer death rates peaked for males during 1990s (female peak expected 10-20 years after)

• Residual causes at age 85+ increased from 27% to 40% of deaths as progress against top 3 causes left “void”

Gender 1965 2000


Male 51.9 26.4 1.91%Female 32.0 21.6 1.12%

Male 28.5 10.2 2.89%Female 9.6 9.3 0.09%


Table 12: Current Cigarette Smoking (%) at Ages 45-64 and 65 Years and Older, United States

Age 45-64

Age 65+


13

Gender 1960-1962 1999-2000


Male 60.3 50.7 0.45%Female 66.4 57.9 0.36%

Male 68.8 68.3 0.02%Female 81.5 73.4 0.27%


Table 13: Prevalence of Hypertension (%; SBP >= 140, DBP >= 90, or Medicated) at Ages 55-64 and 65-

74 Years, United States

Age 55-64

Age 65-74

Gender 1960-1962 1999-2000


Male 233 210 0.27%Female 262 223 0.42%

Male 230 210 0.24%Female 266 229 0.39%


Table 14: Mean Serum Cholesterol Levels (mg/dL) at Ages 55-64 and 65-74 Years, United States

Age 55-64

Age 65-74


14

Gender 1960-1962 1999-2000


Male 41.6 16.5 2.37%Female 70.1 26.2 2.52%

Male 38.0 19.2 1.76%Female 68.5 37.4 1.56%


Table 15: Prevalence of High Serum Cholesterol (%; SC >= 240 mg/dL) at Ages 55-64 and 65-74 Years,

United States

Age 55-64

Age 65-74

Gender 1960-1962 1999-2000


Male 9.2 32.9 -3.37%Female 24.4 43.1 -1.49%

Male 10.4 33.4 -3.08%Female 23.2 38.8 -1.34%


Table 16: Prevalence of Obesity (%; BMI >= 30 kg/m2) at Ages 55-64 and 65-74 Years, United States

Age 55-64

Age 65-74


15

Chronic Disease Risk Factor Changes 1960-2000

• Substantial declines in – cigarette smoking

– hypertension

– serum cholesterol

• Substantial increases in obesity, but not enough to offset favorable declines in other risk factors

Data for Short-Term Trends1984, 1989, 1994, and 1999 National Long Term Care Survey (NLTCS)– Nationally representative sample of elderly

Medicare enrollees aged 65+– Combines cross-sectional and longitudinal

design with stable instrumentation (Stallard, 2000; Stallard and Yee, 2000)

– “One of the best designed surveys for analyzing national disability trends” (Freedman et al., 2002)

– See http://nltcs.cds.duke.edu/index.htm


16

Age 1984 1989 1994 1999Annual Rate of Decline; 15 yr.

65-69 3.32 3.15 3.04 2.38 2.19%70-74 5.15 4.64 4.11 4.12 1.48%75-79 8.83 8.42 7.90 6.32 2.21%80-84 15.95 15.87 13.30 12.61 1.55%85-89 27.86 27.97 25.89 22.50 1.42%90-94 46.89 42.52 45.30 39.04 1.21%95-99 66.43 61.90 60.15 52.11 1.61%Age standardized rate 9.59 9.19 8.49 7.48 1.64%

Source: Author's calculations based on NLTCS.

Table 17: Unisex Prevalence (%) of Disability Satisfying HIPAA ADL Trigger, United States 1984 to 1999, Select Years

Year

Age 1984 1989 1994 1999Annual Rate of Decline; 15 yr.

65-69 1.07 1.01 0.85 0.52 4.67%70-74 2.16 2.09 1.84 1.64 1.84%75-79 4.48 4.37 4.25 2.76 3.18%80-84 9.85 9.43 8.45 6.35 2.88%85-89 19.31 18.95 18.41 12.49 2.86%90-94 34.13 31.38 32.91 27.81 1.36%95-99 50.14 44.60 49.84 37.24 1.96%Age standardized rate 5.47 5.25 5.01 3.71 2.56%

Source: Author's calculations based on NLTCS.

