Overlap Fallacy
I understand the difference between trending and development. So, if we have a loss triangle, should we trend the losses and then calculate development factors? (As opposed to calculating development factors, then trending losses)?
It looks like you're new here. If you want to get involved, click one of these buttons!
I understand the difference between trending and development. So, if we have a loss triangle, should we trend the losses and then calculate development factors? (As opposed to calculating development factors, then trending losses)?
Comments
Theoretically it shouldn't matter whether you apply trending or development first, but for practicality, it's probably simpler to do development first, then trending.
Reason:
If you applied trending first, you would have to do it for every cell in the development triangle, which would be more work than if you just applied the trend to the ultimate for each accident year.
Further Explanation:
It largely depends on the specific use case and the assumptions you're making. Generally speaking, calculating development factors before trending the losses is the more common approach. This allows you to understand the "pure" development pattern of losses without the influence of inflation or other time-dependent factors.
Once you have your development factors, you can then apply trending to project these losses into the future. This way, the trending accurately reflects changes in external conditions, like inflation, without getting mixed up with the internal development of the losses themselves.