Insurance quote going from £800 to 950 to 1700!

SouthLondonUser

Aylesbury_Duck wrote: »

You're presumably using someone else's address to obtain dummy quotes? How can you be sure that doing so doesn't place some sort of marker on that address, pushing up genuine quotes from that address in future?

The one and only thing I am sure of is that insurance pricing algorithms are black boxes which often provide totally nonsensical prices. I have had cases of huge variations in prices for minor changes, like one of the two identical cars costing more for no reason (the only difference was the color, blue vs red), or big differences in prices for minor variations in the job description (both of which describe my job perfectly well), or insurance being cheaper if a vehicle is declared as kept on the road than in a garage, etc.

Aylesbury_Duck wrote: »

Yes, sharing a flat with someone with a history of claims will probably affect you because postcodes are risk-assessed. Lots of claims from a particular postcode area suggests the policyholders in that area are higher-risk clients, therefore their premiums will be loaded accordingly.

Claims, I can understand. But quotes? An address with a history of getting many quotes means... what exactly? Could you please clarify? It seems to me you are confusing claims and quotes!

SouthLondonUser

elsien wrote: »

As all your cheap quotes were via comparison websites, have you tried doing the same with your correct details instead of going direct to the insurer?

Yes, at the end, with Confused.com. I even entered my driving licence details and I got a quote for £620. I did not get other quotes with all my correct details on the other websites.

I can understand a quote with the correct details being cheaper than one with fake data. But, again, I cannot think of any possible logical explanation why, using the very same identical data, quotes should vary as much as they have. Well, no explanation other than the fact that pricing algorithms are unreliable black boxes which spit out nonsensical results, of course. Does any one else have any idea?

I can understand insurance companies would run checks on you if you declared, say, 4 points last week and 1 point now. But the problem is that there are also lots of perfectly legitimate reasons to 'test' different quotes: eg there is nothing wrong in trying to see how different the price would be if wife or husband is the main drivers (yes, I know fronting is illegal, but it's not fronting if a couple genuinely use the car together 99% of the times), if you keep it in a garage, if you install a tracker, etc. I don't do these 'tests' with my real data because I am afraid the insurers might somehow use it against me. What I do is basically use fake data to narrow down my list of options.

Aylesbury_Duck

SouthLondonUser wrote: »

The one and only thing I am sure of is that insurance pricing algorithms are black boxes which often provide totally nonsensical prices. I have had cases of huge variations in prices for minor changes, like one of the two identical cars costing more for no reason (the only difference was the color, blue vs red), or big differences in prices for minor variations in the job description (both of which describe my job perfectly well), or insurance being cheaper if a vehicle is declared as kept on the road than in a garage, etc.

Claims, I can understand. But quotes? An address with a history of getting many quotes means... what exactly? Could you please clarify? It seems to me you are confusing claims and quotes!

Colour might make a difference if there are statistics to show more claims are made per mile for a particular colour car. If red car drivers have more accidents than blue car drivers, why should they not pay higher premiums?

It's long been known that choosing the "right" job title from a list of genuinely applicable titles can lower your premiums. Again, a risk profile will have been attached to each.

I'm not confusing claims and quotes. An address with a history of getting many quotes suggests many things, some or none of which might be true. A driver trying to establish how much a claim or penalty has really added to their policy premium. Someone gaming the system by trying out different job titles (see earlier point). Someone trying to decide whether or not to declare a minor accident by seeing what the effect would be on their premiums, etc etc. All of which might be completely innocent behaviour (except for the last one) but any of which might suggest someone is a greater insurance risk.

Don't get me wrong, I'm not an advocate of insurers' methods but put yourself in the position where you are asked to insure someone's vehicle and their driving. Would you want to understand every possible thing you could about them, their motivation and their exposure to risk before you took a few hundred quid off them in exchange for liabilities that could run into seven figures?

SouthLondonUser

So what would your recommendation be if I want to check which car is cheaper to insure, if taking an advanced driving course or install a specific security device makes a difference, etc etc etc? Just use my real data, and suck it up if the insurers wrongly decide I am up to something dodgy purely because I am getting lots of quotes?

I can sort of understand that insurers might use it against me if I get lots of quotes, as they might think I am trying to game the system. But I do not understand what the relevance of an address with lots of quotes is. Again, if I share a flat with a dodgy driver, I can understand insurers classifying the address as ‘riskier’, but if I share a flat with someone who has simply been getting lots of genuine quotes?

Of course I do understand that insurers use any possible piece of information to detect and prevent fraud. My gripe, again, is that the pricing algorithms are black boxes, in which not even the insurers know what factors have what effect. Again, does any one have a logical explanation for the price difference I have witnessed? How can the same insurer quote £850 from one site, £970 from 2 others, and £1,700 from its own website, if every single detail was identical? To me, this just means their algorithms don't make sense; is there a better explanation? I can understand lots of things: I can understand insurers pricing on the basis of details which look totally irrelevant to me, but, again, when the details are all identical? What can possibly justify such a huge variation in prices?

