Construction Defects CLRS September 19, 2000 Minneapolis
Dec 16, 2015
Construction Defects
CLRS
September 19, 2000
Minneapolis
Presenters:
Chandu Patel, FCASSr. Actuarial ManagerKPMG LLPModerator
Ronald Herrig, FCASActuaryShand Morahan & Co.Panelist
Paul KantonSr. Claims SpecialistMarkel West / Associated
International Insurance Co.
Panelist
Peter MackSr. Claims SpecialistMarkel West / Associated
International Insurance Co.
Panelist
Scope of Presentation
• Background
• Legal Developments
• Actuarial Issues
• Emerging Issues
Background
California Population Growth
• Between 1980 and 1996:– the US population grew by 16.7%.– California’s population grew by 36% – 2.8 million new housing units were built in
California (1 for each 3 new residents)– demand for housing exceeded supply by
600,000 units.
When Housing Demand outpaces Supply
• Builders rush to meet this demand
• Unskilled construction labor enters the market
• Laborers are unsupervised
• Short cuts are taken to save time
In Mathematical Terms:
Unskilled Labor
+ Poor/No Supervision
+ Unrealistic Deadlines
Substandard Housing
“So, what exactly is a ‘Construction Defect’”?
Categories of Defects
1. Defects in Design, Workmanship and Materials
Defect Effect
Framing Structural Failure
Roofing Water Intrusion
Windows Water Intrusion
Categories of Defects
2. Soil Problems– Improper compaction
Subsidence– Inadequate grading
Lateral Mvmt
– Inadequate drainage– Expansive Soil– Seismic Activity
Are These Defects Insurable?
Policyholders – Yes, General Liability policies provide coverage for completed operations.
Insurance Companies – Yes and No, policy covers damage caused by defect, but not cost to repair the defect.
Courts – Sooner or later, we’ll decide!
Legal Developments
Important California Court Cases, Legislation affecting Construction
Defects
1. Montrose I (1993)
2. Montrose II (1995)
3. Stonewall (1996)
4. Calderon Act (1997)
Montrose Chem’l v. Superior Court (of LA County) - 1993
Issue: Insurer’s obligation to defend with respect to proceedings related to the discharge of hazardous substances.
Ruling: Complaint need only allege that damages may have occurred to trigger the defense obligation.
Impact: Leads to the defense of more claims; increases severity of ALAE.
Montrose Chem’l v. Admiral Insurance Co. - 1995
Issue: Use of continuous injury trigger for duty to defend hazardous waste actions.
Ruling: All insurers “from shovel to gavel” have potential liability. All past, current and future policies may apply.
Impact: Could lead to increased claim frequencies; lower severities.
Stonewall Ins Co. v. City of Palos Verdes Estates (1996)
Issue: Application of “Montrose” phenomenon to construction defect cases.
Ruling: Continuous injury trigger does apply to construction defect cases.
Impact: Litigation and claim counts increase significantly.
Calderon Act (1997)
• Requires communication between HO Association and Builder as a pre-condition to filing a lawsuit.
• Encourages Mediation between parties.• Initially increased filing of lawsuits.• Ultimately delayed the filing of lawsuits.• Generally, ineffective in resolving claims
and avoiding lawsuits.
Actuarial Analysis
Scope
• Loss Characteristics
• Chain Ladder Method and Pitfalls
• Frequency and Severity Approaches
• Summary
Characteristics of Losses
• Developers / General Contractors
• Sub-contractors / Artisans
Developers / General Contractors
• Shorter Report Lag• Longer Closure Pattern
• Lower Frequency• Higher Severity• Higher ALAE to Loss
Developers / General Contractors
• Shorter Report Lag• Longer Closure Pattern
• Lower Frequency• Higher Severity• Higher ALAE to Loss
Sub-contractors / Artisans
• Longer Report Lag• Shorter Closure Pattern
• Higher Frequency• Lower Severity• Lower ALAE to Loss
Other General Characteristics
• Lots of Legal Expense– Active Plaintiff’s Bar– Coverage Litigation– Duty to Defend
• Long Statute of Limitations
• Many Cross Complaints
Questions an Actuary must ask:
• Are the risks we insure Developers/General Contractors or Sub-contractors?
