Increased Consistency is a Target for 2014

By Bruce Langstroth, FCIA

I’ve written about consistency before but mostly in conjunction with other topics. This month I want to focus on consistency alone.

When I talk about consistency, I’m talking about consistency in practice between different practice areas. Our Standards of Practice describe both general and practice-specific requirements, which take the form of principles and more specific prescriptions. But in neither case do the standards actually require there to be consistency between practice areas. Guidance material—educational notes and research papers—generally focuses on specific practice areas and rarely, if ever, attempts to articulate guidance with respect to multiple practice areas.

The result is that practice within two different fields will be consistent only by accident. To the extent that practice has evolved in different ways and serves fundamentally different purposes, this may be acceptable. But some of those inconsistencies may be difficult for the profession to explain and, in the extreme, may damage its credibility. Important stakeholders may have difficulty believing or understanding how practitioners can come to materially different conclusions.

An example of this issue is mortality improvement. Assumptions regarding rates of mortality and mortality improvements are made within pension, life insurance, actuarial evidence and (possibly) enterprise risk management practice. The purposes for which these assumptions may be used can include pricing, costing, funding, accounting, financial reporting, solvency, regulation, economic capital, and many others. In what circumstances and for what groups is it acceptable to use different assumptions, and where not?

To be more specific, a Canadian life insurer (with a defined benefit pension plan) will make mortality improvement assumptions in relation to its life and health insurance, annuities, and pension plan for the purposes of its own financial reporting. The Actuarial Standards Board has promulgated standards which impose an element of consistency on life and health insurance and annuities but not with respect to those lines of business and pensions. Does it make sense that those assumptions are different and, if so, where and when?

One point of view holds that there is a general underlying trend towards longer lives and that, in the absence of being able to demonstrate that different groups should have different rates of mortality improvement, we should assume the same trend for everyone. Presumably, if we hold this view, we would assess all available information and articulate an appropriate range of mortality improvement for all purposes based on that information.

Another point of view is that different populations can and should exhibit different rates of mortality improvement and that the assumptions ought to reflect those differences. This would imply choosing subsets of available information (that are most applicable to the populations in question) and developing assumptions from that data.

The problem is difficult to resolve because robust, credible data to assess rates of mortality improvement for different groups and reach a defensible conclusion do not exist. We fall short of being able to demonstrate that population mortality improvement is (or is not) different than that of pensioners, insured, annuitants, disabled lives, and other groups. In the absence of those demonstrations, we end up with different assumptions being made by individual practitioners, or groups of practitioners, based on the data and the views that they have brought to the assessment.

This is, of course, the sort of situation that calls for standards or guidance. As the Practice Council looks to 2014 it, and its committees, will be working with the Actuarial Standards Board and other stakeholders to attempt to increase the level of consistency and ensure that we as a profession are able to justify those consistencies that we agree should persist.

Bruce Langstroth, FCIA, is Chair of the Practice Council.

Canadian Institute of Actuaries/Institut canadien des actuaires