LDF- IBNR Question

[drake835]

I have a project I'm working on that is an accrual and forecast for a general liability claims made policy. The problem is I only have occurrence based ldf''s. This does not have to be specific, it is really more of an estimate. However, I need to remove the IBNR portion of the LDF. I am wondering what do you guys/gals think is a good pick (out of thin air is fine I guess) for percentage of development that is attributable to IBNR. I assume it varies by year so let's say for each year going back 4 or 5 years.

Thanks for you thoughts and assistance.

[stoodawg]

You could use industry LDF's to obtain an estimate

[Levin]

What is this accrual and forecast thing? I'm in P&C in the US, and those aren't terms I am familiar with. Sounds like reserving or pricing, but just curious about the terms themselves and the context they came up in within your class.

[MountainHawk]

Do you have development factors for counts?

[KidCA]

For a mature claims made policy, what matters is the rate of development and the latency of reporting. These factors vary broadly for underlying coverage and exposure, or, line of business and industry.

For GL, say for retail exposures, you have 95% of all claims reported within the 12month policy year. The other 5% is due to some reporting lag, but probably the same type of claims--e.g. slip and fall. If I were converting my Occ LDFs to CM, I don't think I would see much of a reduction at all. Maybe use (LDFocc-1)*Reduct_%+1, where the reduction may be 97.5% in the second year CM.

BUT, I'm skeptical if this scenario is your case. CM is written to reduce premium costs and limit liability to the next 12 months where the current environment (injury and legislative) is better known. Why convert to a CM policy when there is no latent exposure? If there is some severe and probable covered exposure, say mold, then usual GL factors don't make sense anymore to use. You may just want to use another source, like AM Best's "Other Liability-CM" triangles.