chrisplotz

Thanks for your comment. The default scales with our property coverage (and you can top it up too) ending up at $50k total. What number would you be looking for here?

Each insurable loss like a car accident, house fire or heart disease, has two contributors: inherent risk, and behavior. (Inherent risk is risk you can do nothing about - age, gender, genes).

To the degree that AI can help reduce or measure behavioral risk, it's not only a good idea but only fair. I think most people would admit pricing insurance based on people's behavior is fair. They certainly do not like the idea of paying for others' right to drive fast and stop short. AI is a tool that is accelerating this trend.

All of the scary things mentioned in the article actually have nothing to do with AI, but discrimination based on inherent risks. Pricing insurance on ZIP certainly does not require AI. And while genes are a medical factor, getting old is the #1 inherent medical risk.

Is it fair to price on these things? For a house, ZIP makes sense - wildfire risk differs. But it is likewise undeniable that the old will cost more to care for than the relatively young. Debating AI won't help - the real debate is about whether we as a society are willing to protect people from the inherent risks of getting old, childbirth, cancer, living in a floodplain or wildfire zone, or even growing up in a poor area. We don't really have an honest debate about that.

Hi guys thanks for the comments, and yes, I see what you mean. Will be reviewing these in full.

Yes, correct, IBNR can come from cat, or non-cat.

That's true, but I've yet to see a primary insurer that doesn't use reinsurance.

The headaches as I understand it are mathematical really - without a limit to cap costs, outliers have the potential to completely mess up the pricing. Anything where one claim could destroy the predictability of results is hard. So only a few wildly inaccurately priced GRs can take down a whole book - hence capping. If you aren't likely to have outliers, you don't need limits, and some HNW insurers are removing limits entirely since they just price accordingly.

Yes, I hear you on fighting the price war, and agree navigating a balance of price and coverage will be important. In our view, we intend to keep prices low through the use of technology and our dividend of risk profit, but definitely charge a sustainable premium for the cover we provide. Policyholders will be able to see where the money goes so it'll be an ongoing conversation.

Would love to get in touch, will drop you a line.

Thanks for the feedback, happy to consider. Could you be more specific about what you would expect? That would help us make adjustments.

Yes, you're right - "abnormal" is a bit in the eye of the beholder. It would have been a relative assessment - a 3k deductible is much more than most people carry (even HNW, an area I've worked in before).

On guessing rebuild, in CA you shouldn't have to guess, as it's the insurer's legal requirement to recommend something accurate, and the DOI supports the consumer there. Generally those increases will do the job on like-for-like replacement - the big problem is if your Building limit has not been reviewed for years, as they get out of date.

Guaranteed is definitely expensive since it has cost insurers some big claim headaches, so not surprised it's even harder to find now as some insurers are hurting after all the fires.

Typically Guaranteed replacement is offered by HNW providers such as PURE, AIG, and Chubb. Travelers might do so too. It's hard to find though, so talking to an insurance broker that can access those specialist markets (i.e. they do not sell online) might be your best shot. Also, most insurers will offer some kind of Extended Replacement Cost which pretty much does the job too.

Interesting development though is in CA, insurers are legally required to recommend you a replacement cost coverage amount that is adequate for your place. That probably has a lot to do with the changes you've seen on "guaranteeing" something vs just recommending.

Awesome, I will - thanks!

Property insurance attritional losses are fairly predictable, and reinsurance is there to smooth out Catastrophe loss years; they'll be there to support us and bring our Loss Ratio back under control. So key is to charge enough to cover attritional (i.e. predictable) losses + reinsurance premiums.

But yes in those bad years where there's no UW profit, there's no dividend - everyone's contribution was needed.

Thanks! Oh IBNR... I'm surprised too, but glad it got a chance to shine!

And you are right - UW profit shouldn't be a big part, but it's a contentious issue right now with interest rates where they are. Since it shouldn't be a big part of our profit model, we're looking to put our money where our mouth is and return it.

I've seen "haven't read my policy" numbers in the 80s and 90s depending on the survey. And yes - we're property insurance people but know that health can be even crazier. I understand a lot of the times the price for the procedure isn't really decided until they find out who's paying for it.

Underwriting profit hasn't always been a big component of insurer profits - typically "float" or that interest on the premiums held has been the big driver. But in a low interest rate world, companies have had to keep more and more underwriting profit in order to exceed cost of capital. As they need more UW profit, so the fight over claims gets worse...

You are right though that money needs to be set aside for bad years - in insurance we call it "reserving", and actually it is already accounted for before the 5-15%. It is stashed away in the loss ratio as "incurred but not reported" or is paid in reinsurance premiums, which are a fixed cost. 5-15% is what is left over after all that (and admin expenses).

Geography is important, but your specific street address isn't required for this tool. And really, most insurance pricing is done at a higher level than street address (zip, or insurer-defined territories) so actually it is one of the last bits of info needed. For rough advice as this gives, we don't need it.

Thanks - and good luck with your renters insurance search. You're right we don't currently give advice for comparing through the tool - but I've written about it on our blog, and have a even more low tech comparison tool there (google sheet!). https://blog.goodcover.com/save-time-and-buy-home-condo-rent...

Hopefully that could help you out in the mean time.

Regarding returning unclaimed premium - good question. The way insurers make money is by 1) collecting premiums and holding on to them, generating some interest, and 2) keeping more premium than they need to pay out in losses and expenses, which is called "Underwriting Profit."

Typically Home/Renters insurance is written in such a way that a company tries to keep 5-15% of premiums as Underwriting Profit.

We think that the conflict over underwriting profit is at the heart of why the insurance experience is bad - there's not much incentive in improving a user experience that you don't want users to use... So we want to give that 5-15% back in the good years where we don't need it to pay claims. So, what you would see is a dividend at the end of the insurance year. It's not going to be much, but it is "putting our money where our mouth is" on our commitment to policyholders.

Turns out that's really hard to do legally and financially - mutual insurers would technically do this, but starting one is a hugely capital intensive process. We're on the path, and hope to be able to share more about the process soon!

Thanks!

You make a really good point on privacy - insurance is a trust business and managing this right is really important. We delete Dec pages w/in 30 days (we don't need to keep them past giving you the advice), so that's a start and something we don't mention on the site but should. We also never expose user data, or show you what you've uploaded.

Unfortunately though to get insurance advice, property details are pretty important, and if you are concerned with your Liability coverage net worth is pretty important for a service to know too. What we don't need are names - actually people could black out that personal info (even address) and it would be fine!

Chris Lotz, cofounder of Goodcover here. The team and I are around and happy to answer any questions.

Would love to hear any feedback too! Particularly around the user flow, and whether our instructions make sense.

Thanks!