Table talk
When valuing a portfolio, an actuary must often make a decision as to what tables to use for the risk. The ideal is to use tables which have been created from the portfolio's own experience, preferably using a statistical model to account for the various risk factors. It is even possible to generate an individual mortality table or survival curve for each member of the portfolio.
Not every portfolio is large enough to warrant such an approach, of course, and the actuary will have to fall back on some kind of standard table. But how to pick which one? In theory, it doesn't matter much which table is used due to the equivalent-reserve calculation described in the SIAS paper, "Financial aspects of longevity risk". This method avoids the distortions which can arise from doing a crude "actual v. expected" (A/E) comparison of what experience data there is against a standard table.
However, in practice it might not always be possible to do a proper equivalent-reserve calculation. For example, a full valuation run might be quite expensive in terms of time and resource, and the only option available to the actuary might be an equivalent-reserve calculation for a specimen policy. This exposes the actuary's calculations to a number of second-order risks and effects which it is good to minimise. Our recommendation for picking a standard table is therefore twofold:
1. Use a table where the applicable percentage is as close to 100% aspossible. Thus, if the equivalent-reserve calculation suggests 80% of table A or 105% of table B, the latter is preferable as it limits second-order effects.
2. In pensioner valuations, use as recent a table as possible to catch the correct shape of the cohort effect. In the UK, it is perhaps moot whether a SAPS table is better than PCA00 due to the small number of years separating them. However, either table is clearly preferable to any modification of PA92, which is now over fifteen years old.
Actuaries also need to be aware of presentational aspects, too. Actuaries using the latest tables will project an image of being up-to-date. This is not just important for clients, but will also help keep a regulator's mind at ease.
Previous posts
Size isn't everything
Great Expectations
When fitting statistical models, a number of features are commonly assumed by users. Chief amongst these assumptions is that the expected number of events according to the model will equal the actual number in the data. This strikes most people as a thoroughly reasonable expectation. Reasonable, but often wrong.
Add new comment