CMI Working Papers 21 and 22

The 00 series tables have been published in two CMI working papers. The mortality rates themselves are also available in spreadsheets. These final tables, together with full documentation of their derivation, will be published formally in a CMI Report in due course.

Please note that it is the responsibility of any actuary or other person using a published mortality table to ensure that it is appropriate for the particular purpose to which it is put.

The CMI will not be seeking approval for any specific projections of future mortality associated with the new tables. The CMI is evaluating two methodologies for mortality projections - P-spline and Lee-Carter - and is trying to explain the benefits and shortcomings of each. However it is not seeking approval for either methodology, nor does it rule out alternative approaches.

The development of these tables has been published in three working papers:

• CMI Working Paper 8 (Aug 2004), contained initial findings and proposals on which tables to graduate.
• CMI Working Paper 12 (Apr 2005), contained proposed graduations for the assured lives tables.
• CMI Working Paper 16 (Sept 2005), contained proposed graduations for the annuitant and pensioner tables.

The CMI has considered the feedback received on the draft tables and published two further working papers, containing the final 00 series base tables of mortality:

• CMI Working Paper 21. The Graduation of the CMI 1999-2002 Mortality Experience: Final 00 Series Mortality Tables - Assured Lives.
• CMI Working Paper 21. Mortality Rates.
• CMI Working Paper 22. The Graduation of the CMI 1999-2002 Mortality Experience: Final 00 Series Mortality Tables - Annuitants and Pensioners.
• CMI Working Paper 22. Mortality Rates.

Formula (1) on page 9 of Working paper 21 is incorrect. This Erratum (added 5 May 2011) gives the corrected formula.

The changes incorporated as a result of feedback to the draft tables are relatively minor:

• The final life office pensioner tables for normal retirements now start at age 20. As a result the rates between ages 50 and 65 have changed from the draft rates to reflect assumptions regarding the experience of ‘healthy’ lives rather than the actual experience.
• A more accurate method of deriving qx from the graduated formulae for µx has been used.
• The naming of some tables has been altered from the draft tables.

More detail on these changes is included in the working papers. In addition, feedback received that has not been incorporated in the final tables is documented in the two working papers, so that actuaries are aware of the concerns expressed that have not been reflected in the final tables, and can make suitable adjustment to the published base tables if they feel it appropriate to do so.

Note that the life office pensioner tables for those retiring early and the combined tables have not been extended below age 50, in the absence of any clear data. Extended rates for these tables were subsequently developed in Working Paper 26. However, the CMI is not seeking approval for these rates from the profession.