Workers Comp. Reserving - How and When Should You Slice the Cake? Mark J. Mahon, FCAS, MAAA MBA, Inc. CLRS - September 23, 2002
It’s Easy to Want the Data - Receiving It Is Not So Easy zTo Start - Keep Things Simple
It’s Easy to Want the Data - Receiving It Is Not So Easy zTo Start - Keep Things Simple zFollow the Client’s Operations
It’s Easy to Want the Data - Receiving It Is Not So Easy zTo Start - Keep Things Simple zFollow the Client’s Operations zWhy Do You Want the Data?
It’s Easy to Want the Data - Receiving It Is Not So Easy zTo Start - Keep Things Simple zFollow the Client’s Operations zWhy Do You Want the Data? zTime Consuming and Expensive
Traumatic vs. Occupational Disease (Black Lung) zCoal Mine Compensation Rating Bureau of Pennsylvania.
Traumatic vs. Occupational Disease (Black Lung) zComparison of OD to Traumatic yCumulative
Traumatic vs. Occupational Disease (Black Lung) zComparison of OD to Traumatic yCumulative yLong Report Time
Traumatic vs. Occupational Disease (Black Lung) zComparison of OD to Traumatic yCumulative yLong Report Time yClaims take time to Adjudicate
Traumatic vs. Occupational Disease (Black Lung) zComparison of OD to Traumatic yCumulative yLong Report Time yClaims take time to Adjudicate yLifetime Benefits
Traumatic vs. Occupational Disease (Black Lung) zComparison of OD to Traumatic yCumulative yLong Report Time yClaims take time to Adjudicate yLifetime Benefits yLow Frequency/High Severity
The OD System zFrequency times Severity.
The OD Frequency zIBNR Claims yUses Typical Accident Year Triangles yClaims assigned to Year based on Last Date of Exposure yClaim Count Development used to estimate IBNR Claims
OD Frequency zThree Pieces yKnown Claimants Receiving Benefits at the valuation date (Awarded Claims) yAn Estimate of Pending Claims that will be Awarded yAn Estimate of IBNR Claims that will be Awarded
The OD Frequency zAwarded Claims
The OD Frequency zNew Federal Black Lung Legislation yIncrease in Reported Claims yIncrease in Awarded Claims yRetroactive
OD Severity zLife Tables zLifetime Benefits zClaimant’s Age zAnnual Benefits zMarital Status zEscalation (State - No, Federal- Yes)
OD Severity zLife Tables used to Model Future Benefits
OD Severity zAct 57 yOnly Claims Filed After Mid 1996 yWC Benefits Reduced for Social Security yWC Benefits Reduced for Pensions yAMA Guidelines Made it Harder To Achieve PT Disability Status
The OD Reserve zThree Pieces >Known Awarded Claims: Sum of Each Claimant’s future payments from severity model >Pending Claims: Pending Claims x Award Ratio x Average Severity of Pending Claims >IBNR Claims: IBNR Claims x Award Ratio x Average Severity of Awarded and Pending Claims
OD System - Concluding Remarks zLifetime Payments/Long Tail
OD System - Concluding Remarks zLifetime Payments/Long Tail zRelatively Few Claims
OD System - Concluding Remarks zLifetime Payments/Long Tail zRelatively Few Claims zParallels Actual Process
OD System - Concluding Remarks zLifetime Payments/Long Tail zRelatively Few Claims zParallels Actual Process zReviewed by Interested Parties
OD System - Concluding Remarks zLifetime Payments/Long Tail zRelatively Few Claims zParallels Actual Process zReviewed by Interested Parties zExpensive
OD System - Concluding Remarks zLifetime Payments/Long Tail zRelatively Few Claims zParallels Actual Process zReviewed by Interested Parties zExpensive zIs This Really the Only Way to Make Estimates?
Traumatic - Indemnity vs. Medical zVery Common Split zAlso Breakdown Medical into Medical Only and Medical on Indemnity zDifferent yDevelopment yInflation yLaw Changes
Tail Factor - Inverse Power Curve
Tail Factor - Relate to Larger Body
Tail Factor - Decay Factor
Traumatic Losses By Injury Type zDeath, Permanent Total, Permanent Partial and Temporary Claims
Traumatic Losses By Injury Type zDeath, Permanent Total, Permanent Partial and Temporary Claims zProportion of Losses by Injury Type yAct 57 Impacts Largest Claims ySeverity Subsequent to Law Change should be Smaller
Traumatic Losses By Injury Type zPermanent Total Claim Count Development
Traumatic Losses By Injury Type zPT Count Development after Act 57
Traumatic Losses By Injury Type zAct 57 Also made it Easier to Settle Claims yFewer Claims Being Reported as PT yRetroactive - Existing PT’s Settled and Reclassified
Traumatic Losses By Injury Type zReserve Estimation – Be Careful yIncurred Loss Triangles Double Count Impact of Settlements Understating Reserve Need yPaid Triangles Overstate Reserve Need
Traumatic Losses By Injury Type zSafety Program Impact yDrop in Number of Injuries yShift to less Severe Accidents
Losses in Higher Layers zBe Careful - Reserves in Higher Layers may be Inadequate zLoss Development Under Estimates Higher Layers in Most Recent AY yImmature years don’t yet have any big losses yChange in claims handling yHistorically there are no Higher Layer Losses in layer but because of new severe classes or inflation new losses are expected
Losses in Higher Layers zLatest Years’ Lower than Historical
Losses in Higher Layers zImmature Years have No Losses
Losses in Higher Layers zUse Industry Statistics to Fill-in Excess Losses yPayroll x Loss Cost x Excess Loss Factor yAs AY matures rely on actual data more and more zFor Insurers Do Gross, Ceded and Net Separately
Losses By Business Segment zMany Self-Insureds have Different Operating Units yE.g., Construction vs. Shipbuilding yFinancial Data Already Split yUsually Easy to get Actuarial Data as Coding already there
Losses By Business Segment zMany Self-Insureds have Different Operating Units yE.g., Construction vs. Shipbuilding yFinancial Data Already Split yUsually Easy to get Actuarial Data as Coding already there zData by Jurisdiction yUSL&H Benefits Higher than State Benefits
Combining Data zIncreases Credibility for Companies with Sparse Data
Combining Data zIncreases Credibility for Companies with Sparse Data zProvides a Guide to Overall Level
Combining Data zIncreases Credibility for Companies with Sparse Data zProvides a Guide to Overall Level zUse Combined Data as Minimum as usually the Sum of the Slices are greater than the Total yLess Variation in Total ySmaller Slice with Tail muted by Larger Slice without Tail yShifting Losses to Longer Tail Slice
Keeping Score zPurpose of Slices - Improve Reserve Estimation Accuracy zKeep Tabs on your Estimates - Keep A Scorecard
Keeping Score
zKeep Score of Your Estimates
Workers Comp. Reserving - How and When Should You Slice the Cake? Mark J. Mahon, FCAS, MAAA MBA, Inc. CLRS - September 23, 2002