2004: How does Buffett choose metrics for incentives?

AUDIENCE MEMBER: Good afternoon, my name is Stan Leopard, and I’m from Menlo Park, California. I’m very pleased to be here, Warren and Charlie.

I first heard about you, Warren, in the late ’80s, and began reading your writings. Unfortunately, I didn’t invest until the late ’90s. You have shaped my business thinking, and as I listen to you, and as I continue to read what you write, and the things you recommend to read, it continues to shape my thinking.

My question’s about compensation. And I’ve seen your writing, and I heard the earlier comments today. And they still leave me, as a guy who is a business owner, not quite sure how to act to design compensation for managers.

For most of my career, I’ve been the senior manager in my businesses, but now I’m in a situation where I’m looking to own a majority interest of businesses that I don’t manage every day directly, and I’m very concerned with this compensation issue.

When I think about things, like, return on equity, or growth, or risk, or like that, but if you could speak a little more towards the specific of how you approach the getting it to the right things to measure and incent, I’d appreciate that.

WARREN BUFFETT: Yeah. It’s a very good question, and it’s — you know, there is no formula that applies across all industries or businesses.

You take something like return on equity. You know, if you pay way too much for the business that you buy, the person who runs it is going to get a lousy return on your equity. And they may get a good return on the tangible assets employed in the business, but your purchase price may defeat them, in terms of earning good returns.

If you base the — on earnings on tangible equity, you know, there are businesses like a network television station where, you know, if you have an idiot nephew, you can put him in charge, and they’ll earn huge returns on equity as long as they manage to stay away from the office.

And there are other businesses where you have to be a genius to earn 7 or 8 percent returns on equities. So there is no single yardstick.

To have a fair compensation system, both you and the manager have to really understand the economics of the business. In some businesses, the amount of capital employed is all-important. In some businesses, the amount of capital employed doesn’t mean anything.

So we have certain businesses where we have charges for capital and all of that, and where we have other businesses where that would just be an exercise to go through, and it wouldn’t really change any results, anyway.

We have a great preference for making them simple. I mean, we concentrate on the variables that count to us, and then we try to put that against the backdrop of the competitive nature, or the economic — the true economics — of the business they’re in, and really reward where they’re adding value, even if that value is from a very low base in a lousy business. And we make it — the base — very high if they’re in a very easy business.

And it hasn’t been a problem. But I would say it would’ve been an enormous problem if we’d brought in some compensation consultants, because they would have wanted something that would spread across the whole group, and it would have had all kinds of variables. And they particularly would’ve wanted something that would’ve to come in every year and redo in some way, so that they would have a continuing stream of income.

You know, if I knew what kind of a business you were looking at it, it’s easier to talk about what kind of a system to have.

If you had a group of television stations, just to pick an example — let’s say they were network television stations, all of a reasonable size. You know, you would probably figure that a chimpanzee could run the place, and have 35 percent pretax margins. And you might want to pay for performance above some number like that.

But there’s — it’s silly to have something that starts at 10 percent or 15 percent, when you do that. And a lousy manager will always suggest an arrangement like that.

Charlie and I have seen all kinds of compensation arrangements where, basically, you get paid for showing up. But they try to make it look, by constructing some mathematics around it, like you really had to achieve something.

But in the end, if you get a great manager, you want to pay him very well.

You know, we’ve got great managers, for example, at a place like MidAmerican. And somebody mentioned that there’s a big carrot out there for them if they achieve the results that we’ve set out. And that’ll be a check I’ll be very happy to write.

CHARLIE MUNGER: Yeah, if you want to read one book that will demonstrate really shrewd compensation systems in a whole chain of small businesses, read the autobiography of Les Schwab, who had a bunch of tire shops — has a bunch of tire shops — all over the Northwest. And he made a huge fortune in one of the world’s really difficult businesses by having shrewd systems. And he can tell you a lot better than we can.

WARREN BUFFETT: Yeah, and he worked that out himself. I mean, it’s an interesting book, and, you know, selling tires, how do you make any money doing that? And —

CHARLIE MUNGER: Hundreds of millions selling tires.

WARREN BUFFETT: Yeah, yeah. It’s a — and people like Sam Walton. I mean, the compensation system, I will guarantee you, at Walmart, or Charlie’s involved in Costco, they’re going to be rational because you had very rational people running them. And they wanted to get the best — they wanted to attract good managers, and they wanted to get the best out of them. And they had no use in paying for mediocrity.

But that does require a knowledge of the business. I mean, you don’t want to let — if you don’t understand a business, you know, you’re going to have a problem with both the manager and the consultant in terms of getting film-flamed on how you pay people.