2004: How should companies decide to pay dividends or buyback stock?

AUDIENCE MEMBER: Hi, my name’s Charlie Rice, and I’m a stockholder in from St. Louis, Missouri.

I’d appreciate hearing your comments on publicly-held companies using their cash for dividends versus stock buybacks?

WARREN BUFFETT: Well, we — the equation is pretty simple, but the practice doesn’t necessarily follow logic. The —

It’s obviously — as long as you’re telling the truth to your shareholders about what’s going on so that you aren’t manipulating the stock downward or something — when a stock can be bought well below its business value, that probably is the best use of cash.

It’s something The Washington Post did on a huge scale back in the 1970s. Teledyne may have bought 90 percent, or something, or close to it, of their stock back.

And that was the reason a very significant percentage of companies bought stock back in the past, because they actually thought it was selling for less than it was worth.

Like I say, that that can be abused if you do various things to bury your stock in one way or another, but that wasn’t the usual case.

Stock repurchases were relatively unpopular in those days. They’ve become quite popular now.

And to the extent that I’ve been around a good number of them, and been able to pick up on what I thought was the underlying rationale, if not the professed rationale, you know, I think it’s often done for people that are hoping that it causes their stock price not to go down, and their — and often done at prices that don’t really make a lot of sense for continuing shareholders.

If we wanted to return a bunch of cash to shareholders, we would — if our stock was undervalued — we would go to the shareholders, and say, “We think it’s cheap, and we think that this cash can be better used by you than by us.

“And we will, therefore, have — be repurchasing at what we think is a discount intrinsic value.” And the people that remain will be better off, and the people that get out will get out at a little bit better price than they would otherwise.

In terms of dividends, you get into an expectational situation. And for most companies that follow a — that pay a cash dividend — it doesn’t make sense to bounce around the dividend from year to year, although private companies frequently do that.

And we do it ourselves with our subsidiaries. They — some subsidiary can pay us a lot of money one year, and not so much money the next year.

But with public companies, people do — a lot of people do buy stocks to obtain dividends, and they hope for regularity, and that there’s a signally aspect to it and everything.

So I would say that once you establish a dividend policy with a public company, you should think a long time before you change that policy in a material way.

But I think the best use of cash, if you don’t have a good use for it in the business, if the stock is underpriced, is to repurchase it. And if it’s overpriced, you got no business buying in a single share. But a lot of companies do it.

Charlie?

CHARLIE MUNGER: Yeah, dividends are a very interesting subject. If you count the unnecessary stock trading, and the cost of investment advice, and the cost of making a lot of errors, and the trading costs in and out, I don’t think we’d be too extreme to say that now the total amount that’s paid out in dividends is roughly equal to the amount that is wasted in all this trading and investment advice.

So that the net dividends that come to the shareholders are approximately zero. This is a very peculiar way to run a republic. And very few people comment about it.

WARREN BUFFETT: Yeah, actually I did in an article, some time ago in Fortune. The frictional costs to American shareholders in sort of changing chairs for all American business as a whole, those frictional costs, are probably not much different than the entire amount paid out by American corporations.

So — but getting to the individual corporation level, a company that expects to regularly earn more than it can profitably employ in its business, should be paying out dividends.

Take a subsidiary of ours like See’s Candy. We would love to expand See’s Candy to double or triple its present size, but it doesn’t work. We’ve tried it a lot of different ways. So it should be paying out its earnings.

If it was a public company, and it was at one time, you know, you could argue that something approaching a 100 percent payout would make sense there.

But most managements worrying about earnings falling off at some time in the future would rather establish a lower level, and therefore, ensure regularity of dividends by going with a conservative level. I — you know, we —

It’s obviously something we think about at Berkshire when we have 30-odd billion dollars around. If we can’t figure out a way to employ that over time, you know, it’s a mistake to keep it in corporate form.

But we have this expectation, and I think it’s a reasonable expectation, that we get the — put it to work.

If we ever came to a different conclusion, if our stock — we thought our stock was significantly undervalued, we’d probably figure in terms of disbursing it through repurchases, particularly where now dividends and capitals gains are neutral for individuals.

And if our stock was not underpriced, and we fell, we would probably do something by a dividend.

It’s not going to happen soon, however. (Laughs)