Table 18: Unisex Prevalence (%) of LTC Institutionalization, United States 1984 to 1999, Select Years

Year


17

Why is Disability Declining?Contributing Factors --• Improvements in physical health• Improvements in cognitive health• Improvements in diagnosis and

treatment of chronic and disabling illnesses

• Innovations in preventive medicine• Pharmaceutical innovation

Why is Disability Declining?Contributing Factors --• Improvements in diagnosis and

treatment of mental health disorders

• Improvements in health-related behaviors

• Decreases in hazardous exposures• Improved technology of assistive

devices


18

Why is Disability Declining?Contributing Factors --• Expanded use of assistive devices

and environmental supports• Changes in reimbursement for home

health care• Expanded elder care and social

support• Improved levels of education• Improved socioeconomic status

Quantitative Estimates 1984-1993There were minimal impacts of population composition, device use, survey design, role expectations and living environments on functional limitations 1984-1993– Trend may be due to change in “underlying

physiological capability” (Freedman and Martin, 1998)

– Or to improved educational attainment (Freedman and Martin, 1999)


19

Quantitative Estimates 1989-1996• Disability declined 1989-1994

– not due to decreases in disease incidence (via Medicare reports; which actually increased),

– but to reduced debilitating effects of disease (McClellan and Yan-Li, 2000)

• These results were replicated for self-reported conditions 1992-1996 (Yan-Li and McClellan, 2001)

Expected Lifetime Disability-Time Beyond Age x in Year y (Sullivan, 1971)

, , ,0

, , ,

,

where

and disability prevalence at age

Dx y t x y x t y

t x y x t y x y

x t y

e p dt

p l l

x t

π

π

∞

+

+

+

=

=

= +

∫


20

At Age 65 1984 1999Annual Rate

of Change

Life Expectancy 16.64 17.40 0.30%

HIPAA ADL Expectancy 1.86 1.57 1.13%

LTC Institutional Expectancy 1.11 0.83 1.94%

Table 19: Unisex Life Expectancy, HIPAA ADL Expectancy, and LTC Institutional Expectancy (in Years at Age 65), United

States 1984 and 1999

Year

Source: Author's calculations based on NLTCS and life tables from Bell and Miller (2002).

Change in Expected Lifetime Disability-Time Beyond Age x

( )

( )

( )

0 0 0

0 0

0

, , , , , ,0

, , ,0

, , ,0

Survival Increment

Morbidity Decrement

Dx y Dx y t x y x t y t x y x t y

t x y t x y x t y

t x y x t y x t y

e e p p dt

p p dt

p dt

π π

π

π π

∞

+ +

∞

+

∞

+ +

− = −

= −

− −

∫

∫

∫


21

At Age 65 1984 1999 ChangeSurvival

IncrementMorbidity

Decrement

Life Expectancy 16.64 17.40 0.76 0.76 -

HIPAA ADL Expectancy 1.86 1.57 -0.29 0.13 0.42

LTC Institutional Expectancy 1.11 0.83 -0.28 0.09 0.37

Year

Source: Author's calculations based on NLTCS and life tables from Bell and Miller (2002).

Table 20: Components of Change in Unisex Life Expectancy, HIPAA ADL Expectancy, and LTC Institutional Expectancy (in Years at Age 65), United

States 1984 and 1999

Implications for LTC InsuranceIs there a Dynamic Equilibrium in

Population Health?

• Morbidity, mortality, and disability reflect different aspects of a common, underlying health process; future changes in these three components will be linked.

• Modeling this linkage could yield more accurate projections of the joint impacts of survival increments and morbidity decrements on lifetime disability-time.


22

HIPAA ADL Disability Projections, United States 1995-2040, Unisex Age 65 Years and Older

0

1,000

2,000

3,000

4,000

5,000

6,000

7,000

8,000

1995 2000 2005 2010 2015 2020 2025 2030 2035 2040

Year

Num

bers

in T

hous

ands

Constant prevalence rates0.6% per year prevalence decline1.2% per year prevalence decline

Open Questions

• Can the rate of disability decline in the U.S. continue to be significantly larger than the rate of mortality decline?

