Asset Share TIA New Problems 6

[maddie]

In this problem, they adjust the avg losses for change in relativities, but I just can't figure out exactly the mechanic and thinking behind it all, even with the calculation shown for the second PY....
Can someone using TIA help me out with this?

[Dedicated111]

Well you have the first year Loss of 500 (which is provided). You have to trend this while also applying the "losses decrease 1% per annum as policy matures". So that gives you 500*1.1*.99. Then we have to adjusted by the change in loss ratio relativities. So if the relativity for year 1 is .96 and year 2 is .93. The change from year 1 to year 2 is .93/.99 (similar to our rating problems in W&M).

Now year 2 = 500*1.1*.99*.93/.96

[Dedicated111]

I'm not sure exactly what you were getting stuck on with this problem. But if you don't understand what .93/.96 means, then, umm..you may need to go back to some of the fundamentals of the Loss Ratio methods in the various W&M chapters. Hope my explanation helped.

[Dedicated111]

I was actually going back through a couple videos and rewatching some of them. You can find this problem worked out by Ken in the Competitive Strategy video (section G of the ratemaking videos)

[maddie]

Thanks!
I got confused with the 0.99, thinking it had something to do with the new business relativities, forgetting about that 1% decrease in loss costs. Which leads me to another question..
Usually, I would've divided by 1.01 instead of multiplying by 0.99. The wording of this problem is a little different though (improve by vs decrease), but then I noticed Ken used the 1.01 in the video. Do you think we would get away with both on the exam?

[Dedicated111]

You would want to divide. You get lucky here because 1/1.01 = .990099

[Vypa]

Originally Posted by Dedicated111

You would want to divide. You get lucky here because 1/1.01 = .990099

I believe there is some ambiguity here as Feldblum multiplies by (1 - loss cost improvement) in the Discount example in the paper. Elsewhere he divides by (1 + loss cost improvement). It's not consistent.

[intelmic]

Originally Posted by Vypa

I believe there is some ambiguity here as Feldblum multiplies by (1 - loss cost improvement) in the Discount example in the paper. Elsewhere he divides by (1 + loss cost improvement). It's not consistent.

+1 on inconsistency.