The Value of the FICO Score in Insurance Underwriting

high and low fico score meter and a car graphic

When Rashida Tlaib, Democrat candidate for 13th Congressional District, Detroit, Wayne County, Michigan, won her bid for the U.S. House of Representatives, she described to her constituents she would be going to the Congress to help throw out the (insert expletive) President of the United States. Since being sworn in, she has set off several firebrand episodes where her unique style has left her cohorts on their heels. She along with others have made comments about the role of government which if embraced would change the paradigm for government involvement.

One case in point is her concern for car insurance rates in her district. They are among the highest in the nation, which she attributes to the method used by insurance companies to determine premium rates. She has specific concern for the use of FICO scores as part of the underwriting process.

This is a credit score generated by a data analytics company, Fair Isaac Corp. (FICO) established by Bill Fair, an engineer, and Earl Isaac, a mathematician, in 1956 based in San Jose, California. It is based on “predictive analytics” which means they take information and analyze it to predict what’s likely to happen. It is estimated 90% of all lending is affected by the FICO score. Thirty-five percent of the credit score is generated by payment history, 30% from amount of debt, 15% from length of credit history, 10% from new history, and 10% from credit mix (www.credit.com™, Inc.). The score algorithm uses these five elements to generate the score a consumer would have. Just how these elements are mathematically used is ambiguous to the consuming public–the score itself definitely impacts credit being rendered. Since the early 1990’s, it also had a major impact on insurance companies using it to identify which customers are least likely to file claims.

The irritant to Ms. Tlaib is in the element of the credit mix. Because of where they live, people have difficulty establishing good credit sources like mortgages and bank installment loans which have the tendency to lift a person’s credit score, while credit cards, department store cards, gas cards, etc. are not favorably looked upon when determining FICO scores. She has the impression that the practice of insurance companies using credit mix scores discriminates against her constituents since they cannot easily access the type of loans which generate better credit scores.

Subsequently, they don’t reflect accurately the risk a driver may be to an insurance company.

One has to ask with the other elements for establishing rates, i.e. deductibles, what type of car you drive, how often and how far, driving record, age, sex, marital status, and other “add ons” considered, and with credit mix only representing 10% of premium, how much is it really affecting her constituents. Her concern just may be another case of finding something to justify her individual existence.

As a point of interest, my insurance career started in 1974 with a very reputable casualty company whose loss percentage was around 70%. Our underwriting questions for automobile coverage was age of drivers, distance to work, type of vehicles, conviction for DUI, reckless driving, any existing damage to vehicle, and limits desired for coverages. There was no sending request from State Motor Vehicle Department regarding driver license, verification of ages, or number of citations. As time went on, those records were being required, every household occupant had to be listed, pictures of vehicles had to be submitted, credit scores became a requirement, number of tickets and accidents had to be accounted for. Fast forward to 2015 when I retired, the loss percentage was still in the low 70%. All these predictive analytics just haven’t proven to be influential in establishing rates or a method to reduce claims. It makes one wonder if just knowing how one drives, social habits like drinking, smoking, or using drugs and eliminating some of the other underwriting requirements would impact the premiums people are paying for their privilege to drive. Ordering FICO scores and MVRs (motor vehicle records) every six months can’t be cheap. If they don’t seem to impact overall costs, the observation would be maybe it’s time to move on to another way of underwriting; or are insurance companies caught in the army worm scenario where if you place an army worm on a glass rim and place another directly behind until the rim is covered they will continue to march around just blindly falling the one in front. Maybe some of the new upstarts will not just get on the glass rim and start marching. Maybe!

Be the first to comment

Leave a Reply