Society of Actuaries Annual Meeting

Society of Actuaries
Annual Meeting
Disability and Group Life Experience Studies:

What Have We Learned?
Results of IDEC Study

Robert Beal
Consulting Actuary, Milliman, Inc
SOA Study Parameters

• Study Period: 1990-1999
• Approximately 80%+ of industry experience represented
• Results measured relative to 85 CIDA
• Incidence and Terminations studied separately.
• Most results measured in terms of indemnity.
• What’s new?
– Data from all 12 contributors
– Analysis by contract type: A&S, OE, DBO
– Analysis by occupation
– Termination experience by diagnosis
– Total Disability vs. Total & Residual
SOA 2004 New York Annual Meeting - 132OF, Disability and Group Life Experience Studies: What Have We Learned?
1
Major Lessons From the SOA DI Study
1. Significant improvement trend in claim incidence
over the 10 years.
2. Blue/grey occ classes have had much more favorable
experience relative to 85 CIDA than the white
collar/professional occ classes.
3. Substantial claim incidence improvements in business
issued 1996+.
4. Wide disparity of incidence results by occupation,
particularly in Occ Class 1.
5. Excellent incidence from executives/managers,

accountants, engineers and teachers.
6. Poor incidence from physicians, dentists, nurses,
insurance agents, stockbrokers, chiropractors,
podiatrists.
7. Medical occs have had improving incidence only in the
last few years.
8. Lifetime benefits have major impact on claim incidence.
2
9. CA has uniformly worse experience – though positive

signs in 1996+ issue years.
10. FL has poor experience in white collar – EP >=90+.
11. Multi-life does not always have lower incidence than
single life.
12. Claim termination experience has been more difficult
to improve.
13. Longer BP’s decrease claim termination rates.

14. COLA decreases claim termination rates.
15. Significant differences in claim terminations by
diagnoses.
16. Residual benefits product somewhat lower claim
terminations.
3
Claim Incidence Experience
by Contract Type
Overall Claim Incidence Results - % 85 CIDA

1990 - 1999
A&S 102%
OE 63%
DBO 119%
Claim Incidence Trends

A&S Experience
% 85 CIDA by Calendar Year
130%
120%
110%
100%
Occ CI 1
90%
Occ CI 2-4
80%
70%
60%
50%
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
Key Reasons for Improving Incidence Trend:

• Tighter underwriting & contracts
• Favorable economy
4
by Year of Issue and Policy Year
A&S - % 85 CIDA – Occ Class 1
160%
140%
120% Prior to 1990
100% 1990-92
80% 1993-95
60% 1996 & Later
40% ALL
20%
0%
1
6-10
11+
Year
Year
Year
Year
Year
s
s
Year
Year
Observations:
• Prior to 1990 issues incidence is higher than more recent years
• 1996 & later still looks exceptional
• Significant increase after year 2 (contestable period)
• General decreasing trend after year 5

by Year of Issue and Policy Year
A&S - % 85 CIDA – Occ Classes 2-4
100%
80% Prior to 1990

60% 1990-92
1993-95
40% 1996 & Later
20% ALL
0%
1
6-10
11+
Year
Year
Year
Year
Year
s
s
Year
Year
Observations:
• Differences aren’t as significant among issue year groups
• Relatively small differences by policy year
• Similar decreasing trend after year 5
5
by Key Occupations
Medical Occupations in Class 1 – Incidence by Year
200% Percent of 85 CIDA
190%
180%
Medical Occs
170%
Phys/Surg
160%
150%
140%
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
Observations:
• 1993-94 increase in incidence
• Recently returning to 1990-92 level

