Buckeye Actuarial Continuing Education
April 26, 2011
Construction Defect Claim Management and Reserving
Presented by:
 Nathan Voorhis, FCAS, MAAA
 Robin Leibrock, JD
 Steven Jokerst, FCAS, MAAA
2
What is a Construction Defect Claim?
 CGL Policy – completed operations
 Builder, developer, contractor, subcontractor
 Construction, repair, remodel
 Residential and commercial buildings
 Resultant BI or PD, not the work itself
 Types of defects
 Faulty design, workmanship or material
 Roofing, flashing, soil preparation, framing,
waterproofing, doors & windows, carpentry
 Construction defect
 Multiple defendants, defects and policies
3
Where is this a Problem?
 Rapid growth and poor construction
 Litigious environment and highly organized plaintiff bar
 Continuous trigger occurrence
 California and other western states
 Certain states in northeast and southeast
4
Legal Decisions and Statutes
 (1995) Montrose vs. Admiral – known and progressive loss a
covered occurrence
 (1996) Stonewall Ins. Co. vs. City of Palos Verdes Estates Montrose applied to CD claims
 (1997) Calderon Process / (2002) Steinberg Bill – establish
procedures for filing CD claim
 (2001) Presley Homes vs. American States – broad duty to
defend
 California Statute of Limitation and Repose
 Patent defect – 3 years if reasonably apparent
 Latent defect – 10 years if not apparent
 (2004) L-J, Inc. vs. Bituminous F&M – no coverage for “your
work”
5
Coverage Endorsements
 Revise
policy language to clearly reflecting intended
coverage
 Occurrence
 Known or continuous loss (Montrose) exclusion – must
first become aware during policy period
 Prior work exclusion – no coverage for work completed
prior to stated date
 Redefine “occurrence”
 Exclusions for specified hazards and operations
 EIFS, mold, subsidence, imported drywall
 Roofing, residential construction
6
Coverage Endorsements
 Additional insured endorsement – subcontractor policy
covers GC for work performed on his behalf
 Other insurance endorsement
 Failure to complete your work
 Subrogation against third parties
 Abandoning the project
7
CD Claim Adjustment/Adjudication
 Claim made by owner, developer, builder, general contractor
 Multiple Tenders
 Named insured; additional insured
 Co-carrier for Named Insured
 Indemnitee
 Document intensive
 HO Matrix
 Notice of Completion dates, Close of Escrow dates
 Multiple policies, contracts, job and correspondence
files
 Trigger of Coverage
8
Liability Claim vs. CD Claim
 Liability Claim
 One or few plaintiffs
 Few defendants
 Known loss date
 Few damages / injuries
 One policy period triggered
 Shorter Statute of Limitation (BI 1-6 years; PD 1-10
years)
 Typically the primary focus is on Liability, rather than
Coverage or Damages
9
Liability Claim vs. CD Claim
 CD Claim
 Multiple plaintiffs – sometimes 100’s of homeowners
 Multiple defendants – design professional, developer,
general contractor, multiple subcontractors
 Undetermined loss date
 Multiple damages
 Multiple policy periods
 Longer Statute of Limitation (breach of contract 3-20
years)
 Typically the primary focus is on Damages and
Coverage, rather than Liability
10
Contracts, Coverage and Allocation
 CD claims typically involve multiple contracts (design
professional/owner; owner/developer; developer/builder;
builder/GC; GC/sub; Sub/sub)
 The coverage available and priority of coverage must be
analyzed for each contract and each policy triggered
 Allocation of claim costs
 Defense expenses – each policy is obligated to answer
so most courts require cost sharing by equal shares;
some courts allow sharing on a prorata basis
 Loss – time on risk; prorata by limits; combination
11
Difficulties with Traditional Actuarial Methods
Very long reporting lag – Pure IBNR dominates
Not ideal to combine with other book of general liability claims. Even if
premises/operations claims are excluded, development pattern is different
from typical products/completed operations pattern
If construction defect triangles broken out separately, exposure base for
Bornhuetter-Ferguson approach is subjective as there is no “construction
defect” premium since triangles broken out by cause of loss.
