2007: Is volatility a good measure of risk?

AUDIENCE MEMBER: Hi. I’m Bob Kline (PH) from Los Angeles.

Pursuing your earlier comments on sigmas from a different angle, the conventional wisdom in the investment world is that an investment risk can be measured by the volatility of the price of the investment in the marketplace.

To me, this approach has it backwards. Since changes in price are determined by the changes in the opinions of investors in the marketplace, why would a rational investor substitute the opinions of the marketplace, as reflected in the volatility of the price, for his own assessment of the risk of the investment?

And consultants take this idea further by tracking the volatility of a portfolio manager’s results in an attempt to measure risk. So could you guys expand on your thoughts on this?

WARREN BUFFETT: Yes. Volatility is not a measure of risk.

And the problem is that the people who have written and taught about volatility do not know how to measure — or, I mean, taught about risk — do not know how to measure risk.

And the nice thing about beta, which is a measure of volatility, is that it’s nice and mathematical and wrong in terms of measuring risk. It’s a measure of volatility, but past volatility does not determine the risk of investing.

I mean, actually, take it with farmland. Here in 1980, or in the early 1980s, farms that sold for $2,000 an acre went to $600 an acre. I bought one of them when the banking and farm crash took place.

And the beta of farms shot way up. And, according to standard economic theory or market theory, I was buying a much more risky asset at $600 an acre than the same farm was at 2,000 an acre.

Now, people, because farmland doesn’t trade often and prices don’t get recorded, you know, they would regard that as nonsense, that my purchase at $600 an acre of the same farm that sold for 2,000 an acre a few years ago was riskier.

But in stocks, because the prices jiggle around every minute, and because it lets the people who teach finance use the mathematics they’ve learned, they have — in effect, they would explain this a way a little more technically — but they have, in effect, translated volatility into all kinds of — past volatility — in terms of all kinds of measures of risk.

And it’s nonsense. Risk comes from the nature of certain kinds of businesses. It can be risky to be in some businesses just by the simple economics of the type of business you’re in, and it comes from not knowing what you’re doing.

And, you know, if you understand the economics of the business in which you are engaged, and you know the people with whom you’re doing business, and you know the price you pay is sensible, you don’t run any real risk.

And I don’t think Charlie and I — certainly Berkshire — I don’t think we’ve ever had a permanent loss in marketable securities that was, what, 1 percent, maybe, half a percent of net worth.

I made a terrible mistake in buying Dexter Shoe, which cost us significantly more than 1 percent of net worth where I bought an entire business then.

But I was wrong about the business. It had nothing to do with the volatility of shoe prices or leather or anything else. It just was wrong.

But in terms of marketable securities, I cannot recall a case where we’ve lost that kind of — I mean, we’ve done a lot of things in things — in securities — that had a very high beta. We’ve dealt with a lot of things in securities that had a low beta.

It’s just the whole development of volatility as a measure of risk, it has really occurred in my lifetime. And it’s been very useful for people who wanted a career in teaching, but it is not — we’ve never found a way for it to be useful to us.

Charlie?

CHARLIE MUNGER: Well, it’s been amazing that both corporate finance and investment management courses, as taught in the major universities — we would argue it’s at least 50 percent twaddle, and yet these people have very high IQs.

One of the reasons we’ve been able to do pretty well is that we early recognized that very smart people do very dumb things, and we tried to figure out why, and also wanted to know, who so we could avoid them. And — (Laughter)

WARREN BUFFETT: We will not run big risks at Berkshire. Now, we will be willing to lose, as I put in the annual report, $6 billion in a given catastrophe, but our catastrophe business, run over many years, is not risky.

You know, a roulette wheel will occasionally pay off at 35-to-1, and that sounds like you’re paying out an awful lot of money compared to the amount bet on one number, but I would love to own a lot of roulette wheels.