experience trends for income protection international congress of actuaries paris, france june 1,...

Experience Trends for Income Protection

International Congress of Actuaries

Paris, France

June 1, 2006

Experience Trends for Income Protection

Panelists

Daniel SkwireMilliman, Inc. (U.S.A.)[email protected]

Edward FabrizioGeneral Reinsurance Life Australia [email protected]

Denis GarandDenis Garand and Associates (Canada)[email protected]

mailto:[email protected]



Experience Trends in the United States

(Individual Disability Insurance)

Daniel D. Skwire, FSA

Principal and Consulting Actuary

Milliman, Inc.

1985 Commissioner’s Individual Disability A Table

(1985 CIDA)

• Most recent published table for individual disability

• Experience from 1970’s and early 1980’s

• Required table for some reserve calculations

• Expected basis for recent intercompany study

Characteristics of 1985 CIDA

• Sex-distinct

• Not smoker-distinct

• Four occupation classes– Class 1: Executive and Professional– Class 2: Clerical and Office– Class 3: Light manual duties– Class 4: Heavy manual duties

• Accident, sickness, and combined causes

Society of ActuariesIndividual Disability Experience Committee (IDEC)

1990-1999 Intercompany Study

• Twelve contributing companies

• About 64% of total inforce premium

• Incidence Study: Claims incurred 1990-1999

• Claim Termination Study: Claims paid 1990-1999

• Results show Actual-to-Expected (A/E) Ratios relative to 1985 CIDA

1990 – 1999 IDEC Study

A/E Claim Incidence Ratios - A&S ContractsExpected = 85 CIDA

40%

50%

60%

70%

80%

90%

100%

110%

120%

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Calendar Year

Rat

ios

By Count

By Amount


A/E Claim Incidence Ratios (by Amount) - A&S ContractsExpected = 85 CIDA

40%

50%60%

70%80%

90%100%

110%120%

130%

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Calendar Year

Rat

ios

Occ Class 1

Occ Classes 2-4


A/E Claim Incidence Ratios (by Amount)A&S Contracts - Occupation Class 1

Expected = 85 CIDA

40%

60%

80%

100%

120%

140%

160%

1990 1991 1992 1993 1994 1995 1996 1997 1998 1999

Calendar Year

Rat

ios

Medical Occs

Non-medical Occs

Claim Incidence Rates Trends by Elimination Period and Gender

0%

25%

50%

75%

100%

125%

150%

Under 30 30 60 90 180

Elimination Period

Rat

ios

Male

Female



A/E Claim Incidence (by Amount)By Benefit Period - Expected = 85 CIDA

40%

60%

80%

100%

120%

140%

160%

Occ Class 1 Occ Class 2-4

85 CIDA Occupation Class

Ratios Short-term

To Age 65-70

Lifetime

1990 – 1999 IDEC StudyClaim Incidence Rates

Non-Medical Occupations (ranked by exposure)

Occupation A/E Ratio

Executives & Managers 61%

Lawyers 63%

Other Sales 65%

Accountants 62%

Engineers 57%

Insurance Sales 101%

Stockbrokers 142%

Teachers 57%

Other Occupations 64%

Total Non-Medical Occupations 64%

1990 – 1999 IDEC StudyClaim Incidence Rates

Medical Occupations (ranked by exposure)Occupation A/E Ratio

Physicians & Surgeons 87%

Dentists 62%

Psychologists 53%

Nurses 100%

Pharmacists 51%

Chiropractors 139%

Veterinarians 60%

Podiatrists 73%

Other Medical Occupations 80%

Total Medical Occupations 88%

1990 – 1999 IDEC StudyClaim Termination Rates

• 1985 CIDA known to have outdated claim termination rates

• Regulators adopted 1985 CIDC table as temporary measure

• 1985 CIDC consists of adjustments to 1985 CIDA

Age of Claim AdjustmentMonths 1-6 36-44%Months 7-12 50-75%Months 13-18 80-102%Months 19-24 105-120%Year 3 137%Year 4 120%Year 5 120%Year 6+ 100%

• IDEC Study measures experience relative to 1985 CIDA


A/E Claim Termination Ratios (by Amount)By 85 CIDA Occupation Class

Expected = 85 CIDA

0%20%40%60%80%

100%120%140%160%180%

Duration of Disablement

Rat

ios

Occ Cl 1

Occ Cl 2-4


A/E Claim Termination Ratios (by Amount)A&S Contracts - To Age 65-70 and Lifetime BP - Occ Class 1

