2006: Have insurance rates hardened as much as Buffett anticipated?

AUDIENCE MEMBER: My name is John Gosh (PH) from Key West.

Have insurance rates hardened as much as you anticipated, and have you seen a significant flight to quality in the last few years?

WARREN BUFFETT: Yeah. I think you’re probably asking more about reinsurance rates.

Actually, in auto insurance, you can figure it out. Our policies are up more than our premium volume. So the average premium in auto insurance, which, after all, is close to 40 percent of the whole market for insurance — the average premium in auto insurance is actually down a little bit.

But in reinsurance, in which we are a big player, you will — there’s great variances.

If you take insurance for marine risks in the Gulf Coast — drilling rigs and offshore platforms and that sort of thing — those prices are up very dramatically, but they should be.

I think in the last couple of years, there’s been, like, 2 1/2 billion of premium in the Gulf Coast and 15 billion in losses. So if you paid out 15 billion and took in 2 1/2 billion, the more astute of you would figure that you needed a little more money for that particular risk.

We have been, historically — at least in recent years — the largest writer of cat — catastrophe — mega catastrophe insurance in world, and I think we will be this year. In fact, I’m almost sure we will be this year.

Our mix has changed some. Prices are up a lot, but what we don’t know is whether exposures are up even more.

We don’t know whether the experience of the last two years, we’ll say, with hurricanes in this hemisphere, is more to be relied upon than the experience of the last hundred years.

You can take the hundred-year experience and it tells you one thing, and you can take the last couple years and it tells you something else. And which is more meaningful? We don’t know the answer to that.

We do know that it would be kind of silly to assume that the 100-year experience is the relevant criteria when conditions — we know certain atmospheric conditions have changed. We know water temperature’s changed.

But we do not know all of the variables that are into the propensity of hurricanes to occur and the degree to how intense they may be if they do occur. We don’t know the answer to that. We don’t think anybody else knows the answer to it, either.

So we are getting more money for hurricane insurance. We’re getting appreciably more money.

If the last two years are the relevant years, we’re not getting enough. If the last hundred years are the relevant years, we’re getting plenty. And we will know more as time unfolds.

The really scary possibility is that variables are changing in some way so that the change is continuous and that what we’ve seen the last two years is not a worst-case example at all.

And, of course, you get into chaos-type theory with some of these variables where the outcome is not a linear relationship to the input, and you can dream up some pretty scary scenarios on this.

I don’t know whether they’re true and nobody knows.

We are willing to write certain areas, certain coverages, because we believe the prices are adequate, and we can sustain the losses.

We’re willing to lose many billions of dollars in a given catastrophe if we think we’ve been paid appropriately for it.

But it is not like figuring out the odds on flipping coins or rolling dice or something like that. You are dealing with changing variables, and you — the worst thing you could have would be a 100-year history book in making those judgments.

The third quarter, we will have a lot of exposure for wind. We don’t have as much exposure now — well, we may. I’d say we’re getting there. But we don’t have as much — certainly as much as we had a couple of years ago.

Prices — question about prices hardening. Prices are getting — are hardening — in that particular area. And if they get to what the — where we really feel they’re appropriate — you know, we might take on a fair — we will take on — a fair amount more risk.

If they don’t get there, even though they’re higher than last year, we won’t write — you know, we’re not interested in writing it, because it’s a dangerous business.

And we don’t believe in modelers at all. I read all this stuff about modeling. I wrote about that a few years ago. It’s silly. You know, the modelers don’t know a thing, in my view, about what’s going to happen.

And we get paid for making guesses on it. If, over a lifetime, the guesses are decent, we will know that, you know, we were doing the right thing.

But if this year goes by and nothing happens, we still don’t know whether we were right on the prices.

Because if you get a 25 percent rate for something and it doesn’t happen in a year, that does not mean that you didn’t need 40 percent or 50 percent. It just means that if you do it enough times, you will find out whether, overall, your judgments are any good.

It’s still a business we like. We bring a lot to the party. Everybody knows we can pay. You get into the question of creditworthiness.

If there is some super, super catastrophe — and I regard, sort of, the outer limits of that being a $250 billion insured loss — for reference, Katrina was a — presently estimated — was about a $60 billion loss.

So, if something comes along that’s four times Katrina, which could happen, you know, we can pay, and we can comfortably pay. We would probably have about 4 percent of that, maybe 10 billion.

A very large percent of the industry would be in very, very serious trouble.

So, we can play bigger than others, and we can survive better than others if something bad comes along. And we will see, over a five- or ten-year period, how we do. You can’t judge it by any one year.

Charlie?

CHARLIE MUNGER: The record of the past, if you average it out, has been quite respectable.

And why shouldn’t we use our capital strength to get into volatile stuff that makes other people frightened?