Insurance question

I wasn't sure where to post this question. This subforum seemed like it might be the most knowledgeable.

I've been trying to reverse engineer Bruce Berkowitz's investment thesis in AIG.

A common criticism of AIG I see is that it historically underpriced and underreserved its property & casualty business. It is true that the company's initial estimates of reserves were inadequate. However, this appears largely due to its exposure to business lines with long tail risk and that are difficult to price: excess worker's comp and casualty, D&O, etc., in addition to continued payments owed on asbestos contracts written some 25 years ago.

My question is as follows. Assuming it is true that AIG is shifting to business with less tail risk (and assuming that it reserves its remaining lines properly), eventually you should see the numbers begin to improve as the losses associated with this tail risk phase out. How does one determine the "half life" of these bad contracts AIG wrote? I checked the financials but didn't see a disclosure related to this (although it is likely I might have missed it).