Medical vs. Non-medical Occupations

0%

20%

40%

60%

80%

100%

120%

140%

Yr 1 Yr 2 Yr 3 Yr 4-5 Yr 6-10 Yr 11+


Rat

ios

Medical Occs

Non-med Occs


A/E Claim Termination Rates (by Amount)A&S Contracts By Benefit Period

Expected = 85 CIDA

0%20%40%60%80%

100%120%140%160%180%200%

Yr 1 Yr 2 Yr 3 Yr 4-5 Yr 6-10 Yr 11+


Rat

ios

Short-term

To Age 65-70

Lifetime

1990-1999 IDEC StudyKey Conclusions

• Medical occupations have distinct experience from Class 1

• Lifetime benefit periods have poor experience

• Incidence trends are favorable in aggregate, with lots of variation

• Claim termination rates are generally lower (less favorable) than those in 1985 CIDA

Australia & NZ

Disability Income Experience

Edward Fabrizio

Deputy General Manager and Chief Actuary

General Reinsurance Life Australia Ltd

Australian & NZ Market

• Benefit is a monthly annuity style benefit whilst person remains disabled

• Benefit based on income at time of underwriting (generally to max of 75% replacement) and is taxable

• Majority (90%) is annual renewal business (age rated based on age at renewal)

• Non-guaranteed rates

• Full offset if person earns income whilst disabled

• Agreed Value and Indemnity benefits are available

Australian & NZ Market

• Most common deferment periods are 30 and 90 day

• Benefit period is generally to Age 65, but 2 and 5 year periods also common for blue collar workers

• Many extras added :

– Hospital benefit (pays benefit whilst in hospital during deferment period)

– Rehabilitation benefits

– Payment of minimum benefit period on certain accidents / illnesses

IAD 89-93 Standard Table

• Experience Reports for Australia and New Zealand are done on basis of comparing actual results to a standard table – IAD89-93

• IAD89-93 developed by the IAAust based on the Australian disability income experience for this period

• Incidence rates graduated by age/sex and 4 occupation classes

• Termination rates based on % of the US CIDA85 termination rate tables

IAD 89-93 Standard Table

• 4 Occupation classes

– A : Professional and white collar

– B : Other sedentary (light manual technical)

– C : Light Manual (qualified trades persons)

– D : Moderate to Heavy Manual

Australian Experience

• WARNING

Experience results presented for Australia are preliminary only at this stage and are still going through a checking process which may alter the final results

Australia – Latest Experience

Experience – TrendsClaim cost in first 3 years of claim : 1 mth deferment

100

110

120

130

140

150

160

89-93 91-94 92-95 93-96 95-98 98-01

MaleFemale


Incidence Rates - % of IAD89-93Preliminary results for continuing companies only

MalesOcc Class

95-98 98-01 95-98 98-01

A 100% 73% 105% 78%B 94% 81% 101% 97%C 96% 83% 109% 105%D 94% 73% 113% 108%

All 96% 78% 108% 97%

2 week deferment 1 month Deferment

Incidence Experience - Males


Incidence Experience - Females

Incidence Rates - % of IAD89-93Preliminary results for continuing companies only

FemalesOcc Class

95-98 98-01 95-98 98-01

A 96% 72% 97% 73%B 101% 96% 97% 79%C 93% 83% 93% 78%D 90% 73% 84% 53%

All 95% 74% 96% 74%

2 week deferment 1 month Deferment


Occupation RelativitiesPreliminary results for continuing companies only

Weighted Average Incidence Rates as % Male Class A

male B 150% (156%) 175% (156%)male C 238% (243%) 379% (320%)male D 237% (273%) 438% (379%)

Weighted Average Incidence Rates as % Female Class A

female B 154% (138%) 136% (134%)female C 141% (164%) 189% (157%)female D 146% (176%) 142% (174%)

