[dataflow]

> The international code of insurances says goods cannot be insured for more than their worth. The intent was to avoid perverse incentives

Would you mind explaining what the perverse incentive is here? If I want to insure a pillow that I claim is worth $1 million, why should it matter what others are willing to pay for it?

[praptak]

If they let me insure my stuff for 100x of what it's worth, I lose all the incentive to prevent damage.

Even in the legit cases the insurance companies have to account for the "don't worry, it's insured" mindset. Keeping the ceiling on the insurance value is intended to leave at least some of the incentive to prevent the damage with the owner.

The insurance companies cannot rely solely on the "don't be careless" contract clause.

[dataflow]

> If they let me insure my stuff for 100x of what it's worth, I lose all the incentive to prevent damage.

So what, though? Can't they just adjust the premium to account for that? It's not like they can't do their own modeling of what the item is likely worth -- if they see it's 1% of what you stated, then they can just as well cite you a ridiculous premium so that you wouldn't feel it's worth it. What's wrong with that?

[praptak]

In theory nothing, in practice it's just not worth it. Mind that the bad effects would also spread broader than a voluntary contract between two parties.

We'd have to fund the courts to resolve the inevitable insurance fraud accusations, not to mention the additional firefighting crews to put out the additional fires that consume the $1 pillows.

[rocqua]

The difference between gambling and insurance, is whether you have an insurable interest.

It makes the market for insurance much better if everyone actually has insurance. Because it reduces cost. It also keeps the industry legitimate, preventing gambling legislation from applying, and anti-gambling activists from targeting insurers.

You'll have to go to a bookie if you want to gamble.

[dataflow]

I don't follow the logic? How does above-market-value insurance discourage people from having insurance?

I don't get the comparison to gambling either, that reads more like an appeal to emotion than actual reasoning.

[schoen]

You can read about it at

https://en.wikipedia.org/wiki/Insurable_interest

(I don't know if that will make you more sympathetic to the legal rule or not.)

[dataflow]

You're definitely right -- it's interesting, but it's not making me any more sympathetic, because I fail to see why the lack of insurable interest is something the premium can't account for, and they fail to provide any explanation of that.

As far as insurance gambling goes, it feels fundamentally different? In gambling, the "house" that sells you the ticket sets the rules and introduces the element of chance. In insurance, the entity selling the financial product here is in no way in control of the outcome, which is the exact opposite of gambling.

[rocqua]

Because premiums will rise across the board, so people with an insurable interest pay premiums set for people who intend to gamble or manipulate their insurance.

By demanding an insurable interest, insurance companies keep out gamblers and frauds. It also helps strengthen the idea that insurance shouldn't be abused or manipulated for a payout.

[dataflow]

> Because premiums will rise across the board

I don't see why this is true. The insurer still knows the item and its market value. So if the insured amount is higher than the market value then it only needs to increase the premium in those cases, not for everyone else.

[rocqua]

As for gambling. The point isn't gambling is evil, but that others think gambling is evil, so being associated with gambling is bad for business.

[dataflow]

If it's bad for business then don't do it? That doesn't justify an international code.

[js8]

The incentive would be for you to have a "happy pillow accident" in which you get $1M. Of course, you might think that's good for you but the rules have to apply for everybody, by definition.

[dataflow]

> The incentive would be for you to have a "happy pillow accident" in which you get $1M. Of course, you might think that's good for you but the rules have to apply for everybody, by definition.

This doesn't pass the smell test, though. The premium would take care of that. You've told them you have a pillow, and that you want it insured for $1M. They could easily look at it and go "hm, this is worth $10", and give you a absurd premium of $999,900 in exchange for your absurd valuation. So happy accidents won't be worth it anymore. What's wrong with just letting the premium take care of it?

[js8]

You have simply rephrased the actuarial rule "don't insure item for more than its actual value". The "premium" you describe just inflated the value of the item.

[dataflow]

I don't see how this answers my question.

[smallnamespace]

> What's wrong with just letting the premium take care of it?

Offering a deal that nobody honest would take is a waste of time for everyone involved.

[hansvm]

Walking back from the pillow analogy a bit, I'd happily pay for homeowner's insurance that also covered lost wages, a temporary rental place, legal fees, and the other incidentals likely to arise in a fire or flood (as opposed to paying whatever high deductible I'm comfortable with on top of those other large, unknown costs). Adding those to the policy would necessarily go beyond the home value. Is that level of excess allowed?

[dataflow]

> Offering a deal that nobody honest would take is a waste of time for everyone involved.

I'm not suggesting any insurer should be forced to offer a deal. They're welcome to just shrug and tell you to pound sand. What I don't see is the logic behind having an international code prohibiting the offering of such deals. Is the international code trying to dictate to the insurance company what is worth their time?

[smallnamespace]

The international code is also defining the key distinguishing factor of insurance: it makes the insured whole against a risk that they actually have.

There are ways to bet on things where you don’t have that underlying risk: gambling, derivatives markets, prediction markets, etc.

These aren’t insurance and aren’t regulated as such.

[rocqua]

The premium would be 1M. Maybe .99M if they have reason to assume not everyone will be fraudulent.

[dataflow]

Sure, whatever. The exact value of the premium has no bearing on the point I'm trying to make.

[smallnamespace]

> why should it matter what others are willing to pay for it?

Because the actual value of the item determines your incentive to commit fraud.

If you insure a $10 pillow for $10, when you damage your pillow, you personally will definitely be out $10's value in goods in the hope you'll recover that $10 later. Since your only outcome is mildly negative, you don't have any incentive to file a false claim.

If you insure your $10 pillow for $1 million, as soon as the insurance is in hand, will have a strong incentive to destroy the pillow and try to collect a million dollars, since $1 million - $10 = $999,990.

This incentive exists regardless of what premium you had paid for the insurance (since it was a prior cost), and can't really be perfectly mitigated. Yes, you can criminalize fraud, ask for evidence, etc. but courts aren't perfect and it's always possible to be clever and fool people.

Also, some people are honest, and others are dishonest. An insurance company can't perfectly tell ahead of time who is who. Let's say I quote you $500k premium to insure your pillow for $1mm. A fraudster will see this as an opportunity to profit by $500k - $10. An honest person would see this as a terrible deal. Therefore only fraudsters would take this deal. If you continue to work backwards, as an insurance company you know there's no premium that you could quote that would end up in honest people taking this deal—there's no stable equilibrium where the premium charged ends up outweighing the (potentially fraudulent) claims.

Btw, this situation is famously described in George Akerlof's paper The Market for Lemons (he called it "market collapse"):

https://en.wikipedia.org/wiki/The_Market_for_Lemons#Conditio...

Another way to see this: rationally as an insurance company, if you ask me for a policy for $1mm on a pillow, due to the risk of fraud I will likely be quoting you close to $1mm as the premium. You (as an honest person) rationally would never take this policy. Therefore, I shouldn't even bother offering it, to save everyone involved time and energy.

[zabzonk]

depends on the premium, obviously

[dataflow]

What depends on the premium? In my mind, you state the item and the value, they tell you the premium they would cover it at. Where's the perverse incentive, and why is it relevant what anybody else would pay for it?

[listenallyall]

If you intend to insure a pillow for $1 million, expect the premium to cost about $999,950.

[dataflow]

I wrote as much in https://news.ycombinator.com/item?id=41755211

[listenallyall]

Then why did you object to zabzonk's comment?

[dataflow]

Because I don't see what the perverse incentive is?