VHM and EVPV - Brosius

[Infinity]

Brosius interprets VHM as variability from loss occurrence process and EVPV as variability from loss reporting process. Can anyone explain why this is true?

[antisyst3m]

EVPV measures the expected value of the variance of the reported losses in the future period given our reported losses to date. The inner-term is measuring variance around the reported claim component, so this is the easier of the two to understand.

VHM measures the variance around the expected future period's reported loss given reported losses to date. So, if we are calculating the expected value of next year's reported loss on our pollution/asbestos claims given our prior reported losses, our prediction is subject to a lot of variability because many claims haven't occurred yet. So, we are trying to use the past to predict the future when we don't have a complete picture of the past.

Hope that helps.