2 weeks 1 month

2 weeks 1 month


Amount vs Number of Claims

Preliminary results for continuing companies only

Ratio of A/E amounts versus A/E number for Incidence of Claim

Occ Class MALE FEMALE

A 108% 108%

B 104% 104%

C 109% 112%

D 113% 117%


Termination Rates


Ratio of A/E Number of Terminations


95-98 98-01 95-98 98-01

A 81% 81% 77% 80%

B 73% 77% 72% 74%

C 84% 88% 77% 77%

D 88% 84% 72% 64%


Termination Rates – Male: Occ A: 1 mth deferment

0

100

200

300

400

500

600

700

800

900

1000

4 wks

5 wks

6 wks

7 wks

8 wks

9 wks

10 w

ks

11 w

ks

12 w

ks

3 m

ths

4 m

ths

5 m

ths

6 m

ths

7 m

ths

8 m

ths

9 m

ths

10 m

ths

11 m

ths

12 m

ths

13 m

ths

14 m

ths

15 m

ths

16 m

ths

17 m

ths

18 m

ths

19 m

ths

20 m

ths

21 m

ths

22 m

ths

23 m

ths

24 m

ths

36 m

ths

95-9898-01


Termination Rates – Female: Occ A: 1 mth deferment

0

100

200

300

400

500

600

700

800

900

1000

4 wks

5 wks

6 wks

7 wks

8 wks

9 wks

10 w

ks

11 w

ks

12 w

ks

3 m

ths

4 m

ths

5 m

ths

6 m

ths

7 m

ths

8 m

ths

9 m

ths

10 m

ths

11 m

ths

12 m

ths

13 m

ths

14 m

ths

15 m

ths

16 m

ths

17 m

ths

18 m

ths

19 m

ths

20 m

ths

21 m

ths

22 m

ths

23 m

ths

24 m

ths

36 m

ths

95-9898-01


Termination Rates

95-98 98-01 RatioTermination Rate Male Female Male Female Male Female

Mths 1 to 3 0.491 0.490 0.468 0.475 95% 97%Mths 4 to 6 0.326 0.327 0.318 0.307 97% 94%Mths 7 to 12 0.341 0.315 0.303 0.299 89% 95%Mths 13 to 24 0.372 0.353 0.494 0.549 133% 155%Mths 25 to 36 0.197 0.243 0.313 0.383 158% 157%


Actual Termination Rates of Claim by duration of claim


Amount vs Number of Terminations


Ratio of A/E amounts versus A/E number for Termination Rates of Claim


A 96% 98%

B 95% 95%

C 95% 95%

D 93% 97%


Variation in Experience by Company

Preliminary results

50%

60%

70%

80%

90%

100%

40% 50% 60% 70% 80% 90% 100% 110% 120%

A/E Incidence

A/E

Ter

min

atio

ns

Summary for Australia

• Experience has improved significantly– Driven by lower incidence experience (although there are

concerns over quality of data used in the investigation)

• Termination rates during first year of claim continue to deteriorate, however, this is being compensated by better termination rates at later durations

• Incidence and termination rates by amounts are worse than by number (total claims cost by amounts about 15% higher than by number), need to be careful in setting pricing and reserving assumptions – especially for reinsurers on a surplus basis

New Zealand – Latest experience

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

1997 1998 1999 2000 2001 2002

Event Year

Act

ual

vs.

Expec

ted

Male A/ E Female A/ E

Incidence Rate Experience


Termination Rate Experience

0%

20%

40%

60%

80%

100%

120%

1 2 3 4 5 6 7 8 9 101112 1314151617 18192021 222324

Month of disablement

2000- 02 1997- 99


Claim Duration Experience

Deferment Period 1997- 99 2000- 02

30- days 169 146

90- days 146 132

Occupation Class 1997- 99 2000- 02

White- collar 168 141

Blue- collar 154 140

Ratio (%) of actual to expected claim length by deferment period and investigation period

Ratio (%) of actual to expected claim length by occupation class and investigation period


• Important feature of NZ Experience is the interaction with the government accident compensation scheme (ACC)

• ACC provides benefit up to 80% of taxable salary (with maximum dollar cap) in event a person is injured and unable to work

• Covers most forms of accidental injury (work and outside work) and can sometimes overlap with what would normally be considered a sickness

