ICRLL
Do you think it's possible a question could come up in a format for the second advantage of ICRLL - adjusting one Table M to make it for another Table M on LOB with less data? If so, what adjustments would that make to the ICRLL example?
Do you think it's possible a question could come up in a format for the second advantage of ICRLL - adjusting one Table M to make it for another Table M on LOB with less data? If so, what adjustments would that make to the ICRLL example?
Comments
It's possible but probably unlikely in our opinion. In the situation you describe the state/hazard group factor would account for most of the differences between lines of business and then the thin data would introduce more variation in the losses so a higher insurance charge would be needed. So instead of shifting to a higher adjusted expected loss, the ICRLL would shift to a lower expected loss to pick up the additional volatility.