• Is my layer of exposure Primary or Excess?
• Is expense Inside or Outside of Limits?
• Can my data be broken out separately?
• Why’d I become an Actuary?
Potential Reserving Techniques
• Incurred Chain Ladder
• Paid Chain Ladder
• Frequency x Severity
• Other?
Note: Each of the following exhibits is derived from Sub-contractor data. The data used is actual reported data modified by a scaling factor.
Construction Defects CONSTRUCTION DEFECTSReported Loss and ALAE Reported Loss and ALAE - Direct
12/31/1999 12/31/1999
Age in MonthsAx Yr 12 24 36 48 60 72 84 96
1992 0 85,000 350,000 750,000 775,000 800,000 925,000 915,0001993 90,000 1,000,000 2,500,000 3,250,000 4,500,000 7,000,000 9,500,0001994 2,000,000 3,500,000 5,000,000 7,500,000 11,500,000 16,000,0001995 1,500,000 5,000,000 7,500,000 14,500,000 19,000,0001996 750,000 4,000,000 8,500,000 11,500,0001997 1,000,000 2,000,000 2,300,0001998 400,000 700,0001999 0
GL "+12 Mo" "+12 Mo"Incurred Comp. Ops Indicated Indicated Actual
Ax Yr As of 12/98 LDFs Ultimate Incurred Incurred1
1992 925,000 1.102 1,019,350 12,764 -10,0001993 7,000,000 1.125 7,875,000 146,098 2,500,0001994 11,500,000 1.178 13,547,000 541,778 4,500,0001995 14,500,000 1.305 18,922,500 1,563,243 4,500,0001996 8,500,000 1.559 13,251,500 1,654,406 3,000,0001997 2,000,000 2.093 4,186,000 685,055 300,0001998 400,000 3.500 1,400,000 268,896 300,000
44,825,000 60,201,350 4,872,240 15,090,000
Construction DefectsReported Loss and ALAE
12/31/1999
Age in MonthsAx Yr 12 24 36 48 60 72 84 96
1992 0 85,000 350,000 750,000 775,000 800,000 925,000 915,0001993 90,000 1,000,000 2,500,000 3,250,000 4,500,000 7,000,000 9,500,0001994 2,000,000 3,500,000 5,000,000 7,500,000 11,500,000 16,000,0001995 1,500,000 5,000,000 7,500,000 14,500,000 19,000,0001996 750,000 4,000,000 8,500,000 11,500,0001997 1,000,000 2,000,000 2,300,0001998 400,000 700,0001999 0
Construction "+12 Mo" "+12 Mo"Incurred Defects Indicated Indicated Actual
Ax Yr As of 12/98 LDFs Ultimate Incurred Incurred1
1992 925,000 1.100 1,017,500 62,864 -10,0001993 7,000,000 1.250 8,750,000 954,545 2,500,0001994 11,500,000 1.650 18,975,000 3,680,000 4,500,0001995 14,500,000 2.300 33,350,000 5,712,121 4,500,0001996 8,500,000 3.750 31,875,000 5,358,696 3,000,0001997 2,000,000 6.500 13,000,000 1,466,667 300,0001998 400,000 18.250 7,300,000 723,077 300,000
44,825,000 114,267,500 17,957,970 15,090,000
Construction DefectsPaid Loss and ALAE
12/31/1999
CONSTRUCTION DEFECTSAge in Months Reported Loss and ALAE - Net
Ax Yr 12 24 36 48 60 72 84 96
1992 0 25,000 75,000 750,000 775,000 775,000 800,000 850,0001993 23,000 150,000 1,500,000 2,500,000 3,500,000 5,000,000 7,000,0001994 350,000 1,700,000 3,000,000 5,500,000 8,500,000 10,500,0001995 375,000 2,000,000 5,000,000 9,500,000 14,000,0001996 175,000 1,500,000 4,500,000 7,500,0001997 200,000 1,000,000 1,750,0001998 125,000 250,0001999 0