by Key Occupations
Non-Medical Occupations in Class 1 Incidence by Year
Percent of 85 CIDA
140%
120%
100%
80% All Non-Med Occs
60% Exec/Mngrs
40%
20%
0%
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
Observation:
• Consistent incidence improvement since 1990
6
Impact of Lifetime Benefits
on Claim Incidence
Average Incidence – 1990-99
Percentage of 85 CIDA
250%
200%
150%
No Lifetime
Lifetime
100%
50%
0%
Non-Med Occs Med Occs
Geographical Differences
In Claim Incidence
Occ Class 1 – Percent of 85 CIDA
EP Under 90 Days
Issue Year CA FL Other
Prior to 1990 129% 106% 92%
1990-92 127% 91% 83%
1993-95 133% 84% 97%
1996 & Later 133% 57% 91%
EP 90 Days & Over
Issue Year CA FL Other
Prior to 1990 198% 212% 140%
1990-92 182% 160% 122%
1993-95 145% 154% 101%
1996 & Later 84% 91% 67%
7
Claim Incidence
Multi-life vs. Single-life
Business Issued 1990 & Later
Percent of 85 CIDA
Multi-life Single-life Multi/Single
Elimination Periods Under 90 Days
No Lifetime 81% 77% 105%
Lifetime 75% 97% 77%
Total 80% 80% 100%
Elimination Periods 90 Days and Higher
No Lifetime 86% 105% 82%
Lifetime 146% 163% 89%
Total 99% 116% 85%
Observation:
• Multi-life incidence is not always lower than single life
• Multi-life best when EP >=90+, No Lifetime
Claim Termination Trends

by Calendar Year of Incurral
A&S Claims – 1990-99
120%
100% All
% 85 CIDA
80% Pre-1990
60% 1990-92
40% 1993-95
20% 1996 & Later
0%
1
6-10
11+
4-5
Year
Year
Year
Year
s
Year
Year
Claim Duration
Reasons for no significant termination improvements:

• More difficult claims
• Lower incidence might mean avg claims with longer duration
8
Claim Termination Experience
by Gender
A&S Claims – 1990-99
Percent 85 CIDA
160%
140%
120%
% 85 CIDA
100% Male
80%
60% Female
40%
20%
0%
1
6-10
11+
4-5
Year
Year
Year
Year
s
Year
Year
Claim Duration
Observation:
• Female termination experience sharply better than 85 CIDA after 2 years.

by Benefit Period
Average Termination Experience (1990-99)
200%
180%
160%
% 85 CIDA
140% Short BP
120% To 65-70
100% Lifetime
80%
60%
40%
1
-10
-5
11+
Year
Year
Year
6
Year
Year
Year
Claim Duration
9
Claim Termination Trends
Ultimate Claim Durations
Claim Duration 11+ by Attained Age
Percent 85 CIDA
Attained Age By Count By Indemnity Number of Terminations
35-39 425% 382% 38
40-44 187% 129% 88
45-49 184% 133% 219
50-54 130% 108% 269
55-59 108% 86% 290
60-64 75% 59% 268
65-69 65% 43% 64
70+ 68% 58% 34
All Ages 110% 86% 1,274
Observations:
• Long term termination rates converging to around 60% of 85 CIDA
• Big impact on valuation of lifetime claims

By Diagnoses
(Part 1)
Benefit Periods = To Age 65-70 or Lifetime
120%
100%
% 85 CIDA
All Diagnoses
80% Back
60% Musculoskeletal
Other Injury
40%
20%
1
-10
-5
11+
Year
Year
Year
6
Year
Year
Year
Claim Duration
10
By Diagnoses
(Part 2)
320%
270%
% 85 CIDA
220% All Diagnoses

Cardiovascular
170%
Cancer
120% Immunodef
70%
20%
1
-10
-5
11+
Year
Year
Year
6
Year
Year
Year
Claim Duration

By Diagnoses
(Part 3)
140%
120%
% 85 CIDA
100% All Diagnoses

80% Mental
60% Nervous
40% Alcohol & Drugs
20%
0%
1
-10
-5
11+
Year
Year
Year
6
Year
Year
Year
Claim Duration
11
Society of Actuaries
New York City Annual Meeting
October 26 - 28
Session 132 OF
What have we learned?
LTD Experience Committee – Recent Analysis Update
LTD Experience Committee
Participating Companies
AIG/American General Lafayette Life

