Vypa
03-18-2012, 01:24 PM
Hi.
This question is regarding loss cost improvement as policy durations increases in Feldblum's Asset Share Pricing paper.
For all but the "Discount" example provided in the paper, improvement to loss costs (if it exists) as policy duration increases is reflected by dividing the preceding policy year's loss cost by (1 + loss cost improvement).
For example if loss costs for Yr. 1 are $1,000 and the paper states that there is a 3% improvement in each subsequent policy year, for policy year 2 the paper shows the $1,000 original loss cost divided by 1.03, or $970.87 (assuming a 0% loss trend).
However, in the "Discount" example, the improvement to prior year's loss cost is not divided by (1 + loss cost improvement), rather it is multiplied by (1 - loss cost improvement).
For example if loss costs for Yr.1 are $1,000 and the paper states that there is a 1% improvement in loss costs in each preceding policy year, for policy year 2 the paper shows the loss cost as $1,000 x (1-.01), or $990 (assuming a 0% loss trend and no loss ratio relativity improvement).
This same apparent inconsistency exists in the solutions to past exam problems. For instance problem 2006 - #50 divides the prior loss costs by (1 + loss cost improvement) where 2008 - #44 multiplies the prior loss costs by (1 - loss cost improvement).
Am I missing something big here or is there something more subtle at play?
Thanks for your thoughts!
This question is regarding loss cost improvement as policy durations increases in Feldblum's Asset Share Pricing paper.
For all but the "Discount" example provided in the paper, improvement to loss costs (if it exists) as policy duration increases is reflected by dividing the preceding policy year's loss cost by (1 + loss cost improvement).
For example if loss costs for Yr. 1 are $1,000 and the paper states that there is a 3% improvement in each subsequent policy year, for policy year 2 the paper shows the $1,000 original loss cost divided by 1.03, or $970.87 (assuming a 0% loss trend).
However, in the "Discount" example, the improvement to prior year's loss cost is not divided by (1 + loss cost improvement), rather it is multiplied by (1 - loss cost improvement).
For example if loss costs for Yr.1 are $1,000 and the paper states that there is a 1% improvement in loss costs in each preceding policy year, for policy year 2 the paper shows the loss cost as $1,000 x (1-.01), or $990 (assuming a 0% loss trend and no loss ratio relativity improvement).
This same apparent inconsistency exists in the solutions to past exam problems. For instance problem 2006 - #50 divides the prior loss costs by (1 + loss cost improvement) where 2008 - #44 multiplies the prior loss costs by (1 - loss cost improvement).
Am I missing something big here or is there something more subtle at play?
Thanks for your thoughts!