NCCI Experience Rating - B & W for Multi-State Risk

[erickson]

Context

Section 3 on Page 15 of Part Four.

Weighting Value: Each state's weighting value (W) shall be determined based on the total (i.e., multi-state) expected losses of the risk. Use the Table of W and B for each state.

. . . then calculate the average (multi-state) W according to steps 1-3.

Ballast Value: There is no mention of determining each state's ballast value (B) based on the total expected losses of the risk.

The only discussion regarding B involves calculating the average (multi-state) B according to steps 1-3.

Question

Since we are explicitly told to use the (multi-state) total expected losses when looking up each state's W from the table, but we are not told to use the (multi-state) total expected losses when looking up each state's B from the table, it's my interpretation that each state's B should be based on only that state's expected losses.

This doesn't seem ambiguous to me. However, certain All10 solutions (e.g., 1995 #41, 1996 #25) use values for B based on total expected losses.

Any comments?

[Dooby Scoo]

I too have noticed this. I have decided to go with what the All10 has done. I know that's a bad idea, but there isn't explanation of this anywhere in the manual.

And if the All10 did it this way, that means they copied it from the exam solution itself.

If anyone else has any bright ideas, please do share them.