Magnitude of Claims
This might be a bit beyond the scope of the exam but it got me thinking as I was reading through this article.
The paper states on page 5:
"studies show no significant difference in the magnitude
of claims that are filed, but only of the frequency of the claims. [...]
The studies show only that consumers with lower credit scores file more claims, not
that they have greater loss events. It is quite possible the frequency of insured loss events is the
same across populations, but those with higher scores are less likely to file a claim. This may be
because wealthier individuals (with higher credit scores) may not file a legitimate insurance
claim for a broken window or for minor fender bender, instead electing to pay the repairs
themselves so as not to impact their claims history."
This seems to be implying higher credit score individuals may report fewer low severity claims. However, wouldn't this cause the magnitude of claims filed of high credit score individuals to be higher (since they only report high severity accidents)? Which would not align with the finding of no significant difference in the magnitude of claims. The explanation in the text just doesn't make sense to me.
Comments
If you assume that the empirical studies that are referenced compared the size of loss from the two populations in specific strata (rather than comparing the average loss of insureds from the two populations), then the statement makes sense.
I'm sorry it still isn't clicking for me. What do you mean by specific strata?
I mean "claim size buckets." You are extrapolating that both cohorts are bound to have large claims. The research may have compared both the large and the small claims of the two cohorts, and found both types to be of similar size in the two cohorts. Hence the statement that there is no claim magnitude difference.
If both cohorts reported large claims with the same frequency, then you would be right that the average claim of the high-credit-score cohort would be higher. But the more pertinent outcome is that the low-credit-score cohort will have higher aggregate loss cost, due to its higher small claim frequency. By bringing up claim magnitude parity, they are arguing that the low-credit-score cohort is not inherently of higher risk, despite the fact that they may show higher overall loss.
Okay I think think it is making sense to me now. I appreciate the detailed explanation! Thank you!
Sure, good luck.