The funny thing with insurance is that if they ever become too good at estimating risks, they will destroy themselves. You don't need insurance if you can perfectly predict the future.
Though I don't think we particularly close to this point, so I doubt they care too much.
We're going down some kind of bizarre path. Insurance is ultimately about risk sharing. If every risk is modeled then there is no insurance, just expensive prepayment.
Sorry, what? Risk and uncertainty are universal. Everyone has to deal with their existence.
There have been plenty of insurance systems developed throughout time whereby people can hedge their risks. Previously however insurance was informal and consensual - you signed up to a Roman college. You and everyone else in the society engaged in it of your own free choice. If your group couldn't manage its finances it went broke (or more accurately appealed to an aristocrat.)
Nowadays we live in such enlightened times that the elements of choice and freedom have largely been removed.
Yet, the end game is that everyone can predict risk so thoroughly that insurance is pointless.
Not true. Suppose you have a 0.01% chance of needing a $10M treatment in your lifetime. First of all you can't say, "Oh I'll just self insure" because few people have $10M. Second, you may decide that paying $10,000 over the course of your lifetime is preferable to risking a payment of $10M.
Removing uncertainty doesn't eliminate the need for insurance, it just reduces the opportunity for risky subscribers to socialize their risk, and for insurance companies to reap gross profit.
In some way, that is honest, too. Very low-probability, but highly correlated risks (if it happens to you, it likely will happen to your neighbor, too) simply aren't a good fit for a capitalist system that does not look ahead more than say ten years.
If an insurance company were to insure against such risks it would happily collect money from its customers. After a couple of years, its shareholders would force it to make dividend payouts.
Then if the 'un'thinkable were to happen, it would go broke, and the customers would not get the money they paid for.
'The insurance industry transforms from “recovery after risk” to “prevention of risk:”'
I understand insurance to largely already work like this: high premiums discourage risky behavior. For example, one might be dissuaded from building a home on a flood plain when one finds out what the flood insurance will cost.
One could argue that the social value of insurance is to incentivize you to not take risks we know about, while protecting you from risks we don’t know about.
We all understand that insurance was invented only a few decades ago, right? Its not been some fundamental requirement for most of humanity (still isn't) for most of human history.
I get it; the family has to be protected. But if we'd never heard of insurance, we'd all be living our lives anyway, and adopt a more fatalistic approach. And preventable tragedies would occur more often.
Sometimes I think that insurance is a scam, a con game where we all surrender our money because of scare tactics and threats of unlikely events that probably never occur.
But my point is by hyper-optimising for individuals it will at some point cross a line where it's not really insurance as you understand it any more. The unlikely event is that the insurance company gets it wrong, for example your house floods in an area that was not predicted to flood for millennia. As the insurance company gets better, the chance of an unlikely event gets smaller. How small can that chance get before you decide to just take the risk yourself?
At the extreme, imagine a future world where long term health is so predictable that there's no uncertainty. You definitely can't offer an optional insurance policy here (at least, the insurer either isn't going to make any money, or the policy costs will make buying it pointless). In a similar way, you won't do very well selling retrospective fire insurance, as only those people with burnt-down houses will buy it.
Insurance needs 1) a large pool of coverage, where the combined average costs will cover the peak extremes. It also needs 2) variability: you have to have people with lower-than-average costs to balance out the higher-than-average costs. If everyone has average costs then once again, you've lost the uncertainty and any benefit of insurance.
To get 1), you could enforce insurance on people. To get 2), you have to ensure insurance products aren't ultra-specific. Neither of these are popular!
What is the societal value of adding more (and real-time) data to insurance?
In general, insurance helps spread risk. This means that if something drastically bad happens we distribute the impact over a larger population reducing the individual impact.
In the extreme if you perfectly assess risk insurance ceases to have any value. The insurance costs would perfectly reflect the risk and there would be no benefit to owning insurance.
My hunch is that this kind of data driven insurance could help people avoid risky situations. But I’m not sure that’s a given and it seems like there are many downsides (particularly if the risk is unavoidable).
Insurance with perfect info would be an amazing social good. If they could say "you can build a house there, but it will burn down in a forest fire 15 years from now", we could make an informed decision on if we want to build that house.
I don’t think you can predict risk well enough for that to come true. No matter the predictive power, real life still carries probabilistic nature. We can assign 0.0001 probability to somebody’s house burning down, but in real life it either will, or will not.
They can then choose to skip insurance, and carry a very small risk of loosing a lot of money. Or buy insurance, and carry a very high probability of wasting much smaller amount of money.
reply