Wednesday, August 27, 2008

Trading Out

When asked about timing investment exits, Warren Buffett has said that at Berkshire the preferred investment horizon is "forever." Yet, even long-term investors occasionally sell.

Warren Buffett last year sold Berkshire's stake in PetroChina (ADR's Ticker:PTR), which hasn't done so well as crowed by others in the time since. Sure, Buffett could have timed his top better -- but Buffett's methods don't involve identifying fad peaks or the like, but analyzing value. Berkshire Hathaway (BRK.B -- I'd quote the A shares, but honestly, how many among us will be buying those at current prices?) doesn't promise to outrun bull markets, but to offer safety in times of risk. And why not? It's an insurance company. The fact that it's consistently delivered handsome returns while offering investors a rock-solid and diverse financial foundation says something about the quality of the management that's been at its helm for decades. (Unless you think that every year for these past several decades have been a fluke ...?)

Bailing out of PetroChina at nosebleed prices seems good in principle -- but knowing when a stock is crazily valued is a bit trickier. It requires homework. It requires advance-planned decisions regarding the price at which one is better off investing in other opportunities rather than retaining the investment gain.

One lesson The Motley Fool offered years ago was that paying for quality companies was worthwhile. What I was unable to discern from The Motley Fool is how you tell your slick stock is ripe for market. After accepting that paying for quality can be necessary ... how do you work out when a pick is past its prime?

A look at Warren Buffett's behavior offers an answer: don't buy slick high-fliers unless and until they become bargains. Then, you hold forever. Unless they start looking like they offer downside risk that scares you. Then shift out, into another bargain.

The key is finding a pipeline of bargains. I love ACAS at these prices, but I need to work on my pipeline.

No comments: