Credit

Question 1

What is Kucera’s stance on use of credit in rating?

Answer 1

-credit helps insurers segment risks
-actuarially sound rates are charged
-distribution of premium changes, but not overall premium.

Question 2

What is an insurance score?

Answer 2

-numerical score assigned to a risk based on its characteristics
-provides use to pricing and UW
-relative measure of E[Loss]

Question 3

What is a credit-based insurance score?

Answer 3

-use of items in credit report (# of inquiries in opening new accounts or 30 days past due charges)
-also can use third party models
-strong correlation bw loss and credit score

Question 4

What are some ways insurers use credit?

Answer 4

-criterion to accept / deny a risk in UW
-segment into risk tiers for rating (most common, say 600-700 score is the base relativity)

Question 5

ASOP 12 justification of using credit

Answer 5

-actuary should use risk characteristics related to expected outcomes
-relationship is demonstrated if variation in loss correlated to the risk characteristics
-several studies surrounding credit demonstrate this

Question 6

Why should credit be used?

Answer 6

SMOR
statistically significant
hard to manipulate
objective
removal doesn’t change the overall premium level

Question 7

Economic downturn justification

Answer 7

-regulators worried if credit goes down, premium goes up unwarranted
-can apply off-balance factor to charge correct aggregate premium
-distributional shift has little impact on renewals
-competition motivates companies to continuously review rate differences (not charge more for worsening credit, if not correlated with loss) - however, this likely would not reflect in the data fast enough

Question 8

What type of insurance does McCarty’s paper focus on?

Answer 8

personal lines insurance

Question 9

Other rating factors considered inappropriate

Answer 9

life insurance (whites live longer, counter to equal protection for consumers)
-genetic testing for predispostion of inherited disease (outlawed in 1996)
-McCarty is sensitive to any proxies for race, ethnicity, religion, or socioeconomic class

Question 10

Are credit reports often accurate?

Answer 10

No, 2000 study in consumer reports showed half of credit reports contain errors - identify theft is a concern

Question 11

What types of people are disproportionately affected by credit? (negative)

Answer 11

-recent divorcees
-recently naturalized citizens
-elderly
-disabled
-strong religious convictions
-young people with little credit history

Question 12

Why should credit not be used?

Answer 12

FEED
only correlated with frequency (not severity)
over half of credit reports contain Errors
in economic downturns, credit worsens and doesnt correlate with loss
discriminatory towards youth, elderly, religious, immigrants, etc

Question 13

What are the Florida DOI commissioners thoughts on predictive power of credit?

Answer 13

correlation is most likely just indirectly measuring socioeconomic status

Question 14

Why did insurance industry oppose Florida regulatory office limiting credit use?

Answer 14

-didnt have authority to prevent use as UW tool
-no authority to define unfairly discriminatory
-no data to demonstrate these claims
-disproportionate impact definition was too vague

Question 15

McCartys Conclusion (against credit scoring)

Answer 15

-negatives outweigh the positives
-concerned about other tools being adopted that are similar - other proxies for race (like occupation and education)
-disappointed with 2007 FTC report, not objective analysis

Question 16

How have states responded to credit scoring?

Answer 16

-48 states have somewhat limited use of scoring
-some have to notify consumers of use, give regulators access, restricting actions based solely on the model
-some states have banned it, banned to cancel / non-renew, not allowed to use as sole UW decision criteria