Siewart credibility weighting method

[Phil de cube]

What if you wanted to credibility weight the implied development IBNR compared to the expected/loss ratio IBNR? Which LDF would you use, limited or ultimate for the Z factor?

One of the advantages is being able to use direct or indirect development, so I was wondering how the indirect one would work.

Thanks.

[Dash.]

I interpreted that to mean that you would apply the BF method to the Limited Method, and the BF method to the Unlimited method, and take the difference of the two BF methods to get the "Implied" BF method.

[Archen]

Good interpretation.

[Phil de cube]

I interpreted that to mean that you would apply the BF method to the Limited Method, and the BF method to the Unlimited method, and take the difference of the two BF methods to get the "Implied" BF method.

I see expected loss IBNR times (1-z) in second part and I see a possibility of putting the direct IBNR or the implied IBRN in the first part of equation.