We ideally hold 25 to 35 positions, with a core holding at around 5 percent. That gives us most of the free lunch of diversification and allows us to maximize return by owning only our best ideas. There's also a human element to limiting our number of positions. With 25 to 35 stocks, our entire [...]

When we were buying Coca-Cola years ago, I'd lay out for other investors one part of our thesis that enormous demand in emerging markets could eventually turn Coke into a growth stock again but many of them just weren't interested. “Tell me again in a couple of years,” they said. If we're comfortable that value [...]

We're trying to own things that look like value stocks when we buy them, but which turn out to be growth stocks. So we try to build a portfolio of businesses with two primary characteristics. The first is that, even in times of stress, the underlying income-producing assets are strong enough to maintain a value [...]

We spend time speaking with competitors, the one group with a vested interest in telling us why our thesis is wrong. Before we bought [used-car retailer] CarMax in 2001, we went through Sunday newspapers from every market they're in and visited car dealers advertising against them in the paper. We'd describe how CarMax was going [...]

Making judgments about management is important to us and something I think value managers tend to underweight. You can analyze something statistically, but if you expect to own it for 10 years, management is going to make thousands of decisions you can't predict and may never even know about, which collectively make earnings compound at [...]

We avoid industries in which information arbitrage is extremely important to stock prices. I don’t think we’d buy a single-product biotechnology firm,for example. The same holds true in a crisis situation like Bear Stearns [in 2008]. I had no idea whether it was a zero or if it was going to be fine. In cases [...]

We’re unlikely to invest in a pure cyclical like a steel company, where the returns are governed primarily by macro forces.