Most companies do not have sufficient accident year data to capture tail.
Tail estimation is very subjective.
Legislation has calendar year effect, affecting all accident years along a
given diagonal.
12
General Liability Claim Count
Accident Year Reported Development – CD vs NonCD
120.0%
80.0%
% Reported - CD
60.0%
% Reported - General
Liability NonCD
40.0%
20.0%
12
11
10
9
8
7
6
5
4
3
2
1
0.0%
Ye
ar
s
% Reported
100.0%
13
General Liability Accident Year
Incurred Loss Development – CD vs NonCD
120.0%
80.0%
% Incurred - CD
60.0%
% Incurred - NonCD
40.0%
20.0%
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0.0%
Ye
ar
s
% Incurred
100.0%
14
General Liability Accident Year
Paid Loss Development – CD vs NonCD
120.0%
100.0%
60.0%
% Paid - CD
40.0%
% Paid - NonCD
20.0%
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0.0%
Ye
ar
s
% Paid
80.0%
15
Incurred But Not Enough Analysis
 Case Development Existing Claims
 Incurred loss development approach organized in a report year format.
 Significantly reduced tail
 Must allocate estimated report year IBNR to accident year
•
Can choose among various methods including incurred losses, case
reserves, construction contractors premium, or other basis
 Multiple methods can be used; for simplicity this demonstration uses only
an incurred development approach.
16
Incurred But Not Enough Analysis
Report Year Construction Defect Loss Development
General Liability Incurred Loss Report Year Development
Construction Defect
120.0%
% Incurred
100.0%
80.0%
% Incurred
60.0%
40.0%
20.0%
0.0%
Years
1
2
3
4
5
6
7
8
9
17
Incurred But Not Enough Analysis
Report Year Summary of IBNE ($ Millions)
Report
Year
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Incurred
Loss &
ALAE
$2,842
$4,713
$8,709
$18,646
$19,315
$18,465
$18,695
$30,311
$14,315
$10,143
$20,117
$17,701
$10,912
$9,640
$7,063
$7,300
$218,886
Inc
CLDF
1.000
1.000
1.011
1.014
1.013
1.013
1.013
1.013
1.030
1.045
1.046
1.065
1.108
1.174
1.332
1.318
Ultimate
Loss &
ALAE
$2,842
$4,713
$8,802
$18,916
$19,568
$18,706
$18,939
$30,707
$14,741
$10,599
$21,051
$18,849
$12,090
$11,314
$9,410
$9,619
$230,867
Indicated
IBNE
$0
$0
$93
$270
$253
$242
$244
$396
$427
$456
$934
$1,148
$1,178
$1,675
$2,347
$2,319
$11,981
18
Incurred But Not Enough Analysis
Allocation of IBNE to Accident Year ($ Millions)
Selected IBNE for all Report Years
Accident
Year
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Total
Case
Reserves
$37
$132
$1,895
$328
$1,860
$4,254
$611
$878
$1,058
$88
$138
$297
$523
$1,494
$1,250
$166
$15,009
Case
Reserves
% of Total
0%
1%
13%
2%
12%
28%
4%
6%
7%
1%
1%
2%
3%
10%
8%
1%
100%
$11,981
IBNE
$29
$106
$1,513
$262
$1,485
$3,396
$488
$701
$845
$70
$110
$237
$417
$1,192
$998
$132
$11,981
19
Pure Incurred But Not Reported Analysis
 Accident year analysis of reported claim counts
 Selected Loss Severity
 Selected ALAE / Loss Ratio
 Selected Claims Closed with Pay
20
Pure IBNR Analysis
Accident Year Claim Count Development
120.0%
% Reported
100.0%
80.0%
% Reported - CD
60.0%
% Reported - General
Liability NonCD