• How will biomedical research and health care expenditures affect disability and mortality rates?

• How will trends in these rates impact on LTCI pricing and valuation?


23

ReferencesBell FC, Miller ML. Life Tables for the United States Social

Security Area 1900–2100. Actuarial Study No. 116, Social Security Administration, 2002.

Costa DL. Understanding the twentieth-century decline in chronic conditions among older men. Demography37(1):53–72, 2000.

Costa DL. Changing chronic disease rates and long-term declines in functional limitations among older men. Demography 39(1):119–137, 2002.

Fogel RW, Costa DL. A theory of technophysio evolution, with some implications for forecasting population, health care costs, and pension costs. Demography 34(1): 49-66, 1997.

ReferencesFreedman VA, Martin LG. The role of education in

explaining and forecasting trends in functional limitations among older Americans. Demography 36(4):461–473, 1999.

Freedman VA, Martin LG. Understanding trends in functional limitations among older Americans. American Journal of Public Health 88(10): 1457–1462, 1998.

Freedman VA, Martin LG, Schoeni RF. Recent trends in disability and functioning among older adults in the United States: A systematic review. Journal of the American Medical Association 288(24):3137–3146, 2002.

Kochanek KD, Smith BL. Deaths: Preliminary Data for 2002. National Vital Statistics Reports Vol. 52 No. 13, National Center for Health Statistics, 2004.


24

ReferencesMcClellan M, Yan-Li LL. Understanding disability trends in

the U.S. elderly population: The role of disease management and disease prevention. Paper presented at the Annual Meeting of the Population Association of America, Los Angeles, March 23-25, 2000.

National Center for Health Statistics (NCHS). Health, United States, 2003. NCHS, 2003.

Stallard, E. Retirement and health: Estimates and projections of acute and long-term care needs and expenditures of the U.S. elderly population. Chapter 15 in Retirement Needs Framework, SOA Monograph M-RS00-1, Society of Actuaries, Schaumburg, IL, 2000.

Stallard E, Yee RK. Non-Insured Home and Community-Based Long-Term Care Incidence and Continuance Tables. Actuarial Report Issued by the Long-Term Care Experience Committee, Society of Actuaries, Schaumburg, IL, 2000.

ReferencesSullivan D. A single index of mortality and morbidity. Health

Services and Mental Health Administration (HSMHA) Health Reports 86(4):347–354, NCHS, 1971.

Yan-Li LL, McClellan M. Disease management or disease prevention? Evidence regarding the causes of recent disability decline among American elderly from the Medicare Current Beneficiary Survey. Paper presented at the Annual Meeting of the Population Association of America, Washington, DC, March 31, 2001.


1

Morbidity Improvement and Its Impact on LTC Insurance Pricing and ValuationSOA Spring Meeting – Session 107 PD

Scott A. Weltz, FSA, MAAAMay 21, 2004

Do You Believe?I believe in morbidity improvementBUT.........

Before entertaining the notion that morbidity might improve in the future, we must understand:

– Current insured morbidity experience– Projected morbidity estimates– Then consider population trends


2

Sample Pricing Claim CostsThree carriers – Three assumptions– Uniform plan characteristics and demographics– Similar underwriting criteria– Nationwide rating– Marketed through brokers– Similar claim adjudication procedures

In general, very similar blocks

Sample Pricing Claim CostsPer $10 Daily Benefit

6.2

20.8

100.0

5.0

17.7

89.3

9.6

79.5

30.7

$0

$20

$40

$60

$80

$100

$120

Age 65 Age 75 Age 85

Company ACompany BCompany C


3

Sample Pricing Claim CostsHow can this be?

