Spring 2019 Q5 a)

Comments

• Graham

  October 2023

  I don't know which other exam questions you're looking at but a general answer to your questions are as follows:

  The approach you take can depend on several factors, including the data you have, the credibility of that data, and the specific risks you're trying to capture.

  • Trending the losses to a common level allows you to capture changes in cost structures over time, such as inflation. If you don't trend first, the historical data may not accurately represent current or future experience, potentially leading to inadequate rates.
  • Different questions or scenarios may use different methodologies due to variations in focus. Some might aim to capture excess loss trends more precisely, while others might prioritize other components, like base rates. It's about what you're trying to achieve with your ratemaking model.
  • Whether you need to develop losses to ultimate depends on the quality and completeness of your data and what exactly you're trying to analyze. If you have incomplete data, then developing to ultimate can provide a more accurate reflection of the likely final cost of claims. If the data are fairly complete and credible, you might opt to use undeveloped loss amounts.

  For the purposes of the exam, you just have to work with whatever data they give you and it should be fairly clear what to do based on the type of information they give you in the question. If you are unsure, just make sure to include a brief note explaining your reasoning. If you do that and your answer is logical, the graders should give you credit even if it might not have been the answer they were intending. Questions are sometimes ambiguous.