American United Life Insurance Co. Liberty Mutual
Anthem Life Insurance Company MetLife Ins. Co.
Assurant Mutual of Omaha Insurance Co.
CIGNA Group Insurance Principal Financial Group
CNA Insurance Co Prudential Financial
Florida Combined Life Reliance Standard Life Ins. Co.
Genworth Safeco Insurance Co.
Guardian Life Insurance Co. Standard Insurance Co.
Hartford Life Insurance Co States West
Jefferson Pilot Financial UnumProvident Corp.
1
Experience Committee Members
Edd Bailey - Assurant Paul Hitchcox - ULR

Warren Cohen - Relaince Standard Rick Leavitt - Smith Group
Tom Corcoran - Tillinghast Allen Livingood - UnumProvident
Peter Doucette - Hartford Jack Luff - SOA
Pat Fay - MassMutual Roger Martin - UnumProvident, Chairman
Deb Fredricks - MetLife Chuck Meintel - JHA
Steve Garfield - Standard Eric Poirier - UnumProvident
Ray Siwek - Prudential
• Special thanks to Todd Fuhs and Steve Atkins as prior

committee members
• Independent Vendor: Solucient – Perry Beals, Julia Havey
Committee Focus – Paid claim termination study with separate analysis of

recoveries, mortality, benefit maximums, and settlements.
Timeline
• Initial data request sent out summer of 2003. Initial data analysis, mapping and
validation during the winter and spring of 2004. Resubmission for several
companies in Spring of 2004.
• Present initial review – 2004 SOA annual meeting.
• Distribute detailed results of initial review to participating companies - Fall 2004.
• Develop Experience Report and Experience table – 2005.
• Consider valuation table implications – 2006.
2
Profile of Initial Review

• 19 companies participating.
• More than 1.7 million claims submitted with more than 1.0 million currently in
experience study.
• Not all data submitted by each company was in sufficient detail to be included in
this initial review – most notable exclusion was by calendar year.
• 25 million months of claim exposure over 10+ calendar years.
• Dampening factors will be applied to reduce the influence of those companies
supplying the largest exposures.
• Initial variables reviewed include age, gender, elimination period, duration,
diagnosis of claim, definition of disability, and gross benefit amount.
• Analysis of raw recovery and death rates along with actual to expected ratios
relative to Table95A (t95a).
Recovery Rate - Company Distribution

100.00%
Min 25%-ile Median 75%-ile Max
10.00%
Termination Rate
1.00%
0.10%
0.01%
6
19 8
25 4
31 0
37 6
43 2
49 8
55 4
61 0
67 6
73 2
79 8
4
8
0
2
4
6
8
0
2
4
6
8
13 2
96
0+
1-
10
12
13
14
15
16
18
19
20
21
22
-1
-2
-3
-3
-4
-4
-5
-6
-6
-7
-7
-8
1
7-
24
Claim Duration
3
Mix comparison to Table95a
• Current study is about 5 times the size of t95a study.
• The mix of claims by EP is skewed more toward longer EP.
• Current table is about 1% more female.
• Current table is also skewed toward younger age mix & the age shift holds true for both genders.
• t95a exposure is from 1986 – 1996 with 80% from 1990 – 1996. Current table has at least 15%
more exposure at each overlapping calendar year and up to 6 times more in 1996. Overall, in
overlapping CY’s, the current table has twice the exposure.
Age Mix Comparison EP Mix
22% 75%
t04s t95a 58%
60%
16%
47%
% of table
40%
45%
34%
10%
30%
7% 13%
t04s T95a 15%
4%
0%
<30 30-34 35-39 40-44 45-49 50-54 55-59 60+
< 75 90 180
Recovery Rate - EP 180

