Long tail non life business and IBNER

[Dutch Delight]

We’re having trouble with our IBNER being larger than our history of settled claims; ie since some claims are open for a very long time they tend to get overrepresented among the open, reported claims. Furthermore, these long claims have proven themselves to be much more expensive than the settled claims.

Is there anyone who knows of a method which handles this problem? When I’m talking long tails I mean up to 30 years and even longer…

As it is now we are using the chain ladder method, but we have realised that we probably are underestimating of our claims reserve.

Someone who have looked into Schniepers method (1991)? Tried this, but came to strange results, maybe due to repayments (negative values).

I would be delighted for any kind of input! Being overseas, I might be a bit slow in replying any comments (sleeping while the rest of you are working ).