2002: How does Buffett ensure he doesn't fool himself?

AUDIENCE MEMBER: OK. My name’s Paul Tomasik from Illinois.

I’d like to talk about your thinking, if you don’t mind.

In the “Fortune” magazine article that you sent to all the shareholders, you referenced a practice by Darwin that when he found something that was contrary to his established conclusions, he quickly wrote it down because the mind would’ve pushed it out.

And if you read “The Origin of Species,” Darwin’s very careful to avoid fooling himself. He very carefully asks and answers the hard questions.

It’s a feedback mechanism, and you’ve picked up on one of his feedback mechanisms to avoid fooling yourself.

So the two questions are this:

If you look — model — how you think, Charlie thinks, how physicists think, how mathematicians think, you see the same pattern.

You want to use logic. You’re dedicated to logic. But logic’s not enough. You have to avoid fooling yourself, so you build feedback mechanisms.

So the first question is, do you see it that way? That you’re thinking just like mathematicians, physicists, and some of the other exceptional businessmen, by being logical and being careful to have feedback mechanisms?

And the second question is about other feedback mechanisms.

Your partnership — sitting next to you is a great feedback mechanism. Hard to fool yourself when you partner with Charlie Munger.

WARREN BUFFETT: Right.

AUDIENCE MEMBER: This meeting’s a feedback —

WARREN BUFFETT: Hard to fool him, too. (Laughter)

AUDIENCE MEMBER: But that’s not an accident.

The meeting is one level of feedback mechanism, the way you attack the annual report letter is a feedback mechanism.

So you could comment, both of you, on other feedback mechanisms you developed? Thank you.

WARREN BUFFETT: Well, you’ve come up with two very good ones. I mean, there’s no question that Charlie will not accept anything I say because I say it, whereas a lot of other people will.

You know, I mean, it’s just the way the world works.

And it’s terrific to have a partner who will say, you know, you’re not thinking straight.

CHARLIE MUNGER: It doesn’t happen very often.

WARREN BUFFETT: There’s no question, the human mind — what the human being is best at doing is interpreting all new information so that their prior conclusions remain intact. I mean, that is a talent everyone seems to have mastered.

And how do we guard ourselves against it? Well, we don’t achieve it perfectly.

I mean, Charlie and I have made big mistakes because, in effect, we have been unwilling to look afresh at something.

You know, that happens.

But we do have — I think the annual report is a good feedback mechanism. I think that reporting on yourself, and giving the report honestly, whether you do it through an annual report or do it through some other mechanism, is very useful.

But there — I would say a partner, who is not subservient, and who himself is extremely logical, you know, is probably the best mechanism you can have.

I would say that on the contrary, to get back to looking things you have to be sure you don’t fall into, I would say the typical corporate organization is designed so that the CEO opinions and biases and previous beliefs are reinforced in every possible way.

I mean, having staff surround you that know what you want to do, you are not going to get a lot of — you’re not going to get a lot of contrary thinking.

I mean, most staffs, if they know you want to buy a company, you’re going to get a recommendation.

Whatever your hurdle rate, if it’s 15 percent internal rate of return, which very few deals ever work out at, you know, or 12 or — they’re going to come back, and they’re going to come back with whatever they feel that you want.

And if you arrange your organization so that you basically have a bunch of, you know, sycophants who are cloaked in other, you know, titles, you’re not going to get — you are going to leave your prior conclusions intact, and you’re going to get whatever you go in with your biases wanting.

And the board is not going to be much of a check on that. I’ve seen very, very few boards that can stand up to the CEO on something that’s important to the CEO and just say, you know, “You’re not going to get it.”

You’ve hit on a terribly important point. All of us in this room want to read new information and have it confirm our cherished beliefs. I mean, it is just built into the human system.

And that can be very expensive in the investment and business world.

And, like I say, I think we’ve got a pretty good system. And I think that most of the systems aren’t very good, that exist in corporate America, to avoid falling into the trap you’re talking about.

Charlie?

CHARLIE MUNGER: Yeah, I think it also helps to be willing to reverse course even when it’s quite painful.

As we sit here, I think Berkshire is the only big corporation in America that is running off a derivative book.

And we originally made the decision to allow the General Re derivative book to continue, and it’s a very unpleasant thing to do to reverse that decision, yet we’re perfectly willing to do it.

Nobody else is doing it, and yet it’s perfectly obvious, at least to me, that to say that derivative accounting in America is a sewer is an insult to sewage. (Laughter)

WARREN BUFFETT: I would second that. I might not have chosen those exact words, and we may not even use those words in describing why we got out of it, but —

Yeah. And in the first quarter of this year, we’ll show quite a bit of income — and anything we say here, we ought to put on the internet, Marc — but I think we’ll show, what, 100 and, I don’t know, 60 million or something like that of financial — maybe it’s 140, I’ll take a look here.

Well, you’ll have 160-odd million of income in that funny little line we have from financials income.

But that will be after an $88 million loss, in terms of getting — the first steps — of getting out of the General Re — what used to be called General Re Financial Products — derivative book.

You know, those losses were there. I mean, some of that is a shutdown loss, 30-odd, 30 million or thereabouts is severance pay and that sort of thing.

But the truth is that derivative accounting is absolutely terrible in this country, and there are a lot of companies that will not want to face up to what would be involved if they actually got out.

Now you’re seeing derivative accounting unwound at Enron, in a very major way. And believe me, it’s not being unwound at a profit, except to the extent that the bankruptcy court lets them disaffirm certain contracts.

I mean, it is — there was no place where there was as much potential for phonying numbers at a place like Enron than the derivative kind of data.

They were marking-to-model, they were doing all these things.

You give a whole bunch of traders the ability to create income by putting little numbers down on a piece of paper that nobody can really check, and it, you know, it can get out of control. It will get out of control.

And so we decided, finally, to bite the bullet on it, and we’d get out of it.

And it would — incidentally, we would not have reported $88 million of loss if we’d stayed in it. Might have reported a tiny profit or something, but, in the end, you know, the loss was there.

And there will be — it could well be — some more to come in that, because once you get into derivatives — I think our longest contract may run 40 years or something like that.

The guy who put the 40-year contract on the book probably got paid, you know, that week for putting it on, virtually. And, you know, we’ve got a bunch of assumptions as to how it’s all going to work out over 40 years. You couldn’t devise a worse system.

And, in the end, you know, we didn’t want to be in the business when we got in it, and we are now in the process of getting out.

But you don’t get out of fast — out fast — in something like this.

I mean, it’s, you know, it’s a little like hell. It’s easy to get into, and it’s hard — very hard — to get out of. (Laughs)