5.0%
Actual Recovery
4.0%
Expected Recovery - t95a
Termination Rate
3.0%
2.0%
1.0%
0.0%
6 18 30 42 54 66 78
Monthly Claim Duration
4
Death Rate
1.4%
1.2%
Actual Death
Expected Death - t95a
1.0%
Termination Rate
0.8%
0.6%
0.4%
0.2%
0.0%
4 16 28 40 52 64 76
Monthly Claim Duration
Industry Termination Summary

200%
160%
141%
150% 141%
131%
110%
% t95a
114% 114%
108% 102% 103%
96%
100%
73% 97%
102% 96%
82% 79% 85% 88% 87% 83% 84%
79% 79% 77%
50%
1-3 m o 4-6 mo 7-9 m o 10-12 Yr 2 Yr 3 Yr 4 Yr 5-7 Yr 8-10 Yr 11- Yr 15+ All
mo 14
A/E Recovery A/E Death 'Combined' A/E Claim Duration
10
5
Claim Termination Trends by EP 90

200%
A/E Recovery A/E Death
'Combined' A/E
150% 138%
% t95a
112% 114%
120% 117%
106% 107% 105% 106% 109%
94%
100%
103% 103%
92% 96% 95% 90% 96%
88%
78% 81% 82% 87%
50%
4-6 mo 7-9 mo 10-12 Yr 2 Yr 3 Yr 4 Yr 5-7 Yr 8- Yr 11- Yr 15+ All
mo 10 14
Claim Duration
11
Claim Termination Trends by EP 180

200%
A/E Recovery A/E Death
182%
'Combined' A/E
157%
133% 133% 149%
150%
130% 127%
123% 123%
% t95a
117%
108%
100%
107%
95% 97%
81% 84% 82% 80% 81%
78% 78%
73%
65%
50%
4-6 mo 7-9 mo 10-12 Yr 2 Yr 3 Yr 4 Yr 5-7 Yr 8- Yr 11- Yr 15+ All
mo 10 14
Claim Duration
12
6
Male Claim Termination by Duration 196%

200%
A/E Recovery A/E Death 'Combined' A/E 158%
175%
151% 154%
150%
% t95a
125% 116% 115% 121% 116%

109%
95% 95%
100%
96% 96%
75% 63% 89%
77% 77% 76% 79% 77% 78%
66% 72% 73% 74%
50%
1-3 4-6 7-9 10-12 Yr 2 Yr 3 Yr 4 Year Yr 8- Yr 11- Yr 15+ All
mo mo mo mo 5-7 10 14
Claim Duration
13
Female Claim Termination by Duration

200%
A/E Recovery A/E Death 'Combined' A/E
175%
150% 131% 128% 129%

% t95a
104% 119%
125% 107% 110% 112% 109% 100%
92% 96% 121%
100% 98.4%
100% 114% 101% 96%
87% 88% 86% 94% 94%
75% 82%
76%
50%
1-3 4-6 7-9 10-12 Yr 2 Yr 3 Yr 4 Yr 5-7 Yr 8- Yr 11- Yr 15+ All
mo mo mo mo 10 14
Claim Duration
14
7
Claim Terminations by Diagnosis

150% 400%
361%
125%
300%
104%
% t95a - Recover
264%
% t95a - Death
100%
200%
75%
100%
50%
89%
25% 0%
01 06 11 02 03 04 05 07 08 09 10 12 13 U 95-oth
A/E Recovery A/E Death Category
Diagnosis Category
% Cat # Description % Cat # Description
0.9% 1 Maternity or Pregnancy/Childbirth
7.0% 6 Mental Nervous 25.1% 8 Muscoloskeletal/Connective Tissue
1.9% 11 Aids/HIV 3.6% 9 Other
1.6% 2 Reproductive/Urinary Disease 3.4% 10Respiratory
Nervous System & Sense Organ,
2.0% 3 Digestive 12.8% 12 Infectious/Parasitic
9.4% 4 Injury/Poisioning 17.4% 13 Circulatory
7.2% 5 Neoplasms 3.0% U Unknown
All categories except 1, 6, & 11 are
4.8% 7 Emergent Disabling Conditions 90.2% 95-oth mapped to t95a all-oth diagnosis cat 15
Recovery Rates by Calendar Year & Duration

