2014: How will technology impact GEICO's moat?

ANDREW ROSS SORKIN: — And it has to do with the future as well, and it relates to GEICO.

And the questioner asked, “Could you, and perhaps Tony Nicely, please explain how you think about usage-based pricing, tracking drivers electronically and charging premiums accordingly, and how that will affect the auto insurance industry in the U.S. in the next decade, and how these changes impact the moat at GEICO?

I’m also sure you’ve studied the potential impact of self-driving cars on GEICO. Google says it’s now five years away. What does this mean to the future of the profit machine that is GEICO, and if the analysis showed a challenging future, would you ever sell?”

WARREN BUFFETT: Yeah. Well, the answer to the very last question is no.

The usage-based pricing is something that’s popular with some companies that have done a lot of work on it. Probably the one most identified with it is Progressive doing something called “Snapshot.”

And there’s no question that knowing how customers drive, or policy holders drive, and how they use their cars is a valuable input to assessing the proper premium to people.

And insurance is all about comparing the propensity of loss to achieve — or evaluating the propensity of loss to establish the proper premium.

And it’s very easy to understand in life insurance. I mean, if somebody is 90, they’re more likely to die than 20, despite Charlie — situation. (Laughter)

Even at 83 they’re more likely to die than somebody at 20.

The — so you know, that’s obvious. Females live longer. That’s not quite as obvious but it’s been established.

So there are various variables in insurance, and you try to assess those variables and set the proper price for the policy holder. If you lived in a state, for example, where the population of the state was one instead of, you know, 100 million, there’d be a whole lot less chance of an accident, you know, than — because of the lack of density of driving and so on. There’s all kinds of variables.

And through studying usage, by various methods, Progressive being probably the best known on it, they’re attempting to look at some variables and hope they get better information about the propensity of that particular driver to — or the likelihood of that particular driver — to be in an accident.

And we look at lots of variables, they look at lots of variables. We think we’ve got a pretty good system, and so far I think that’s been proven correct. But we’ll continue to look at many variables.

I feel very, very, very good about GEICO, GEICO’s management, and its ability to evaluate risk. And I think there are plenty of other people that are good at it, but I don’t think there’s — in my view, there’s nobody better at it, in terms of auto insurance, than the GEICO people.

So — but we ought to keep asking ourselves, “Can we do it better?” And we do ask ourselves that.

Now, when you get to the self-driving car, that is a real threat to the auto insurance industry. I mean, if that proves successful and reduces accidents dramatically, it will be very good for society and it will be very bad for auto insurers.

So you know, that can happen. I don’t know how to evaluate over how long a period that might take or what percentage of cars might be affected with that. But it certainly could happen and it would not cause us to be thinking for one second about selling GEICO. Charlie?

CHARLIE MUNGER: Yeah, some of these things happen a lot more slowly than you might think. I went to a program at Harvard, oh, at least 30 years ago, describing how color movies were going come to the house on demand, and they were just around the corner.

Well, they’ve come, but it was 30 years later. I have a feeling that self-driving cars having a huge impact on the market may take quite a while. And so I’m not —

WARREN BUFFETT: That would be my guess, but we could be wrong, you know.

CHARLIE MUNGER: Yeah, it could be wrong.

WARREN BUFFETT: But — (Laughter)

CHARLIE MUNGER: But —

WARREN BUFFETT: But if we are wrong, we’ll be wrong together. (Laughs)

It is hard to figure out how it could have a major impact in 10 years, but it may not work at all, who knows?

But GEICO will be doing more business, a lot more business, in my view, five years from now than now, and ten years from now.

You know, 30 years from now, you’re young enough to find out, and I will go away peacefully without knowing.

OK. (Laughter)