GL "+12 Mo" "+12 Mo"Paid Comp. Ops Indicated Indicated Actual
Ax Yr As of 12/98 LDFs Ultimate Paid Paid1
1992 800,000 1.223 978,400 49,306 50,0001993 5,000,000 1.352 6,760,000 527,392 2,000,0001994 8,500,000 1.602 13,617,000 1,571,746 2,000,0001995 9,500,000 2.135 20,282,500 3,160,737 4,500,0001996 4,500,000 3.175 14,287,500 2,192,037 3,000,0001997 1,000,000 5.463 5,463,000 720,630 750,0001998 125,000 12.500 1,562,500 161,015 125,000
29,425,000 62,950,900 8,382,862 12,425,000
Construction DefectsPaid Loss and ALAE
12/31/1999
Age in MonthsAx Yr 12 24 36 48 60 72 84 96
1992 0 25,000 75,000 750,000 775,000 775,000 800,000 850,0001993 23,000 150,000 1,500,000 2,500,000 3,500,000 5,000,000 7,000,0001994 350,000 1,700,000 3,000,000 5,500,000 8,500,000 10,500,0001995 375,000 2,000,000 5,000,000 9,500,000 14,000,0001996 175,000 1,500,000 4,500,000 7,500,0001997 200,000 1,000,000 1,750,0001998 125,000 250,0001999 0
Construction "+12 Mo" "+12 Mo"Paid Defects Indicated Indicated Actual
Ax Yr As of 12/98 LDFs Ultimate Incurred Incurred1
1992 800,000 1.150 920,000 51,852 50,0001993 5,000,000 1.280 6,400,000 565,217 2,000,0001994 8,500,000 1.600 13,600,000 2,125,000 2,000,0001995 9,500,000 2.400 22,800,000 4,750,000 4,500,0001996 4,500,000 4.400 19,800,000 3,750,000 3,000,0001997 1,000,000 11.000 11,000,000 1,500,000 750,0001998 125,000 62.750 7,843,750 588,068 125,000
29,425,000 82,363,750 13,330,137 12,425,000
Problems with Chain-Ladder
• GL experience isn’t representative of CD.
• Company CD history may not be extensive enough.
• Oldest accident years may still be developing, maybe substantially.
• Little industry-wide experience available.
Will Bornhuetter-Ferguson work better?
• B-F is LDF dependant – the same LDF issues will apply
• How do you pick an a priori ULR? – short history makes this difficult; highly judgmental.
Frequency x Severity Approach
• Severities are low and stable/decreasing– Lower for Subs than for General Contractors– Trending downward as:
• More Policies are exposed
• More insurers are brought into litigation
• Frequency is the key to projections
Construction DefectsSeverity Information
12/31/1998
Reported Open Paid ReportedAccident Counts Counts Losses LossesYear 12/31/98 12/31/98 12/31/98 12/31/98
1992 15 5 800,000 925,0001993 223 78 5,000,000 7,000,0001994 570 175 8,500,000 11,500,0001995 688 215 9,500,000 14,500,0001996 535 208 4,500,000 8,500,0001997 168 73 1,000,000 2,000,0001998 38 10 125,000 400,000
2237 764 29,425,000 44,825,000
Paid per O/S perAccident Closed Open ReportedYear Severity Severity Severity
1992 80,000 25,000 61,6671993 34,483 25,641 31,3901994 21,519 17,143 20,1751995 20,085 23,256 21,0761996 13,761 19,231 15,8881997 10,526 13,699 11,9051998 4,464 27,500 10,526
Total 19,976 20,157 20,038
Will this continue once fully developed?
