CAS 2017 Fall 14(b)
Hi,
In the examiner report, the solution states that "Candidates were expected to correctly identify the underlying loss distribution given the graphed excess severities" but the solution is assuming exponential distribution applied to loss excess 1M.
Why is the underlying distribution exponential applied to loss excess 1M, not ground up loss?
Thanks!
Comments
The data we are given is grouped so all we know is losses above 1M follow an exponential distribution based on our calculations for e(x). The memoryless property of the exponential distribution allows us to extrapolate e(x) for values of x larger than 1M but it doesn't allow us to extrapolate in the other direction because we don't have more refined information.
In other words, since we have grouped data starting at 1M, we only know e(x) for x>1M rather than x>0 so we don't know how X behaves for 0 < x < 1M. It could be X is a mixed distribution where it's exponential after 1M and something else before hand. To avoid making additional assumptions, we apply the exponential distribution only to loss excess of 1M.
Hi, do we expect to draw any chart/plot in the worksheet during the exam?
No. Currently, according to the CAS we are not expected to draw any chart/plot in the worksheet during the exam. https://www.casact.org/sites/default/files/2024-03/CAS_Exam_New_Item_Types_FAQ_03_2024.pdf
You should definitely be able to sketch charts on your scratch pad as they can greatly help you answer questions.
We will keep monitoring the CAS for announcements on this topic and update you all as we learn more.