Why is Insurance Legal?
Somebody sent me Geico’s list of unusual insurance policies the other day and asked me how I might price them. Leaving aside the throwaways like alien abduction and werewolves (although there is absolutely something to be written on alien abduction insurance), there’s some food for thought in there.
Think about ransom reimbursement insurance, for example. You’d want to know things like where that person travels and how likely kidnappings are of a person of that profile in that area. You’d want to try to estimate what the ransom demand might be. And you’d probably want some kind of dynamic pricing model or ability to reset rates given that the odds of those variables could change on a dime.
Of course, the price might also change if you were able to get the insured to agree to do things (like not travel to places where they were likely to be kidnapped) that reduced the odds of a claim. But while people typically don’t buy insurance so they can’t do things, there are likely some things they would never do anyway that they are willing to agree to:
Markel is not Geico, but it is a wildly successful specialty insurance company, and I had the opportunity to see its then CIO and now co-CEO Tom Gayner speak at an event some years ago. He ended up discussing the company’s successful business insuring competition fishing boats (i.e., small boats with big engines that are dangerous). The reason, he said, that they were so good in this space was because they had started in it early and kept very good records, so they now had a proprietary data set that let them price the risk of insuring these boats better than anyone else.
That’s a real edge (data!) and when you have one of those you need to push it as far as you can, as Markel has done in the marine insurance space.
What’s the point? Insurance is basically reverse gambling. In gambling, you pay small amounts of money to hopefully win big ones, but your expected value is negative because you probably won’t win anything. In insurance, you pay small amounts of money to hopefully avoid losing big ones, but your expected value is negative because you probably will pay more into the system than you will ever get out of it.
When it’s put that way, why did the government for so long prohibit gambling while requiring insurance? Part of the answer to that has to do with the difference between upside and downside, I get it, but both are scams in the sense that the house has a huge informational advantage over the customer in such a way that it makes the transaction inherently one-sided. If you’re buying insurance, in other words, you’re probably getting duped.
This is why I hate insurance and would rather self-insure against any event that is controllable and/or affordable. But insurance does have its place in protecting against uncontrollable events that might lead to bankruptcy (which is why I’m not lobbying too hard to get rid of it).
When it comes to insurance, or really anything, think about your tolerances and your data and your exposures and your costs. If you bet blind, you’re likely to lose, but the catch-22 is that even if you bet smart, you could still lose. So maybe just get the high deductible plan?
— By Tim Hanson