Construction DefectsCalendar Year Statistics
12/31/1999
Cumulative Incremental Cumulative IncrementalCalendar Reported Reported Reported Reported ReportedYear Count Count Losses Losses Severity
1992 0 0 0 0 01993 33 33 175,000 175,000 5,3031994 183 150 3,350,000 3,175,000 21,1671995 538 355 8,250,000 4,900,000 13,8031996 1053 515 14,775,000 6,525,000 12,6701997 1596 543 25,300,000 10,525,000 19,3831998 2237 641 44,825,000 19,525,000 30,460
1999 2875 638 59,915,000 15,090,000 23,652
Techniques to Estimate Ultimate Counts
• Accident Year Approach
• Calendar Year Approach
• Exposure Growth Approach
Construction DefectsClaim Count EstimationAccident Year Approach
Age in MonthsAx Yr 12 24 36 48 60 72 84 96
1992 0 8 10 10 10 10 15 171993 25 53 95 120 155 223 2651994 120 245 340 443 570 7701995 188 408 523 688 8901996 175 385 535 6751997 80 168 2101998 38 481999 0
Reported Construction "+12 Mo" "+12 Mo"Counts Defects Indicated Indicated Actual
Ax Yr As of 12/98 LDFs Ultimate Counts Counts1
1992 15 1.200 18 2 21993 223 1.500 335 56 421994 570 2.000 1,140 190 2001995 688 2.600 1,789 206 2021996 535 3.250 1,739 134 1401997 168 4.250 714 52 421998 38 6.000 228 16 10
2,237 5,962 655 638
Construction DefectsClaim Count EstimationCalendar Year Approach
Age in MonthsAx Yr 12 24 36 48 60 72 84 96
1992 0 8 10 10 10 10 15 171993 25 53 95 120 155 223 2651994 120 245 340 443 570 7701995 188 408 523 688 8901996 175 385 535 6751997 80 168 2101998 38 481999 0
Incremental Cumulative Growth Calcul-Calendar Reported Reported Pattern ated Selection Increment'lYear Time Count Count Factor Ultimate at 12/98 Expected
1992 t=0 0 01993 t=12 33 33 325.000 10,7251994 t=24 150 183 45.000 8,235 2051995 t=36 355 538 10.500 5,649 6011996 t=48 515 1,053 4.500 4,739 7171997 t=60 543 1,596 2.500 3,990 4,100 8421998 t=72 641 2,237 1.750 3,915 6841999 t=84 638 2,875 1.300 3,738 774
Construction DefectsClaim Count EstimationEstimated Pattern of Reporting
(NEP) Pattern of Claim Reports by Accident Year/Calendar YearC/AY Exposure 12/92 12/93 12/94 12/95 12/96 12/97 12/98 12/99 12/00 12/01
1992 130 2.5% 5.0% 10.0% 30.0% 25.0% 20.0% 7.0% 5.0%1993 2,408 2.5% 5.0% 10.0% 30.0% 25.0% 20.0% 7.0% 5.0%1994 9,750 2.5% 5.0% 10.0% 30.0% 25.0% 20.0% 7.0% 5.0%1995 12,773 2.5% 5.0% 10.0% 30.0% 25.0% 20.0% 7.0%1996 11,063 2.5% 5.0% 10.0% 30.0% 25.0% 20.0%1997 7,960 2.5% 5.0% 10.0% 30.0% 25.0%1998 2,353 2.5% 5.0% 10.0% 30.0%
Relative Reports by AY/CY1992 130 3.25 6.50 13.00 39.00 32.50 26.00 9.10 6.501993 2,408 60.19 120.38 240.75 722.25 601.88 481.50 168.53 120.381994 9,750 243.75 487.50 975.00 2925.00 2437.50 1950.00 682.50 487.501995 12,773 319.31 638.63 1277.25 3831.75 3193.13 2554.50 894.081996 11,063 276.56 553.13 1106.25 3318.75 2765.63 2212.501997 7,960 199.00 398.00 796.00 2388.00 1990.001998 2,353 58.81 117.63 235.25 705.75
Relative Growth: 3.25 66.69 377.13 1086.56 2644.94 5582.25 8322.91 9550.53 8746.25 6289.83
Actual Counts: 0 33 183 538 1053 1595 2235Actual Growth: 0 33 150 355 515 542 640
Projected Claims Reported by Calendar YearEstimated Growth: 734 673 484Estimated Counts: 2969 3642 4126
Actual for 1999: 638
Pattern of Claim Reports by Accident Year/Calendar Year12/02 12/03 12/04 12/05
5.0%7.0% 5.0%
20.0% 7.0% 5.0%25.0% 20.0% 7.0% 5.0%
Relative Reports by AY/CY
638.63774.38 553.13
