Experience Trends
for Income Protection
International Congress of Actuaries
Paris, France
June 1, 2006
Experience Trends for Income Protection
Panelists
Daniel Skwire
Milliman, Inc. (U.S.A.)
dan.skwire@milliman.com
Edward Fabrizio
General Reinsurance Life Australia Ltd
efabrizio@genre.com
Denis Garand
Denis Garand and Associates (Canada)
denis@garandnet.net
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 Actuaries
Individual 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 Contracts
Expected = 85 CIDA
120%
110%
Ratios
100%
90%
By Count
By Amount
80%
70%
60%
50%
40%
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Calendar Year
1990 – 1999 IDEC Study
Ratios
A/E Claim Incidence Ratios (by Amount) - A&S Contracts
Expected = 85 CIDA
130%
120%
110%
100%
90%
80%
70%
60%
50%
40%
Occ Class 1
Occ Classes 2-4
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Calendar Year
1990 – 1999 IDEC Study
A/E Claim Incidence Ratios (by Amount)
A&S Contracts - Occupation Class 1
Expected = 85 CIDA
160%
Ratios
140%
120%
Medical Occs
Non-medical Occs
100%
80%
60%
40%
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999
Calendar Year
1990 – 1999 IDEC Study
Claim Incidence Rates
Trends by Elimination Period and Gender
150%
Ratios
125%
100%
Male
Female
75%
50%
25%
0%
Under 30
30
60
Elimination Period
90
180
1990 – 1999 IDEC Study
A/E Claim Incidence (by Amount)
By Benefit Period - Expected = 85 CIDA
160%
140%
Ratios
120%
Short-term
To Age 65-70
Lifetime
100%
80%
60%
40%
Occ Class 1
Occ Class 2-4
85 CIDA Occupation Class
1990 – 1999 IDEC Study
Claim 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 Study
Claim 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 Study
Claim 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
Months 1-6
Months 7-12
Months 13-18
Months 19-24
Year 3
Year 4
Year 5
Year 6+
Adjustment
36-44%
50-75%
80-102%
105-120%
137%
120%
120%
100%
• IDEC Study measures experience relative to 1985 CIDA
1990 – 1999 IDEC Study
Ratios
A/E Claim Termination Ratios (by Amount)
By 85 CIDA Occupation Class
Expected = 85 CIDA
180%
160%
140%
120%
100%
80%
60%
40%
20%
0%
h
nt
M
Occ Cl 1
Occ Cl 2-4
1
h
nt
M
3
r
Qt
3
r
Qt
5
r
Qt
7
3
Yr
Duration of Disablement
5
Yr
Yr
1+
1
s
1990 – 1999 IDEC Study
140%
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
120%
Ratios
100%
80%
Medical Occs
Non-med Occs
60%
40%
20%
0%
Yr 1
Yr 2
Yr 3
Yr 4-5 Yr 6-10 Yr 11+
Duration of Disablement
1990 – 1999 IDEC Study
Ratios
A/E Claim Termination Rates (by Amount)
A&S Contracts By Benefit Period
Expected = 85 CIDA
200%
180%
160%
140%
120%
100%
80%
60%
40%
20%
0%
Short-term
To Age 65-70
Lifetime
Yr 1
Yr 2
Yr 3
Yr 4-5
Yr 6-10
Duration of Disablement
Yr 11+
1990-1999 IDEC Study
Key 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 – Trends
Claim cost in first 3 years of claim : 1 mth deferment
160
Male
Female
150
140
130
120
110
100
89-93
91-94
92-95
93-96
95-98
98-01
Australia – Latest Experience
Incidence Experience - Males
Incidence Rates - % of IAD89-93
Preliminary results for continuing companies only
Males
Occ Class
A
B
C
D
All
2 week deferment
95-98
98-01
1 month Deferment
95-98
98-01
100%
94%
96%
94%
73%
81%
83%
73%
105%
101%
109%
113%
78%
97%
105%
108%
96%
78%
108%
97%
Australia – Latest Experience
Incidence Experience - Females
Incidence Rates - % of IAD89-93
Preliminary results for continuing companies only
Females
Occ Class
A
B
C
D
All
2 week deferment
98-01
95-98
1 month Deferment
98-01
95-98
96%
101%
93%
90%
72%
96%
83%
73%
97%
97%
93%
84%
73%
79%
78%
53%
95%
74%
96%
74%
Australia – Latest Experience
Occupation Relativities
Preliminary results for continuing companies only
Weighted Average Incidence Rates as % Male Class A
male B
male C
male D
2 weeks
150% (156%)
238% (243%)
237% (273%)
1 month
175% (156%)
379% (320%)
438% (379%)
Weighted Average Incidence Rates as % Female Class A
female B
female C
female D
2 weeks
154% (138%)
141% (164%)
146% (176%)
1 month
136% (134%)
189% (157%)
142% (174%)
Australia – Latest Experience
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%
Australia – Latest Experience
Termination Rates
Preliminary results for continuing companies only
Ratio of A/E Number of Terminations
Occ Class
MALE
FEMALE
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%
36
24
23
m
m
m
m
m
m
m
m
m
m
m
m
m
m
m
m
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
m
th
s
m
th
s
m
th
s
m
th
s
m
th
s
m
th
s
700
22
ks
ks
m
th
s
w
w
ks
ks
ks
ks
ks
ks
ks
w
w
w
w
w
w
w
800
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
12
11
10
9
8
7
6
5