Oh, before overinterpreting data, let’s not forget that the risk of spurious correlation is huge, when dealing with lots and lots of data. Some funny examples here: http://www.tylervigen.com/spurious-correlations

Aylesbury_Duck

SouthLondonUser wrote: »

How can the same insurer quote £850 from one site, £970 from 2 others, and £1,700 from its own website, if every single detail was identical? To me, this just means their algorithms don't make sense; is there a better explanation?

Yes. Customer behaviour. They reason that customers buying from one site will only buy at £850, that those on another site will pay a little more and for the few that go directly to them, they will get away with charging double. It's no different to auto-renewing which is nearly always more expensive than ringing and asking for a new quote from the same insurer. A percentage of customers have inertia and will pay the higher price.

I don't see your problem. You're a savvy consumer who's shopped around and if you go with Hastings, will presumably pay the lowest premium of the three you found. Why worry about others paying more? After all, they are effectively subsidising your lower premium.

I get my gas and electricity from nPower at the moment. It's the same gas and electricity coming through the same pipes and cables as it was from British Gas but it's a different price. I could even go on another nPower tariff and pay more for the same gas and electricity through the same pipes and cables with the same company on the end of the phone. But I don't.

EdGasketTheSecond

SouthLondonUser wrote: »

So what would your recommendation be if I want to check which car is cheaper to insure, if taking an advanced driving course or install a specific security device makes a difference, etc etc etc? Just use my real data, and suck it up if the insurers wrongly decide I am up to something dodgy purely because I am getting lots of quotes?

Yes you should use the proper data and get a proper quote; you'll find it will be much more accurate and consistent. I have used comparison sites and when I've then gone to the insurers own site, I get the exact same quote. Better to use your own data than someone else's and cause them a problem. I have not found any problems using my own data to obtain different quotes for prospective cars, job titles etc.

takman

EdGasketTheSecond wrote: »

Apparently dummy quotes can have consequences; see this thread:

https://forums.moneysavingexpert.com/discussion/5797314

Which is why in that thread I was arguing against insurers storing random data generated from internet quotes but seems I was shouted down so not everyone agrees.

The only "consequence" was that the OP had to allow them to access their licence to check that they didn't have the points they previously stated on a comparison site. Which i personally don't think is an unreasonable request in any way and really they should be doing this for everyone who takes out an insurance policy and not just a select few.

SouthLondonUser

Aylesbury_Duck wrote: »

If red car drivers have more accidents than blue car drivers, why should they not pay higher premiums?

You are assuming there is a statistically sound reason why insuring red cars is more expensive. Unless you are one of the actuaries who devised the pricing algorithm of an insurer, how can you be sure? Can you rule out that it's just a nonsensical result of the black box algorithm? Or that insurers are charging more simply because they can?

The fact that I got 3 wildly different prices from the same insurer using 3 different channels (2 price comparison sites vs the insurer's own site) proves that price differences are not always a true reflection of a different risk profile.. In this case it was self-evident; in the case of charging more to insure red vs blue, or to insure a vehicle with a sticker that must be declared as a modification, it's not. Maybe there is a statistically sound reason. Maybe there isn't. How can you be sure? Do you trust insurers? I most certainly don't!

SouthLondonUser

takman wrote: »

The only "consequence" was that the OP had to allow them to access their licence to check that they didn't have the points they previously stated on a comparison site. Which i personally don't think is an unreasonable request in any way and really they should be doing this for everyone who takes out an insurance policy and not just a select few.

I agree on this. My point is different: how can I be sure that simply getting quotes for different combinations (eg no modifications va some modifications) doesn't result in the insurers flagging me as a potentially dodgy driver who's trying to game the system? That's my concern. Note there is nothing dodgy about checking the impact of this kind of changes on the insurance. And let's not even discuss about the fact that insurers are legally allowed to charge more for the mere presence of a ******* sticker!

Car_54

SouthLondonUser wrote: »

You are assuming there is a statistically sound reason why insuring red cars is more expensive. Unless you are one of the actuaries who devised the pricing algorithm of an insurer, how can you be sure? Can you rule out that it's just a nonsensical result of the black box algorithm? Or that insurers are charging more simply because they can?

The fact that I got 3 wildly different prices from the same insurer using 3 different channels (2 price comparison sites vs the insurer's own site) proves that price differences are not always a true reflection of a different risk profile.. In this case it was self-evident; in the case of charging more to insure red vs blue, or to insure a vehicle with a sticker that must be declared as a modification, it's not. Maybe there is a statistically sound reason. Maybe there isn't. How can you be sure? Do you trust insurers? I most certainly don't!

If you try buying an identical item (tin of Heinz beans?) from several different retailers, you will see different prices. It's called competition. Why should insurance be different, if you're looking at the same product through different channels?

If they were all charging the same price, wouldn't it suggest a criminal conspiracy?