40.0%
20.0%
0.0%
Years
1
2
3
4
5
6
7
8
9
10
11
12
21
Pure IBNR Analysis - Summary of IBNR Claim Counts (000’s)
Selected
IBNR Claim
Counts
Accident
Year
Reported
Claim Counts
Reported
Claim CLDF
Ultimate
Claim Counts
1995
2,800
1.030
2,884
84
1996
13,181
1.035
13,644
463
1997
20,566
1.044
21,468
902
1998
4,571
1.059
4,842
271
1999
3,175
1.101
3,494
319
2000
6,122
1.163
7,118
997
2001
2,450
1.266
3,102
652
2002
2,972
1.404
4,172
1,200
2003
963
1.587
1,528
565
2004
291
1.883
547
256
2005
214
2.253
481
267
2006
336
2.802
942
606
2007
651
4.108
2,674
2,023
2008
2,279
4.834
11,014
8,735
2009
1,911
6.501
12,423
10,512
9.783
8,937
8,023
99,268
35,876
2010
914
63,392
22
Pure IBNR Analysis
Selected Loss Severity
 Review past historical average paid loss and/or incurred loss severity
ratios
 Exclude unusual claims or years
 Incorporate IBNE into incurred severity calculation
23
Pure IBNR Analysis
Selected ALAE / Loss Ratio
 Review past paid ALAE to paid loss and/or incurred ALAE to incurred loss
ratios
 Ratios above 100% not uncommon
 Alternatively, severity analysis separately for ALAE
24
Pure IBNR Analysis
Selected Claims Closed without Pay
 Review past claims closed without pay to total closed ratios
 Ratios above 75% not uncommon; significant variability by policy year
25
Pure IBNR Analysis - Summary of IBNR
Accident
Year
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Total
IBNR
Counts
(000's)
84
463
902
271
319
997
652
1,200
565
256
267
606
2,023
8,735
10,512
8,023
35,876
Loss
ALAE/Loss Closed With
IBNR
Severity
Ratio
Payment % ($Millions)
$10,000
100.0%
20.0%
$336
$10,000
100.0%
20.0%
$1,853
$10,000
100.0%
20.0%
$3,606
$10,000
100.0%
20.0%
$1,084
$10,000
100.0%
20.0%
$1,278
$10,000
100.0%
20.0%
$3,986
$10,000
100.0%
20.0%
$2,607
$10,000
100.0%
20.0%
$4,802
$10,000
100.0%
20.0%
$2,260
$10,000
100.0%
20.0%
$1,026
$10,000
100.0%
20.0%
$1,070
$10,000
100.0%
20.0%
$2,422
$10,000
100.0%
20.0%
$8,092
$10,000
100.0%
20.0%
$34,940
$10,000
100.0%
20.0%
$42,049
$10,000
100.0%
20.0%
$32,092
$143,505
26
Summary of Construction Defect Losses ($ Millions)
Accident
Year
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Total
Case
$37
$132
$1,895
$328
$1,860
$4,254
$611
$878
$1,058
$88
$138
$297
$523
$1,494
$1,250
$166
$15,009
IBNE
$29
$106
$1,513
$262
$1,485
$3,396
$488
$701
$845
$70
$110
$237
$417
$1,192
$998
$132
$11,981
IBNR
$336
$1,853
$3,606
$1,084
$1,278
$3,986
$2,607
$4,802
$2,260
$1,026
$1,070
$2,422
$8,092
$34,940
$42,049
$32,092
$143,505
Total
Reserves
$402
$2,091
$7,014
$1,674
$4,623
$11,636
$3,706
$6,380
$4,163
$1,183
$1,318
$2,956
$9,032
$37,626
$44,296
$32,390
$170,494
27
Recent Years
 Most recent three years or so still very green
 Possible approaches
•
Hold loss ratio (use earned premium exposure summarizing only
construction contractors classes)
•
B/F accident year approach (use earned premium exposure
summarizing only construction contractors classes)
•
Same approach as described if large and stable enough book of data
28
Q&A
29
Buckeye Actuarial Continuing Education
April 26, 2011