Thursday, September 20, 2007

Conversation with William Bernstein

Thanks to Karl, a bunch of us got to meet William Bernstein today, and the conversation was far more interesting than I expected. Here's a quick summary:
  • Do you consider Berkshire Hathaway a separate asset class? I consider Berkshire Hathaway a closed end fund. Yes, the P/E of BRK is 10, but if you ask private businesses around the country, they'll tell you that they'll have a hard time selling their business for 5 times book value, let alone the 10 that Berkeshire Hathaway is getting. This means that there's a premium for Warren Buffett to be running the fund, and I do not expect him to be running the fund 40 years from now. But if you were to put together a fund that invested in private companies that can't be bought on the market, I would consider it a separate asset class.
  • Why do you suggest using fixed asset weightings for regions, but market cap weighting for stocks, sectors within a region/country? For countries, there have been countries where the market cap has gone to zero. For instance, the rate of return in Peru for the last century has been -99.5%. So for you to rebalance in those cases would be a bad idea. But within regions, the risk is low, and certain countries like the US, Japan, or Britain can be considered regions because their markets are so mature. We then debated between fundamental weights and market cap weights *with* value/small tilts. At the practical level, on the market cap side you have Vanguard, DFA. On the fundamental side you have folks like Rydex and PIMCO. If you were to ask me there's no contest. (Somewhere in there he also mentioned that he was willing to consider REITs and precious metal equity as separate asset classes you rebalanced against)
  • Isn't re-balancing just market timing? There was a paper written quite a while back about this precise issue. The authors were very coy about it. They postulate a world in which nearly everyone was a convex investor (i.e., when something went up they bought more of it, and when something went down they sold it or bought less of it). In that world, you'd make more money if you were a concave investor (i.e., buy more when it's down and sell when it's up). It turns out the model works both ways --- if most of the world is concave, you'd actually make money by being a momentum investor. But of course, the majority of the world is made up of convex investors, which is why rebalancing works. In fact, if you were to buy more of stocks when dividend yields were high, and less of them when dividend yields were low, you'd do very well. Just because you believe in the efficient market does not absolve you of the responsibility to do the math and look at what makes sense.
  • How do you get the data to do this computation? You can subscribe to Morningstar. Or use the Wall Street Journal. Or if you're a DFA advisor you get the data as part of the package.
  • Interest rates dropped 50 basis points. How does that affect you? It shouldn't affect you at all. It's a no-op.
A fascinating discussion, and I was glad to be in the room with such a smart guy and ask him questions.
Post a Comment