In part a you generate an AEP curve for the given data. In part b, we want to simulate the total insured loss (the aggregate loss) by looking at the 86th percentile of the AEP curve.
Since the AEP curve was generated using three discrete loss events, the curve itself is discrete/has step-function like behavior. So the aggregate loss is 0 83.79% of the time, then jumps up to 10m for 0.931 - 0.8379 = 9.31% of the time etc. We can't get a loss amount in between say 0m and 10m because each loss event either happens or doesn't.
So the 86th percentile loss is 10m because 0.86 > 0.8379 and 0.86 < 0.931.
In part a you generate the AEP curve using the three events given. You're told to simulate using the random number 0.86. This means finding the 86th percentile of the AEP curve.
The AEP curve is 0 until 0.8379 and then jumps up to 10m between 0.8379 and 0.931. Since 0.86 lies in this range, we choose 10M.
Comments
In part a you generate an AEP curve for the given data. In part b, we want to simulate the total insured loss (the aggregate loss) by looking at the 86th percentile of the AEP curve.
Since the AEP curve was generated using three discrete loss events, the curve itself is discrete/has step-function like behavior. So the aggregate loss is 0 83.79% of the time, then jumps up to 10m for 0.931 - 0.8379 = 9.31% of the time etc. We can't get a loss amount in between say 0m and 10m because each loss event either happens or doesn't.
So the 86th percentile loss is 10m because 0.86 > 0.8379 and 0.86 < 0.931.
Hi! Could you please explain why we choose 10M instead of 0M?
In part a you generate the AEP curve using the three events given. You're told to simulate using the random number 0.86. This means finding the 86th percentile of the AEP curve.
The AEP curve is 0 until 0.8379 and then jumps up to 10m between 0.8379 and 0.931. Since 0.86 lies in this range, we choose 10M.