Setting assumptions

[Madiba]

I find that I always struggle to score highly when I am confronted with the following kind of question:
Pricing a new benefit (cancer, mental illness etc), in a country with no similar kind of benefit.
In this case I know that I would rely on reinsurance expertise (to assist in obtaining suitable data, tables etc), overseas data (although a lot of adjustments will need to be made), government statistics (hospital data, although would need to adjust for underwriting etc).
My question is, how would one make these adjustments in a real situation? e.g how would you adjust overseas data (I will appreciate a numerical illustration), how would you adjust hospital data for underwriting and other factors that could be a source of differences?
In what order would you make these adjustments in order to produce a table that you will use for pricing the benefit, what would your final rate table look like? e.g start off with a standard table, adjusted by using overseas data, then adjust using government stats?

 

[Steve McIntosh]

The actual detail of how you would make adjustments in a real situation will vary by company and depend on factors such as the situation given, purpose of the calculation and the data available. In practice, the adjustments may be subjective at times and the detail is likely to be commercially sensitive. However, the formula on page 20 of Ch11 does give one example of how overseas data may be used.

For the ST1 exam, knowledge of the actual mechanics of how you would make the adjustments is not required in detail. What’s more important is that you can generate a wide range of reasons WHY adjustments need to be made, and hence state the adjustments. A good way of doing this is to consider reasons why the data may not be relevant for use.

As a general rule for the order, I would tend start off with a source of data which is fairly CREDIBLE, in order to give me the detail I require (e.g. standard table or government stats), and then adjust to allow for data from sources that are more RELEVANT (e.g. own data).

I do sympathise, as these sorts of questions are tricky. However, continued practice will definitely help. ST1 exam Sept 05 Q8 (ii), Apr 06 Q6, Apr 07 Q9 are good examples. Once attempted, a read through the examiners reports (and/or ASET) will give you a good indication of points required and the level of detail that examiners tend to look for in these sorts of questions.

Hope this helps.