Superimposed Inflation in CTP Schemes What s in our toolkit? Karen Cutter

Transcription

1 Superimposed Inflation in CTP Schemes What s in our toolkit? Karen Cutter

2 Today s presentation What is superimposed inflation? What are the problems in measuring SI? What s in the toolkit? - Actuarial model - Comparable claims model - Individual claim file reviews

3 What is superimposed inflation? a tendency for benefits for a given injury to increase faster than a suitable standard measure of inflation the increase in the total cost of compensation that has not been explicitly provided by the actuarial model(s)

4 Drivers of SI Many drivers including: - Legal decisions/precedents or scheme dynamics - An increased level of legal involvement - Better preparation by plaintiff lawyers - Claims handling practices of insurers

5 Why is it important? Understanding the level is important for: - good scheme management - input into valuation and pricing assumptions Arguably the area of most subjectivity due to: - problems in measuring it - the nature of SI

6 Problems in measuring SI Changes to claim frequency Changes in order of finalisation of claims Cause vs effect Model structure

7 So why use actuarial models? Problems do not invalidate their use But models should be used appropriately bearing in mind their limitations. Can: - provide indication and estimate of quantum - indicate areas where SI is apparent - direct further investigations into causal impacts

8 Case study: Actuarial model results Normalises for changing claim mix and changes in order of finalisation Claims mix dealt with by adopting common claim profile for all accident periods, using bootstrapping and simulation techniques Order of finalisation dealt with by forcing order to be more consistent, using boosting techniques

9 Case study: Total SI NSW Queensland Very High Superimposed Inflation Moderate Moderate Low High Low /Nil Superimposed Inflation (%) Very High High Low H1 Settlement Year H1 Settlement Year

10 Case study: By HoD 100% NSW % Queensland % 80% Contribution 60% 40% 20% Contribution 60% 40% 20% 0% 0% -20% GDs EL Care Med Legal Oth Head of Damage -20% GDs EL Care Med Legal Oth Head of Damage

11 Case study: By Injury severity 100% 80% NSW % 80% Queensland Contribution 60% 40% 20% Contribution 60% 40% 20% 0% 0% -20% Direct Sev1LR Sev2LR Sev3+LR -20% Unk S1NLR S1LR S2 S3+ Injury Severity Injury Severity

12 Comparable claims model Compares average size of like claims over time Four specific injuries examined - whiplash only, whiplash plus lumbar strain, whiplash plus thoracic strain, seat belt injuries Excluded certain claims to achieve more comparable body of claims

13 Case Study: NSW Injuries examined in context Number large 2% w hip only 6% whip+lumb 7% whip+thor 7% large 27% w hip only 1% Cost whip+lumb 4% w hip+thor 5% seat belt 2% seat belt 5% other 73% other 61%

14 Case Study: NSW Whiplash+Lumbar Average size by HoD Average Claim Size ($000) Settlement Year NEL Past Eco Future Eco Care Medical Plaint Legal Def legal Other

15 Case Study: NSW Whiplash+Lumbar Average awards and proportion receiving 50 Average Award % Receiving 100% % 60% 40% 20% Average Award ($000) Proportion Receiving 0% FEL Care Medicals P Legals Settlement Year

16 Case Study: Queensland Injuries examined in context Number large 2% w hip only 15% large 21% Cost w hip only 6% whip+lumb 10% whip+lumb 12% w hip+thor 8% seat belt 2% whip+thor 9% other 58% seat belt 4% other 53%

17 Case Study: Qld Whiplash+Lumbar Average size by HoD 80 Average Claim Size ($000) Settlement Year NEL Past Eco Future Eco Care Medical Plaint Legal Def legal Other

18 Case Study: Qld Whiplash+Lumbar Average awards and proportion receiving Average Award % Receiving % % 60% 40% 20% Average Award ($000) Proportion Receiving 0% FEL PEL GDs P Legals Settlement Year

19 Individual claim file reviews Used to test various hypotheses e.g: have changes in claims management contributed to SI? What changes could be made to reduce SI? Important that question(s) to examine is clearly defined

20 Individual claim file reviews Define pool of claims from which sample will be drawn Actuarial model and comparable claims model help define the pool (i.e. those claims where SI observed) Sample size big enough that conclusions can be expected to hold for whole pool

21 Individual claim file reviews Design data collection form to collect information about each claim in the sample Collect information not contained on electronic files Some information may be subject to judgement

22 Individual claim file reviews Conducted by suitably qualified, independent claims personnel Double review of some claims to ensure consistency in the review

23 Individual claim file reviews Analysis of sample may: - identify changes in plaintiff/defendant behaviour over time - help identify actions and strategies aimed at reducing SI A repeat file review could be conducted postimplementation of strategies to assess if effective

24 Conclusion SI is and will remain a difficult aspect of scheme management and actuarial work Measurement must be approached in a rigorous manner, but care taken regarding precision of measurement Sources and reasons for SI must be well understood

25 Conclusion Use of a variety of tools provides the best outcome for identifying and reacting to SI - simple comparable claims model for regular monitoring of problem areas - complex actuarial model for scheme wide assessment at least annually - individual file reviews for more complex questions around causes and remedial action