pour l'actuaire désigné for the Appointed Actuary Session: PD-12-Update on Living Benefit Experience Speaker: Emile Elefteriadis

Transcription

1 Canadian Institute of Actuaries Institut canadien des actuaires Seminar Colloque for the Appointed Actuary September 22 23, 2011 DoubleTree by Hilton Toronto Airport Hotel Session: PD-12-Update on Living Benefit Experience Speaker: Emile Elefteriadis pour l'actuaire désigné Les 22 et 23 septembre 2011 Hôtel Hilton DoubleTree Aéroport de Toronto

2 Update on Living Benefit Experience Critical Illness Insurance Critically Canadian CI Base Tables CI Experience Study Status UK Observations Group LTD Termination Experience Study 2

3 Base Tables CSCIT08 Research Paper Critically Canadian: Canadian Critical Illness Standalone Base Incidence Tables Individual Living Benefits Experience Subcommittee 3

4 Purpose-2008 CI Tables To develop a base table from Canadian data which can be used for benchmarking experience and developing pricing and valuation bases. Similar to Staple Inn reports on CI Ready for October/November

5 Format of Research Paper Chapter 1: Introduction Chapter 2: Market Overview Chapter 3: Methodology Chapter 4: The Base Tables Chapter 5: Commentary 150+ pages Bonus: Excel Spreadsheet with development of each impairment

6 Development of each Impairment Follows UK Critical Review/ Critical Path Base crude rates Definition Adjustment Trending to 2008 First-ever adjustment Overlap with other adjustments Prevalence adjustment 31-day mortality and sudden death

7 Reasonableness of tables Munich Re 2010 Industry Claims Underwritten Males and Females 3% 2008 CI Tables Males and Females "Other " accounts for disproportionate share of incidence. 7% 19% 4% 5% 6% 13% 53% 6% 68% 3% 13% CA Stoke HA MS CABG Other CA Stoke HA MS CABG Other

8 ReasonablenessRecalibrated 2008 CI Tables Munich Re 2010 Industry Claims Underwritten Males and Females 3% Males and Females-Recalibrated Re-weighted to produce 7% overall 7% 4% 4% 7% 7% 5% 6% 13% 59% 17% 68% CA Stoke HA MS CABG Other CA Stoke HA MS CABG Other

9 Reasonableness-Why is "Other" so large? 2008 CI Tables Males and Females "Other" Decomposition- Age CI Tables 0% 3% Males and Females -Age 45 14% 21% 6% 16% 3% 8% 15% 7% 53% 3% 8% 18% 11% 14% CA Stoke HA MS CABG Other PKNS KF HVS ALZ PARTIAL AS LOIE BBT MOT MINOR

10 Possible Reasons 2008 CI tables are derived from population data with some modification do not reflect adjustments intended to reflect the process of self-selection, underwriting, and differences in the risk profiles between insured lives and the population in general.

11 UK Insight Compare insured experience relative to population based CI tables. Impairment group Cancer Cardiovascular Alzheimer's Kidney Organ transplant TPD Accident type Other neuro Mortality* Insured / CIBT02 /ELT92* 71% 38% 10% 19% In UK, CIBT02 are analogous to the 2008 CI Tables So many of these "Other" impairments are less frequent in the insured population. 20% 16% 19% 66% 47%* Source: A Critical Table Pricing Critical Illness in the UK on a new Insured Lives table, page 37 by GenRe June 2007, ELT=English Life Tables.

12 Applying UK I/P Munich Re 2010 Industry Claims Underwritten Males and Females 3% 2008 CI Tables Males and Females-Recalibrated to UK I/P 7% 4% 4% 6% 5% 5% 3% 11% 13% 68% CA Stoke HA MS CABG Other 71% CA HA BP ST MS Other

13 More UK comparisonsmales

14 Female relationships

15 Trends Cancer cancer trend generally flat

16 Trend- heart attack AMI - age standardized - rates of hospitalization Male Ontario Male Canada Female Ontario Female Canada

17 Trends-stroke

18 Update on Experience Study Data requested in considerable detail to permit estimation of incidence of incidence by impairment and many other dimensions Five insurance groups have provided data files covering , two more to contribute files next year. Thomson Analytics vendor conducting the study

19 Update on Experience Study many errors and misunderstandings: missing data, claim records with no matching exposures, exposures indicating claims with no matching claim records, exposure records with plan indicators that don't exist, etc. Companies have been contacted by Thomson Analytics to address these issues

20 Update on Experience Study programming for report completed, but will need to be reviewed targeting initial draft report for December 2011, conditional on corrections to data file

21 Update on Experience Study based on UK experience study reporting, long time lags between diagnosis and reporting. e.g. about 90% reported within 1 year of diagnosis, 95% within 2 years.

22 Conclusions and summary Paper to be released November CI Tables appear reasonable Population trends since 2008 slightly decreasing overall Consistent relationships between Canadian and UK population incidence and mortality Experience study in progress, late 2012 target release