CAS Predictive Model

Question 1

Potential benefits of predictive model in insurance?

Answer 1

Reveal Insights into insurance costs
Encourage better risk management
Lower the cost

Question 2

What is the purpose of regulatory best practice for predictive models?

Answer 2

Support parameter evaluation
Improve model understanding
Provide a base line for the regulator

Question 3

How doe best practice help state regulator and industry?

Answer 3

Guide the regulator
Identify the key element
Faciliate resource sharing
Improve model review quality
Aid in training

Question 4

How do regulators benefit from the best practice?

Answer 4

Enhance market speed and competitiveness by improving reviewing quality

Question 5

Core characteristics a model should have

Answer 5

If it’s cost based
If it complies with the state law
If it improve the rating plan

Question 6

Review items on rating factor

Answer 6

Comply with the state law
Not disrupting premium ( no premium swing)
Produce fair rate

Question 7

Review items on data and model

Answer 7

How model was built
Accuracy of input data
Adjustment of raw data
Update frequency of the risk characteristics (update data source)

Question 8

Review items on insurance market place (less material)

Answer 8

Enable competition
Protect model confidentiality
Allow timely review to ensure speed to market