Paid ratio (%) by cause of claim and investigation period

Paid ratio by claim cause and occupation class in 2000-02

Cause of Claim 1997- 99 2000- 02

Accident 52 43

Sickness 79 81

All 72 70

Occupation Class Accident Sickness Total

A 41 83 76

B 43 82 74

C 57 76 71

D 43 80 63

E 44 77 60



- 2%

- 10%

- 3%

- 14%

- 11%

- 24%

1%

- 31%- 35%

- 30%

- 25%

- 20%

- 15%

- 10%

- 5%

0%

5%

Incidence Duration Paid Ratio Cost

Chan

ge

in A

/E R

esult

Male A/ E Female A/ E

Contribution of each component to total change in claim cost from 1997-99 to 2000-02


0%

20%

40%

60%

80%

100%

120%

140%

160%


Numbers

Amounts

Comparison of ratios (%) of actual to expected results in 2000-02

weighted by numbers and amounts


Warning: smoking hazardous to your health

Occupation Class Male Female All

White- collar 142 (13) 135 (16) 138 (10)

Blue- collar 123 (10) 117 (31) 122 (10)

All Occupation

Classes

122 (7) 129 (14) 124 (6)

Ratio (%) of actual to expected incidence for smokers over non-smokers in 2000-02

Ratio (%) of actual to expected claim cost for smokers over non-smokers in 2000-02

Occupation Class Male Female All

White- collar 143 123 134

Blue- collar 132 123 131

All Occupation

Classes

124 121 123


111%95%

106% 113%

0%

20%

40%

60%

80%

100%

120%


Agreed Value benefits versus Indemnity


Self-employed versus Employed

152%

102%80%

126%

0%

20%

40%

60%

80%

100%

120%

140%

160%



86

50

69

99

5942 67

65

20%

30%

40%

50%

60%

70%

80%

90%

100%

110%

120%

130%

140%

80% 90% 100% 110% 120% 130% 140% 150% 160% 170% 180% 190% 200%

Duration

Incidence

Variation in experience by company

Bubbles show claim cost

Summary for New Zealand

• Experience has improved– Driven by increased termination rates

• Difference between self-employed and employee can be an important rating variable – is just as important as smoker status

• Incidence and termination rates by amounts are worse than by number, need to be careful in setting pricing and reserving assumptions – especially for reinsurers on a surplus basis

• Experience varies greatly by company

Denis Garand

Denis Garand & Associates

Experience Trends in Canada

Fraser Garand 2000-2004 Study

• Proprietary study • Financial and technical support from:

– Munich Re– Optimum Re– RGA Re – SCOR Re– Swiss Re

• 13 companies– Small group market, average size 41 employees– 2,600,000 exposure from 2000 to 2004– 26,000 claims of which 18,000 terminate in first 2

years– one province Quebec, overrepresented

• These companies represent:– 34% of the market, groups less than 1000 employees

Demographic data

Female Male

Average age 39.7 40.1

Distribution in % 40% 60%

• Growth of portfolio higher in earlier years of the study

• 72% 119 day elimination period• 91% initial own occ definition

Fraser Garand 2000-2004 Table

• 4 month elimination

• Own Occupation

• Benefit to age 65

• Tables by 5 year age band and gender– Quebec,– Canada except Quebec

Incidence per 1000, study population

Canada 8.4

Canada, except Quebec 6.2

Quebec 15.1

SOA GLTD 1987 Basic “4 months” expected

4.6

Variation Actual /Expected (F-G)

• By age (compare to SOA 87)– Younger female higher incidence– Older males lower incidence

• By amount insured– Lower benefits higher incidence – Higher benefits lower incidence

• By industry (50% to 150%)

Variation Actual /Expected (F-G)

• By region– Some provinces better (no control on industry)

• By cause of disability– Variations by Mental and Nervous,

Musculoskeletal and all other causes

• By year of disability– 2000 highest, 2003 best

• By participating company (~70% to 125%)

Termination study

• 18,000 terminations within the first 24 months of duration.

• Produced termination rates for duration 4-12 months. – 159% versus SOA87 Basic 3 month

• Analysis of claims duration to 24 months by three major causes of disability.– Substantial variation in termination rates.

Caveats

• Study not adjusted for industry.

• Short period of study was in a favourable economic cycle.

• Termination study requires more data

Uses of study

• Improve pricing for current environment (most companies operated on “old” data)

• Adjust valuation reserves for key parameters of region and cause of disability.

• Benchmark company versus industry

experience trends for income protection international congress of actuaries paris, france june 1,...

Documents

cida slide

idec study slide

combined causes slide

claim termination study

idec study measures

experience trends

distinct experience

cida sexdistinct