No-Charge Auto Insurance Pioneer Out to Sell More VWs, Convince Skeptics Rating Model Works
National News • February 1, 2005
Volkswagen of America, looking for ways to attract younger buyers, is going beyond sporty designs, cash rebates and zero percent financing incentives to offer car buyers, regardless of their risk ...
Insurance Journal is not responsible for the content of the message below.
Subject: insurance risk, credit risk, and rating models
Posted On: February 2, 2005, 12:54 pm CST
Posted By: Jared Yanowicz
Comment:
The previous posts have some good points. Although I have never worked in personal auto, I have some experience with multivariate models applied to insurance, and I believe I can explain the conditions under which Nationwide will and won't profit from this. However, I don't claim to be an expert in this area.
If Nationwide (or their consulting company) are simply only looking at one variable, the loss history associated with the model of car, and plan to charge rates based only upon the loss history of the model, they are likely to lose money. As the other posts have pointed out, younger, higher risks drivers are likely to gravitate to Volkswagen.
However, here is one possibility. Assume, for simplicity, that only two factors affect losses - age and car model. A multivariate model is estimated where losses are a function of those 2 variables:
L=F(Age, CarModel). The model is able to isolate the effect that both Age and CarModel have on losses.
Nationwide knows that the average age of the driver will change, but is confident that they have isolated the effect of CarModel. Moreover, Nationwide assumes that they can accurately estimate what the new age distribution will be. Using the effect that the car model has on losses from the multivariate model and the estimated age distribution, Nationwide is now able to successfully price this new Volkswagen program AS A POOL and charges enough to make a profit.
Another possibility is that Nationwide knows that they will lose money during the first year of the program. However, after the first year, they plan on dropping the bad risks and keeping the good ones. Many of these good drivers may not have even considered Nationwide if they had not been drawn in by Volkswagen's promotion. Since we are dealing with a pool of mostly young people, if the drivers that are not dropped stay with Nationwide, they may enjoy many of years of profitable business from them. From Nationwide's perspective, they would hope that they would make enough money off of the good drivers in years 2,3,4...etc to more than make up for the money that they lost in year 1.
Now comes the issue of credit risk. Some of the previous posts have raised the issue the some of the buyer's under this program may get dropped after the first year an might become unable to pay for insurance. This, in turn, could prevent them from driving, which may diminish their desire or ability to make car payments. This may be a problem for the lenders, and it may be wise for them to tighten their lending standards on Volkswagens to avoid this problem. However, Nationwide will have no exposure to this risk (unless they are also providing the auto loans), and it will probably not effect their behavior or profitability.
Subject: insurance risk, credit risk, and rating models
If Nationwide (or their consulting company) are simply only looking at one variable, the loss history associated with the model of car, and plan to charge rates based only upon the loss history of the model, they are likely to lose money. As the other posts have pointed out, younger, higher risks drivers are likely to gravitate to Volkswagen.
However, here is one possibility. Assume, for simplicity, that only two factors affect losses - age and car model. A multivariate model is estimated where losses are a function of those 2 variables:
L=F(Age, CarModel). The model is able to isolate the effect that both Age and CarModel have on losses.
Nationwide knows that the average age of the driver will change, but is confident that they have isolated the effect of CarModel. Moreover, Nationwide assumes that they can accurately estimate what the new age distribution will be. Using the effect that the car model has on losses from the multivariate model and the estimated age distribution, Nationwide is now able to successfully price this new Volkswagen program AS A POOL and charges enough to make a profit.
Another possibility is that Nationwide knows that they will lose money during the first year of the program. However, after the first year, they plan on dropping the bad risks and keeping the good ones. Many of these good drivers may not have even considered Nationwide if they had not been drawn in by Volkswagen's promotion. Since we are dealing with a pool of mostly young people, if the drivers that are not dropped stay with Nationwide, they may enjoy many of years of profitable business from them. From Nationwide's perspective, they would hope that they would make enough money off of the good drivers in years 2,3,4...etc to more than make up for the money that they lost in year 1.
Now comes the issue of credit risk. Some of the previous posts have raised the issue the some of the buyer's under this program may get dropped after the first year an might become unable to pay for insurance. This, in turn, could prevent them from driving, which may diminish their desire or ability to make car payments. This may be a problem for the lenders, and it may be wise for them to tighten their lending standards on Volkswagens to avoid this problem. However, Nationwide will have no exposure to this risk (unless they are also providing the auto loans), and it will probably not effect their behavior or profitability.