The One Number We Need From Insurers

When you get a loan, they always give you two numbers: the (nominal) interest rate and the APR. The first one takes into account only the interest you pay the bank, while the second is supposed to include fees and other items that get tacked on. Since banks get creative with their fees, you should always consider the APR more important than the nominal interest rate: if a loan costs you 1% up-front, that’s a lot of money to add to a potentially enticing interest rate!

When you get an insurance, you don’t really get anything. There is a ton of paper that explains what exactly you are insuring, and to what extent, and an incredibly long list of things that the insurance company decides it will not consider reasons for a payout. Some of them seem reasonable (for instance, if you have a term life insurance, there is always a suicide clause). Others are a bit iffy (no examples here, just look at your phone insurance for examples).

What is really strange is that banks are mandated to include fees in the APR, while insurance companies are not mandated to really do anything. The government prefers to deal with insurance, apparently, on a case-by-case basis. Certain practices are banned, while others are allowed. Some practices are allowed sometimes, others are forbidden on certain days of the month (just an example).

That makes it almost seem like there is no easy number that insurance companies could give us to give an indication of whether it’s worth it or not to get the insurance. Or is there?

Actually, there is. And it’s an astonishingly simple number: it’s the percentage of premiums returned (paid back) to the insured, the PPR.

In short, that’s the percentage of the money you give the insurance company that the company is actually going to pay back to you in average. How does that compute? Well, simple: you take all the money paid out by the company to insurance casualties, divide by the total amount of money paid by the insured, and there you go.

How does that help you? First, you should know that a lot of forms of insurance are what you and I would call a scam. It’s people that get money from you well knowing you are never going to get that money back, or not even a decent fraction of it. They are really terrible deals in general, and you should know that.

Second, you should know what parts of your insurance are worth it, and what parts aren’t. For instance, you may get really good collision coverage for your car, but is it really worth adding underinsured driver protection? How about that homeowners’ insurance: should you add flood or earthquake coverage?

Third,** you should know how a change in your insurance reflects on your premium.** I was horrified when I moved seven miles and my motorcycle insurance doubled. It would have been really good to know what percentage of that increase was actually due to the crazy motorcycle driving habits in Pacific Beach, instead of La Jolla.

Fourth, and finally, the PPR would allow you to shop around. What good is an insurance that costs less, but pays out a lower percentage of premiums? If insurance A quotes you $100 a month and pays back 20%, while insurance B quoted $90 but pays back 10%, A is the better deal (by a whisker: you give A $100 and it gives you $20 back in average = $80. B costs you $90 and gives you $9 back = $81).

But the downside! Let’s look at possible downsides.

First and foremost, some of the money you pay an insurance company goes to fraud prevention. That money is really both yours and the company’s, so it kinda should be counted partly in your favor.

Imagine for example that 10% of insurance claims are fraudulent. Since you are honest, you will never see those 10%, so it doesn’t do any good to add them to the payout. While the company could claim a PPR of 35%, you would never see 10% of that, because it goes to fraudsters.

At the same time, fraud prevention is largely the company’s decision. I am sure there are people that abuse the system, but how do we know the company doesn’t abuse fraud detection? If the PPR included the cost of fraud prevention, then the insurance company would have an incentive to deny claims.

In the end, the rational thing to do is to exclude fraud prevention costs from the PPR, but to also exclude insurance claims deemed fraudulent. That way, the insurance company can use as much fraud prevention as it sees fit, but it can’t claim the fraudulent claims as part of the payout.

With the fraud piece gone, we should look at how an insurance company spends its money. There is administrative cost, which is considerable. There is packaging cost, which is the cost of figuring out how much money you should pay. There is marketing cost, which is the money spent convincing you of your insurance needs (and your need to switch insurance). There is profit and dividend for shareholders. And payouts.

Which one of those costs is relevant to you? None of them, really, except the administrative cost, tangentially. You see, one of the things that really is important is how quickly you get your money paid out. You usually insure things you use (like a car or your home), so that not having money to pay for their replacement means you have to pay out of your own pocket.

As a consumer, you have a vested interest in getting your money back (it is your money, since the premiums are yours) in a reasonable amount of time. It doesn’t do you any good if your car is broken, if your insurance company pays you 2 years from now. Also, it doesn’t do you any good if you have to take time off from work to deal with the insurance company.

Let’s just say that insurance payments have to be quick. Let’s say they have to be within 30 days, or they get depreciated. Let’s say that every month you have to wait for your money, the insurance company has to “pay” a pro-forma (accounting only) penalty equal to your worst credit card rate. That makes computation a lot harder, but that’s what we have computers for. The alternative would be another, second number, that tells you what was the average wait time per dollar (to avoid the company paying out small sums faster than large sums).

You would probably be disgusted by the PPR, because it’s so small. Interestingly,** there are lots of web site that tell you exactly the opposite**. They usually don’t back up their claims with numbers, and tell you that insurance companies pay out most of the money or even take a loss, because they invest your premium while you haven’t gotten a payout.

That definitely happens, and it’s probably a leading source of revenue. At the same time, there are forms of insurance that are so ludicrously overpriced (rental car insurance, for instance) that the claim they are paying out most of the premiums is absurd. Let’s be serious: if it costs me more to insure against the loss of a car than to rent the car in the first place, there is something wrong with the business model.

At the same time, insurance companies should be able to make a profit. They spend lots of money streamlining products, figuring out how much money you are likely going to cost, and giving you an accurate idea of your risk. You shouldn’t really think that a PPR of 100% is a reasonable goal.

But, really, how many times have you watched a commercial for insurance switching and thought, “Hey, I am paying for this commercial! Turn it off and give me the money!”