Why are insurance companies' actuarial models so out of whack with one another?

Factual Questions
1

As anyone who’s ever shopped around for insurance (i.e. everyone that can be trusted with a knife and fork) knows the insurance quotes you get for all kinds of things are way, way, out of line with one another. For instance for car insurance I have usually found that the Admiral Group (Admiral, Bell, Elephant, some woman only thing I forget, etc) works out great for me over here. But even within Admiral, who presumably use very similar models within its group, I quite often switch the insurer every year for a car because one will come out better than the others when I do my checking (and it’s not just a lazy existing customer fee either, as I say other companies within the group will quote differently).

Admiral hasn’t been my only insurer, mind, and I have had quotes from many. And the weirdest thing is when I compare a significant quantity of sites. I can get prices, quite literally, from circa £500 a year, not bad with my car, to £35 000!!! Even if I ignore the prices that make very little sense and are presumably there to say “we don’t wanna insure you”, £500-£1500 is a pretty reasonable scale.

It seems to me that there is something weird going on here and at the majority of it has to be with inadequete risk assessment, somewhere. Admiral insured me at the age of 19 with a 4 litre beast, for about a quarter of the price a 1 litre ford fiesta would have been with them, so clearly they’ve got something clever going on amongst their internal calculations (and I have never made a claim on insurance, nor do I ever expect to) but in general presumably insurance companies will not come up with quotes that they can expect to lose money on, so what’s up with higher quotes?

All a long winded way of restating the question again. I remake it:

If an insurance company sets its quotes too high, then no one is going to buy insurance from it. If it sets its quotes too low, it will lose money when people do business with it. Neither it, nor its competitors’ actuaries have that many variables to consider when quoting retail insurance - we’re basically talking car, driving record, location, demographics (but no all demographics for legal reasons). So why are quotes so far out of whack with one another? Especially when this is an obvious place for the Winnner’s Curse to apply to the insurance company’s quotes - suggesting that there is no systematic problem with overly low quoting.

2

I think it’s mostly because they place different values on the information they have, they have different information, and they have different methods of managing overall risk. For example, if one company has a lot of insured divers in one town, they may increase the odds they will have to cover both sides of an accident. Another example might be if has paid out several claims near where you live, they might assess the geographic risk differently.

3

Premiums aren’t controlled solely by the actuarial departments. Marketing also has some say in it, for instance.

4

AMNAA (I am not an actuary) but based on my impressions as a insurance customer over time it seems auto insurance companies often tend to be focused on certain lines of business they are familiar with with respect to age groups, risk types etc, and have developed profitable models for those cohorts. If you fit within those actuarial comfort parameters you will be quoted reasonable rates, if you are outside those parameters you will quoted a punitive “go away” rate.

This can lead to oddball circumstances such as “safe” drivers being quoted expensive rates by companies that specialize in risky drivers. They really don’t want your business. It’s not their bread and butter.

5

I was speaking to a referral partner who handles insurance and retirement planning. He was explaining why his company’s business insurance was better for some industries than others. Basically, the company decided that they wanted x% increase in total revenues to cover their total cost. Corporate’s original plan was to raise all industries up, but the guys involved in sales crunched some of their own numbers to show that certain industries were not only cheap in terms of claims, but also cheap in terms of servicing costs (like billing and customer service). So the sales guys convinced corporate to shift most of the rate increases to the industries that generated more of the costs.

Now… I have no evidence this is the true story rather than just a sales line, but it seems rational and I can see how it could be applied to different demographics of auto insurance.

6

A lot of it has to do with the way rating categories are grouped.

This is one of the most funadamental parts of insurance, and what distinguishes insurance from primitive mutual aid societies.

All insurance companies group things based on different levels of risk, but each has its own groupings and its own dividing lines, and exactly where the dividing lines are can make a huge difference.

[Auto insurance companies - at least in the US - are constantly advertising that the average person who switched to their company saved X%. And it’s true, but meaningless. Even if the overall rates are the same for Company A as for Company B, there are any number of people who are cheaper for Company A, and a corresponding number who are cheaper for Company B. A high percentage of these people will switch to the company which is cheaper for them, and both Company A and Company B get to run ads proclaiming that people saved money by switching to them.]