1592.00 557.20 398.00588.13 470.50 164.68 117.63
3593.13 1580.83 562.68 117.63
Projected Claims Reported by Calendar Year276 122 43 9
4402 4523 4567 4576
Construction DefectsFrequency x Severity CalculationsBased upon Data as of 12/31/98
Ultimate Losses = Incurred Losses + (IBNYR Count x Selected Severity)IBNYR Count = (Ultimate Claim Count - Total Claim Count as of 12/98)
Incurred Losses as of 12/98: 44,825,000Total Claim Count as of 12/98: 2,237
Ultimate Claim Count Assumptions: 1) Calendar Year Approach 4,100 2) Exposure Growth Approach 4,576 3) Accident Year Approach 5,962
Severity Assumptions: 1) Low (Decreasing AY Sev.) 15,000 2) Medium (Current O/S Sev.) 20,000 3) High (Increasing CY Sev.) 25,000
Calculation of Ultimate Losses:
F x S Approach Selected Frequency:CY EG AY
Severity: 4,100 4,576 5,962
Low 15,000 72,770,000 79,910,000 100,700,000
Medium 25,000 91,400,000 103,300,000 137,950,000
High 35,000 110,030,000 126,690,000 175,200,000
Construction DefectsSummary of Estimated UltimatesBased upon Data as of 12/31/98
Current Data:Incurred Losses at 12/98: 44,825,000Incurred Losses at 12/99: 59,915,000
Method Results:Reported ISO GL Chain-Ladder 60,201,350Reported CD Chain-Ladder 114,267,500
Paid ISO GL Chain-Ladder 62,950,900Paid CD Chain-Ladder 82,363,750
Frequency x Severity Estimate 1 72,770,000 Estimate 6 137,950,000 Estimate 2 79,910,000 Estimate 7 110,030,000 Estimate 3 100,700,000 Estimate 8 126,690,000 Estimate 4 91,400,000 Estimate 9 175,200,000 Estimate 5 103,300,000
Range of Estimates: 60,201,350 - 175,200,000Median of Estimates: 100,700,000Mean of Estimates: 101,364,115
Selected Ultimate Incurred Losses: 105,000,000
What's the Correct Answer?
Sample Industry data – Sub-Contractors
• ALAE to Loss RatioRange 75% to 105%
• Mature Reported Average Severity
Range $22,000 to $23,000
As with any industry data, discretion should be used in applying these factors to individual company analyses.
The Future
Mitigation Efforts
• Settlement Efforts
• Coverage Restrictions
Settlement Prior to TrialPros
– Reduces expensive Trial Costs– Juries are biased toward Homeowners– Encouraged by Calderon
Cons– Mediations can be numerous, lengthy and
expensive– Involve many parties, each with own interests
Coverage Restrictions
• Montrose Exclusions – Could prohibit coverage for losses known to the insured before the policy’s inception.
• Cost Inclusive Coverage – Contains ALAE within the Policy Limits
• Aggregate Limits – Contains Catastrophic Exposure.
What does the Future Hold?
• CD Litigation expanding to Other States
• Continued Aggressive Litigation
• Challenge our Actuarial Techniques
Expansion to Other States
• Already seeing litigation in Nevada, Colorado, Florida and Washington
• Arizona’s population grew 20.4% from 1990-1996, compared to 6.4% nationwide.
Continued Aggressive Litigation
• Construction Defect litigation is a no-lose situation for Plaintiff’s attorneys.
• Army of Plaintiff’s Experts/Army of Defense Experts
• New Concepts of Liability being proposed
Challenge our Actuarial Techniques
• Exposure is Comparable to Environmental/Toxic exposures.
• Communicate with Claims and Underwriting Departments.
• Continually question/monitor methods.
• Incorporate Educated Judgment into analysis.
Sometimes they’re all trick questions!