2.5% 8%
7%
2.0%
Term ination Rate - Y rs > 1
6%
Term ination Rate Y r 1
1.5% 5%
4%
1.0% 3%
2%
0.5%
1%
0.0% 0%
90-92 93-95 1996 1997 1998 1999 2000 2001 2002
Yr 2 Yr 3 Yr 4 All Yr 1
Calendar Year
16
8
Death Rates by Calendar Year & Duration

Termination Rate 1.00%
0.75%
0.50%
0.25%
90-92 93-95 1996 1997 1998 1999 2000 2001 2002
Yr 1 Yr 2 Yr 3 Yr 4 All Calendar Year
17
9
1
Session 132
What Have We Learned?
Group Life Experience Committee Update
Susan R. Sames
Tillinghast
October 27, 2004
©Towers Perrin
© 2004 Towers Perrin
Group Life Experience Committee Members

Sue Sames, Chair Tillinghast
Jack Luff SOA
Jay Barriss MassMutual
John Bettano Prudential
Charlie DeWeese DeWeese Consulting
Marissa Limjoco MetLife
Marty Loughlin MJL Associates
Rocco Mariano Prudential
Gary Piccolo UnumProvident
Kari Powell Guardian
John Schwegel Fort Dearborn
Chris Svedin Beneficial Life
Reg Yoder Principal
Special thanks to Karen Edgerton and Ray Biondi as prior members of the Committee
Independent Vendor: Solucient
2
Group Life Experience Committee has been

working on two studies simultaneously
Mortality Study
Incidence study to support pricing (and possible update to IRS
Table I rates)
Measures death, waiver and accidental death and
dismemberment (“AD&D”) claims vs. insurance exposure for
group life plans
Waiver Reserve Study; a.k.a, Update to Krieger
Claim termination study to support valuation (and pricing)
Measures deaths and recoveries vs. exposed group life waiver
claims
Timeline for both experience studies
2002
Design study and issue call for data
2003
Receive and review submissions
2004:
Audit data
—Some down time while IDI and LTD were worked on
Next Steps
Resolve remaining data issues (2004)
Distribute detailed results of initial review to participating
companies (through early 2005)
Release study results (early to mid 2005)
Consider waiver valuation table implications (late 2005 - 2006)
3
Mortality Study is at data audit stage

Identifying and resolving data issues
Complex structure due to linking multiple files; e.g., group data,

individual exposure, self-administered exposure, and claims for
basic, optional, and accidental death and dismemberment
Nearly 20 companies submitted data
Key issues include:
Few companies submitted self-administered data
Many companies had difficulty providing waiver provision data
Relatively little data on AD&D
Dampening factors will be applied to largest contributor
Data Structure for Mortality Study

Lives and volume information was collected for
Claims (death, waiver, AD&D) and
Exposure (individual versus self-administered)
…across the following parameters:
gender
age
waiver provision (e.g., lifetime, no waiver)
type of coverage (e.g., basic, supplemental, optional)
group size
group effective date
SIC code
4
Sample pivot table for mortality study

The mortality study will have the following views for various
combinations of those parameters:
REPORT LAYOUT
Deaths Waiver
Exposure Claims Rate A/E Claims Rate A/E
Central Age
17
22
27
32
37
42
47
52
57
62
Subtotal 17-62
67
72
77
82
87
92
Subtotal 67 and Above
Total
Waiver Study is also at data audit stage
Waiver study was less complex, only one file per participating
company
23 companies submitted data
Two separate databases following Krieger’s format:
1. Select period claim durations up through ten years
—Age at disability and length of disability
2. Ultimate period
—Claim durations beyond ten years by attained age
Key issues include:
Truncating exposure period appropriately, e.g., company may
have submitted exposure that covers all 10 years but really
has claims and recoveries for only the past five years.
Addressing impact of systems changes and clean-up efforts
5
Data Structure for Waiver Study
Lives and volume information was collected for each claim
…for the following parameters:

gender
age at disability
duration of disability
attained age
termination reason, e.g., death, recovery, expiration of benefits
Sample pivot table format for select period