– Dearth of credible & uniform insured data– Dynamic environment

Plan design changesUnderwriting improvementsClaim management improvementsMix of business shifts

Conclusion– Developing LTC morbidity estimates is difficult

Actual Claims ExperienceIndustry claims study

– $2 billion of insured claims– 4.6 million life years– 12 companies

Developed industry benchmarks

Normalized for risk characteristics


4

Actual Claims ExperiencePer $10 Daily Benefit

5.8

25.3

107.0

5.4

25.9

101.2

4.7

113.7

26.7

$0

$20

$40

$60

$80

$100

$120

Age 65 Age 75 Age 85

Company ACompany BCompany C

Actual Claims ExperienceWith rigorous experience analysis of fully

credible insured data, it is possible to develop reasonable claim cost assumptions today....

....for the early policy durations

Projecting experience is a different story


5

Projecting LTC MorbidityPredominant methods

1. Connect-the-Dots

2. Underwriter’s Dream

3. Freelance

Table I-a: Co nne c t-the -Do ts

0102030405060

65 70 75 80 85 90

Attained Age

LT

C C

laim

Co

st

Actual Issue Age 65Actual Issue Age 75Actual Issue Age 85


6

Table I-b: Co nne c t-the -Do ts

0102030405060

65 70 75 80 85 90 95

Attained Age

LT

C C

laim

Co

st

Actual Issue Age 65Actual Issue Age 75Actual Issue Age 85Projected Ultimate

Table II-a: Unde rwrite r's Dre am

0102030405060

65 70 75 80 85 90 95

Attained Age

LT

C C

laim

Co

st

Actual Issue Age 65Actual Issue Age 75Actual Issue Age 85Pricing


7

Table II-b: Unde rwrite r's Dre am

0102030405060

65 70 75 80 85 90 95

Attained Age

LT

C C

laim

Co

stActual Issue Age 65Actual Issue Age 75Actual Issue Age 85PricingProjected Ultimate

Table II-c : Unde rwrite r's Dre am

0102030405060

65 70 75 80 85 90 95

Attained Age

LT

C C

laim

Co

st

Actual Issue Age 65Actual Issue Age 75Actual Issue Age 85Projected UltimateS i 1


8

Table III: Fre e lanc e

0102030405060

65 70 75 80 85 90 95

Attained Age

LT

C C

laim

Co

st

Actual Issue Age 65Actual Issue Age 75Actual Issue Age 85Projected

Table IV-a: Comparis on of ProjectionsIssue Age 65

0102030405060

65 70 75 80 85 90 95

Attained Age

LT

C C

laim

Co

st

Actual Issue Age 65CTD ProjUW Dream ProjFreelance Proj


9

Table IV-b: Comparis on of ProjectionsIssue Age 65

0102030405060

65 70 75 80 85 90 95

Attained Age

LTC

Cla

im C

ost

Actual Issue Age 65CTD - 1% Morb ImpUW Dream ProjFreelance Proj

Table IV-c: Comparis on of ProjectionsIssue Age 65

0102030405060

65 70 75 80 85 90 95

Attained Age

LTC

Cla

im C

ost

Actual Issue Age 65CTD ProjUW Dream - 1% Morb ImpFreelance Proj


10

Table IV-d: Comparis on of ProjectionsIssue Age 65

0102030405060

65 70 75 80 85 90 95

Attained Age

LTC

Cla

im C

ost

Actual Issue Age 65CTD ProjUW Dream ProjFreelance - 1% Morb Imp

Table V: Generational Morbidity ImprovementFreelance with Morb. Improvement Varying by Issue Age

0102030405060

65 70 75 80 85 90 95

Attained Age

LTC

Cla

im C

ost

IA 65 - 1.5% MI; 15 YearsIA 75 - 1% MI; 5 YearsIA 85 - 0% MI


11

Valuation Survey

Survey of LTC companies in late 2003– 17 individual carriers– 6 group carriers– Over $4.6 billion in-force premium

Morbidity improvement assumptions– Reserve calculations (STAT, TAX, GAAP)– Asset adequacy

Valuation Survey

* Note: Some companies do not hold GAAP reserves or do not value their business on a best estimate basis internally.

200N/A*

1111.50%

2111.25% - 1.49%

5441.00% - 1.24%

711110.00%

GAAP*TaxStatAnnual

Morbidity ImprovementRate

Morbidity Improvement RatesIndividual Companies


12

Valuation Survey

11116 - 20 Years

* Note: Some companies do not hold GAAP reserves or do not value their business on a best estimate basis internally.

