Actuarial Notes

2024 Canadian Pensioner Mortality Research Project

By Kai F. Kaufhold, A. Jibreel Ismail and Stefan J. Ramonat · Canadian Institute of Actuaries · 2026 · Document 226032

Study guide: Exam RET-301

A research report, dated March 2026, by Ad Res Advanced Reinsurance Services GmbH for the Canadian Institute of Actuaries, recommending an updated set of mortality tables (CPM2024) for Canadian retirees and surviving spouses as of January 1, 2024. Relative to the CPM2014 tables it uses more than three times the experience data, applies mortality trends specific to Canadian pensioners, adds separate tables for surviving spouses, and offers Light and Heavy variants for plans whose experience differs from the industry average.

On the syllabus

  • Exam RET-301 — Outcomes 1a, 1b, 1c (Assumptions); Sections 1 and 8 only (CIA Report, Mar 2026; the syllabus names it "2024 Canadian Pensioner Mortality Research Project").

1 Executive summary

  • The updated tables reflect the mortality rates of Canadian retirees and surviving spouses as of January 1, 2024.
  • Eight innovations over CPM2014 are listed, among them: more than three times the volume of experience data; trends specific to Canadian pensioners used to bring the tables forward to 2024; separate tables for widows and widowers; Light and Heavy variants; and rates graduated with weighting by pension amount.
  • 1.1 Background
    • The CIA retained Ad Res in September 2021 to update the Canadian Pensioner Mortality tables; 48 pension plans contributed experience data.
    • The CPM2014 study examined experience from 1999 to 2008, with CPM2014Priv, CPM2014Publ and CPM2014 tables, based solely on retiree experience.
  • 1.2 Method
    • The data set holds close to 1.6 million individual records and over 300,000 deaths, against approximately 100,000 for CPM2014; survival models are parameterized with spline basis functions.
  • 1.3 Results
    • Four combined tables are created: male retirees, female retirees, male surviving spouses and female surviving spouses, with Heavy and Light tables by plan category and a single male surviving spouse table.
    • Annuity factors and remaining life expectancies are tabulated at selected ages for CPM2014 and CPM2024.
  • 1.4 Acknowledgements

2 Data and exploratory analysis

  • 2.1 Data received
  • 2.2 Data preparation
    • 2.2.1 Data excluded
    • 2.2.2 Study period
    • 2.2.3 Age range
    • 2.2.4 Data reconciliation and summary
  • 2.3 Exploratory analysis by risk factors
    • 2.3.1 Gender and participation type
    • 2.3.2 Disability retirement
    • 2.3.3 Pension amount
    • 2.3.4 Industry and workplace characteristics

3 Graduation method

  • 3.1 Overview of graduation methods
  • 3.2 Choice of method
  • 3.3 Parametric form
    • 3.3.1 Mortality evolution by age
    • 3.3.2 Mortality changes over time
  • 3.4 Producing mortality tables
    • 3.4.1 Concerning amounts-weighting
    • 3.4.2 Extension to younger ages

4 Modelling the combined experience

  • 4.1 Capturing subgroups
  • 4.2 Trends and COVID-19
    • 4.2.1 Base improvement trends
    • 4.2.2 Seasonal variation
    • 4.2.3 Excess mortality
    • 4.2.4 COVID-19 impact on trends
  • 4.3 Weighting by pension amount

5 Modelling by plan categories

  • 5.1 Defining categories
  • 5.2 Categories retained
  • 5.3 Creating Heavy and Light tables

6 Trends and table effective date

  • 6.1 CPM trends compared to industry scale
  • 6.2 CPM trends pre-2020
  • 6.3 CPM trends compared to general population
  • 6.4 Analyzing trends to the end of 2023
  • 6.5 Choice of table effective date

7 Final mortality rates

  • 7.1 Names for the mortality tables
  • 7.2 Charts of mortality rates
  • 7.3 Comparison of combined tables to CPM2014
  • 7.4 Impact relative to CPM2014

8 Discussion

  • The study results in four combined and six plan category mortality tables considered applicable as of beginning-of-year 2024, reflecting 2011 to 2021 experience with an allowance for excess mortality due to COVID-19 in 2020.
  • 8.1 Use of surviving spouse tables
    • The surviving spouse tables are conditional on the retiree dying before their spouse; no experience for contingent spouses was provided.
    • A practitioner might apply them solely to surviving spouses in pay and defer to the retiree tables for contingent spouse mortality while the retiree is assumed to be alive.
  • 8.2 Use of Heavy and Light tables
    • Heavy or Light may be warranted where the loadings that would be applied to the combined tables are significantly higher or lower than 100%.
  • 8.3 Public versus private sector
    • A distinction between public and private plans would not be useful in the CPM2024 study; the Heavy and Light plan category tables serve as more appropriate plan-specific benchmarks.
  • 8.4 Features of the data not used for tables
    • Disability status
    • Age at retirement
    • Marital status or pension form

References

Appendix A: Overview of survival modelling

Appendix B: Benefits of chosen graduation method

Appendix C: Age component

Appendix D: Time trend component

Appendix E: Seasonal and spikes components

Appendix F: Establishing the table effective year

Appendix G: Weighting by pension amount

Appendix H: Extension to younger ages

Appendix I: Producing the final mortality tables

Appendix J: Parameters under final models

Appendix K: Summary of statistical tests

Appendix L: Charts of model fit and residuals

Sources