4
Australia – Latest Experience
Termination Rates – Male: Occ A: 1 mth deferment
1000
900
95-98
98-01
600
500
400
300
200
100
0
36
24
23
m
m
m
m
m
m
m
m
m
m
m
m
m
m
m
m
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
th
s
m
th
s
m
th
s
m
th
s
m
th
s
m
th
s
m
th
s
700
22
ks
ks
m
th
s
w
w
ks
ks
ks
ks
ks
ks
ks
w
w
w
w
w
w
w
800
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
12
11
10
9
8
7
6
5
4
Australia – Latest Experience
Termination Rates – Female: Occ A: 1 mth deferment
1000
900
95-98
98-01
600
500
400
300
200
100
0
Australia – Latest Experience
Termination Rates
Preliminary results for continuing companies only
Actual Termination Rates of Claim by duration of claim
Termination Rate
Mths 1 to 3
Mths 4 to 6
Mths 7 to 12
Mths 13 to 24
Mths 25 to 36
95-98
Male Female
0.491
0.326
0.341
0.372
0.197
0.490
0.327
0.315
0.353
0.243
98-01
Male Female
0.468
0.318
0.303
0.494
0.313
0.475
0.307
0.299
0.549
0.383
Ratio
Male Female
95%
97%
89%
133%
158%
97%
94%
95%
155%
157%
Australia – Latest Experience
Amount vs Number of Terminations
Preliminary results for continuing companies only
Ratio of A/E amounts versus A/E number for Termination Rates of Claim
Occ Class
MALE
FEMALE
A
96%
98%
B
95%
95%
C
95%
95%
D
93%
97%
Australia – Latest Experience
Variation in Experience by Company
Preliminary results
A/E Terminations
100%
90%
80%
70%
60%
50%
40%
50%
60%
70%
80%
90%
A/E Incidence
100%
110%
120%
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
Incidence Rate Experience
Male A/E
Female A/E
100%
90%
Actual vs. Expected
80%
70%
60%
50%
40%
30%
20%
10%
0%
1997
1998
1999
2000
Event Year
2001
2002
New Zealand – Latest experience
Termination Rate Experience
120%
100%
80%
60%
40%
20%
0%
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Month of disablement
2000-02
1997-99
New Zealand – Latest experience
Claim Duration Experience
Ratio (%) of actual to expected claim length by deferment
period and investigation period
Deferment Period
1997-99
2000-02
30-days
169
146
90-days
146
132
Ratio (%) of actual to expected claim length by occupation
class and investigation period
Occupation Class
1997-99
2000-02
White-collar
168
141
Blue-collar
154
140
New Zealand – Latest experience
• 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
New Zealand – Latest experience
Paid ratio (%) by cause of claim and investigation period
Cause of Claim
1997-99
2000-02
Accident
52
43
Sickness
79
81
All
72
70
Paid ratio by claim cause and occupation class in 2000-02
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
New Zealand – Latest experience
Contribution of each component to total change in claim cost
from 1997-99 to 2000-02
Male A/E
Incidence
Duration
Female A/E
Paid Ratio
5%
Cost
1%
Change in A/E Result
0%
-5%
-2%
-3%
-10%
-15%
-11%
-10%
-14%
-20%
-25%
-24%
-30%
-35%
-31%
New Zealand – Latest experience
Comparison of ratios (%) of actual to expected results in
2000-02 weighted by numbers and amounts
160%
140%
120%
100%
Numbers
80%
Amounts
60%
40%
20%
0%
Incidence
Duration
Paid Ratio
Cost
New Zealand – Latest experience
Warning: smoking hazardous to your health
Ratio (%) of actual to expected incidence for smokers over nonsmokers in 2000-02
Occupation Class
Male
Female
All
White-collar
142 (13)
135 (16)
138 (10)
Blue-collar
123 (10)
117 (31)
122 (10)
122 (7)
129 (14)
124 (6)
All Occupation
Classes
Ratio (%) of actual to expected claim cost for smokers over nonsmokers in 2000-02
Occupation Class
Male
Female
All
White-collar
143
123
134
Blue-collar
132
123
131
All Occupation
124
121
123
Classes
New Zealand – Latest experience
Agreed Value benefits versus Indemnity
120%
100%
80%
60%
111%
40%
95%
106%
113%
Paid Ratio
Cost
20%
0%
Incidence
Duration
New Zealand – Latest experience
Self-employed versus Employed
160%
140%
120%
100%
80%
152%
126%
60%
102%
40%
80%
20%
0%
Incidence
Duration
Paid Ratio
Cost
New Zealand – Latest experience
Variation in experience by company
Incidence
140%
Bubbles show claim
130%
cost
120%
110%
100%
99
90%
86
65
80%
70%
60%
6769
42
50%
59
50
40%
Duration
30%
20%
80%
90%
100%
110%
120%
130%
140%
150%
160%
170%
180%
190%
200%
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
Experience Trends in Canada
Denis Garand
Denis Garand & Associates
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
Average age
Distribution in %
Female
39.7
40%
Male
40.1
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
Canada, except Quebec
Quebec
8.4
6.2
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 412 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