200N/A*

322Forever

22211 - 15 Years

2116 - 10 Years

711110 - 5 Years

GAAP*TaxStatYears of Improvement

Applied from Policy Issue Date

Duration of Morbidity ImprovementIndividual Companies

Summary

LTC morbidity is ever-changing

1. Evaluate available data2. Develop well thought-out projections3. Consider morbidity improvement

Population dataTranslating to insured environmentDo not attempt this step in a vacuum from Steps #1 & #2


1

Session 107 PDMorbidity Improvement and Its Impact on LTC Insurance Pricing and Valuation

Use of LTCI Morbidity Improvement -Practical Considerations

Presented byRonald M. Wolf, FSA, MAAA

1

Outline

• Is the use of morbidity improvement material?

• Is it supportable (for an insured population)?

• (When) is it properly used?• Guidelines for consideration


2

2

Materiality

• Consider four alternatives:

1 No No

2 Life No

3 Life* Life*

4 10 y** 10 y**

Alternative Morbidity MortalityPremium For

Same ROIROI For Same

PremiumImprovement In

*1.5% improvement per year**Improvement for10 years, then level for life

3

Materiality (continued)

It is material for pricing

1 No No $100 15%

2 Life No $ 86 23%

3 Life* Life* $ 91 22%

4 10 y** 10 y** $ 99 17%

Alternative Morbidity MortalityPremium For

Same ROIROI For Same

PremiumImprovement In

*1.5% improvement per year**Improvement for 10 years, then level for life


3

4

Materiality (continued)

It can be material for in-force modeling

1 No No $100

2 Life No $240

3 Life Life $164

4 10 y 10 y $110

Alternative Morbidity MortalityPV Pre Tax

Book Profits*

Improvement In

*Discounted at pre-tax earned rate

5

Supportable/Applicable/Appropriate?

• Population improvement implies insured improvement(?)– Seems plausible at face value– True for life insurance mortality

• Supported by insured experience (?)– 1984-1999 SOA study shows improvement in

incidence by issue year and duration– Some carriers have seen improving A/E ratios (some

have seen deterioration)– Difficult to measure due to changes in underwriting

and selection curve– Inherent tendency of insured to anti-select


4

6

Supportable/Applicable/Appropriate? (continued)

• Potential long term societal trends– Fewer family care givers, more single persons– Change in care delivery, contract interpretation– Rate increase policies, limitations– Morbidity/mortality for the very elderly

• Context of other assumptions– Mortality – level and improvement– Voluntary lapse– Current morbidity – good/bad (rate increases); credible– Interest rates (hedging; improving?)

• Nature of task

7

Appropriate Assumptions Differ By Task

Task Nature of Assumptions

Pricing Moderately adverseStatutory ALRs No prescribed table; sound valueGAAP ALRs Best estimate plus marginAsset Adequacy Moderately adverseEV Reporting Best estimateDFCA

Internal Best estimateExternal Moderately adverse

Appraisal “Best foot forward”


5

8

Characteristics of Assumptions

• Best estimate– Equally likely to be high or low

• Best estimate plus margin• Moderately adverse• Sensitivity/stress test

– Plausible/possible– Minimal chance of being worse than…

(historical tie-in)• Lack of directly credible data should imply some

conservatism

9

If Used, How Much Morbidity Improvement May Be Appropriate

• Stallard data indicates morbidity improvement about 1% per year higher than mortality improvement (recently)

• Material difference in effect of lifetime vs. temporary improvement

• Current industry practice varies considerably


6

10

Guidelines For Consideration

• Morbidity improvement should not be used without mortality improvement

• Consider nature of other key assumptions

• Consider nature of the assignment or application

• Lifetime improvement is not moderately adverse


Anaheim - Morbidity Improvement and Its Impact on … Morbidity Improvement and Its Impact on LTC...

Documents

Transcript of Anaheim - Morbidity Improvement and Its Impact on … Morbidity Improvement and Its Impact on LTC...