[In picking stocks] You really have to know a lot about business. You have to know a lot about competitive advantage. You have to know a lot about the maintainability of competitive advantage. You have to have a mind that quantifies things in terms of value. And you have to be able to compare those values with other values available in the stock market.

Finding a single investment that will return 20% per year for 40 years tends to happen only in dreamland. In the real world, you uncover an opportunity, and then you compare other opportunities with that. And you only invest in the most attractive opportunities. That's your opportunity cost. That's what you learn in freshman economics. The game hasn't changed at all. That's why Modern Portfolio Theory is so asinine.

Without numerical fluency, in the part of life most of us inhibit, you are like a one-legged man in an ass-kicking contest.

Our biggest mistakes, were things we didn't do, companies we didn't buy.

Berkshire's past record has been almost ridiculous. If Berkshirehad used even half the leverage of, say, Rupert Murdoch, it would be five times its current size.

I have concluded that most PhD economists under appraise the power of the common-stock-based "wealth effect," under current extreme conditions... "Wealth effects" involve mathematical puzzles that are not nearly so well worked out as physics theories and never can be... What has happened in Japan over roughly the last ten years has shaken up academic economics, as it obviously should, creating strong worries about recession from "wealth effects" in reverse.

A different set of incentives from rising in an economic establishment where the rewards system, again, the reinforcement, comes from being a truffle hound. That's what Jacob Viner, the great economist called it: the truffle hound - an animal so bred and trained for one narrow purpose that he wasn't much good at anything else, and that is the reward system in a lot of academic departments.