New mortality improvement scale
On April 28, 2017, the Canadian Institute of Actuaries (CIA) issued the draft Task Force Report on Mortality Improvement. The purpose of the task force was to construct “a mortality projection scale for the purpose of reflecting future mortality improvement in Canadian actuarial work.” The task force was composed not only of pension plan experts, but also actuaries from other practice areas, such as insurance.
The task force work led to the construction of a new two-dimensional mortality improvement scale, MI-2017. This scale assumes that historical mortality improvement rates will transition smoothly to an assumed long-term mortality improvement rate over a 10–20 year period, varying by age. The new scale is based on data up to 2015; the CPM-B, currently the most commonly used scale for pension plan valuations, is based on data up to 2007. The long-term mortality improvement rate was thus established taking more recent historical averages into consideration, but also using a long-term rate that is basically an average of the projection rates used for public plans in Canada, the United States and the United Kingdom. The CPM-B scale was based primarily on projections for the Canada Pension Plan.
Establishing an assumption for the mortality improvement rate requires a considerable amount of judgment. It must take into account a number of unknown variables, such as future technological and medical advances that will have an impact on the longevity of Canadian citizens. The uncertainty is even greater for the oldest ages (90 and up), for which little data is available.
Impact on life expectancy and the value of pension plan commitments
As shown in the table below, assumed life expectancy with this new scale combined with the base mortality table CPM2014 is about 1% - 1.5% higher than the life expectancy calculated using the more common CPM-B scale, for both males and females.
The impact on pension values would be about 0.5% (compounded by the application of any stabilization provision). Using the new scale would result in a similar increase in the value of plan liabilities.
Impact for actuarial valuations of pension plans
The report suggests that the new scale should be considered for most actuarial work in Canada (for actuarial valuation of defined benefits plans, but also for other actuarial work such as establishing insurers’ reserves). This means that sponsors of pension plans that are not fully funded on a going-concern basis could see an increase in their amortization payments and in their current service contributions. A similar impact could also be expected on liabilities and current service cost for accounting purposes if the new scale is also considered the best estimate on that basis.
It is important to point out that for the time being, the new scale would not be used to calculate transfer values payable upon member termination, since the assumptions used for these calculations are prescribed and based on CIA recommendations.
The same is true for the calculation of liabilities on a solvency basis for members assumed to receive transfer-value payments (members who have not retired).
Lastly, note that the report does not present results varying by population segments for which little data exists (socio-economic level, gender, smoker or non-smoker, etc.). However, the task force did conduct a literature review, which tended to indicate that the mortality improvement rate could vary by population segment. Although the report acknowledges that this could be reflected in certain circumstances, it argues that for most actuarial work in Canada, the proposed MI-2017 scale is a reasonable estimate of the mortality improvement rate.
CIA members have until June 30, 2017 to comment on the draft report. Depending on the outcome of the consultation, the results and impact on the value of pension plan liabilities could change in the final version of the report. The CIA might then revise its recommendations on transfer values to reflect this new scale.