B-F Method Key Assumptions

[skaren9]

According to Friedland, claims reported to date provide no information on claims not yet reported.

Don't we have to select the CDFs and convert them to percentages unreported based on claims reported to date? I understand that claims reported to date provide no info on the loss ratio, but that's only one part of IBNR component.

Thanks!

[Present Value]

The CDF in any case doen't include the latest AY. And for all you care, it could come from a completely different process. Friedland was implying more about the magnitude of the IBNR relative to the magnitude of what has been reported to date. Think Hugh White's question.

[Vorian Atreides]

To add to what PV said . . . the CDF's are not dependent on how much is reported/paid for a given AY; the calculations are simply a determination of how claims are actually paid out to get to Ultimate.

The BF still assumes that Expected Claims is the (best) estimate of Ultimate Claims for immature AY. Deviations of the actual development from the expected development are simply a reflection of volatility in losses between AYs and doesn't give rise to a need to adjust IBNR estimates.

Also agree that the Hugh White questions are key to understanding the role the different methods play in addressing reserve estimations when a large loss is realized.