2013: Would telematics help GEICO?

AUDIENCE MEMBER: Hi. Mike Sorenski (PH) from New York.

In regards to GEICO, Warren, last year you said the firm had no plans to adopt usage-based driving technology, similar to what competitor Progressive —

WARREN BUFFETT: Right.

AUDIENCE MEMBER: — called Snapshot.

Is that still the case, and if so, why wouldn’t that technology give GEICO better data to potentially give discounts to customers?

WARREN BUFFETT: Yeah. That still is the case, and Snapshot has attracted a fair amount of attention and there are other companies doing that.

It’s an arrangement, essentially, to tie — well, the term “Snapshot,” perhaps, says it — to get a picture of how people really do drive.

Insurance underwriting, you know, is an attempt to figure out the likely propensity, based on a number of variables, of a person having an accident.

Now, you know, in life insurance, it’s very obvious that somebody 100 is — if you don’t know anything else about them — is more likely to die in the next year than somebody that’s 20.

When you get into auto insurance, figuring out who’s likely to have an accident involves assessing a number of variables, and different companies go at it different ways.

Clearly, on statistics, if you’re a 16-year-old male, you’re more likely to have an accident than I am.

Now, that isn’t because I’m a better driver. It’s because the 16-year-old is probably driving about ten times as much, and he’s trying to impress the girl sitting next to him.

And that doesn’t work with me anymore, so I’ve given it up. (Laughter)

But the — we ask a number of questions, and our attempt, as much as possible, is to figure out the propensity of any given applicant, or the possibility, that they will have accidents.

And there are a number of variables that are quite useful in predicting. And Progressive is focusing on this Snapshot arrangement, and we’ll see how they do.

I would say that our ability to sell insurance at a price that’s considerably lower than most of our competitors, evidenced by the fact that when people call us, they shift to us, and, at the same time, earn a significant underwriting profit, indicates that our selection process is working quite well.

I mean, if your selection process is wrong, if you treat a 16-year-old male and give him the same rate that you’d give a 40-year-old that’s driving their car 3 or 4,000 miles a year, you know, you’re going to get terrible underwriting results.

So our systems, our underwriting criteria, have been developed, you know, over many decades. We have a huge number of policyholders, so that it becomes very credible, these different underwriting cells.

And everybody in the business is trying to figure out ways to predict with greater accuracy the possibilities that a given individual will have an accident.

And Progressive is focusing on this Snapshot approach, and we watch it with interest, but we’re quite happy with the present situation.

OK. Andrew Ross Sorkin?

Oh, Charlie, I’ve got to give you a chance to comment.

CHARLIE MUNGER: I have nothing to add. (Laughter)