REPORT LAYOUT
By Age at Disablement
<20 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75 +
Rate of Expected (Krieger) Rate of Actual/Expected
Exposure Recovery Death Expiry Total Recovery Death Total Recovery Death
Length of Disability
8
9
11
12
8
9
11
12
2 Annual
3
4
5
6
7
8
9
10
10
6
Sample pivot table format for ultimate period

REPORT LAYOUT
Rate of Expected (Krieger) Rate of Actual/Expected

Exposure Recovery Death Expiry Total Recovery Death Total Recovery Death
Central Age
17
22
27
32
37
42
47
52
57
62
67
72
77
82
87
92
11
What have we learned?
Our initial strategy was to allow for more flexibility in

the structure of the submission to increase
participation
Having a tighter structure would have made the
linking much easier
Data audit process has been time consuming
Increasingly difficult for companies to commit
resources; however, the industry is very interested in
the results.
12

Society of Actuaries Annual Meeting

Transféré par

Informations du document

Description originale:

Copyright

Formats disponibles

Partager ce document

Partager ou intégrer le document

Options de partage

Avez-vous trouvé ce document utile ?

Ce contenu est-il inapproprié ?

Droits d'auteur :

Formats disponibles

Society of Actuaries Annual Meeting

Transféré par

Droits d'auteur :

Formats disponibles

Society of Actuaries

Disability and Group Life Experience Studies:

Results of IDEC Study

SOA Study Parameters

Major Lessons From the SOA DI Study

5. Excellent incidence from executives/managers,

9. CA has uniformly worse experience – though positive

Major Lessons From the SOA DI Study

13. Longer BP’s decrease claim termination rates.

Overall Claim Incidence Results - % 85 CIDA

Claim Incidence Trends

Key Reasons for Improving Incidence Trend:

Claim Incidence Trends

80% Prior to 1990

Claim Incidence Trends

Claim Termination Trends

Reasons for no significant termination improvements:

Claim Termination Experience

Claim Termination Experience

220% All Diagnoses

Claim Termination Experience

100% All Diagnoses

LTD Experience Committee

AIG/American General Lafayette Life

Experience Committee Members

Edd Bailey - Assurant Paul Hitchcox - ULR

• Special thanks to Todd Fuhs and Steve Atkins as prior

LTD Experience Committee

Committee Focus – Paid claim termination study with separate analysis of

Profile of Initial Review

LTD Experience Committee

Recovery Rate - Company Distribution

Min 25%-ile Median 75%-ile Max

• The mix of claims by EP is skewed more toward longer EP.

• Current table is about 1% more female.

LTD Experience Committee

Recovery Rate - EP 180

LTD Experience Committee

Industry Termination Summary

Claim Termination Trends by EP 90

LTD Experience Committee

Claim Termination Trends by EP 180

Male Claim Termination by Duration 196%

125% 116% 115% 121% 116%

LTD Experience Committee

Female Claim Termination by Duration

150% 131% 128% 129%

Claim Terminations by Diagnosis

LTD Experience Committee

Recovery Rates by Calendar Year & Duration

Death Rates by Calendar Year & Duration

October 27, 2004

Group Life Experience Committee Members

Group Life Experience Committee has been

Timeline for both experience studies

Mortality Study is at data audit stage

 Complex structure due to linking multiple files; e.g., group data,

Data Structure for Mortality Study

Sample pivot table for mortality study

Waiver Study is also at data audit stage

Data Structure for Waiver Study

 Lives and volume information was collected for each claim

 …for the following parameters:

Complex structure due to linking multiple files; e.g., group data,

Lives and volume information was collected for each claim

…for the following parameters:

Our initial strategy was to allow for more flexibility in