Photo credit: ashokboghani

The market in 1999-2000 got narrow. Few groups and few stocks were leading the rise. Performance-conscious investors, amateur and professional, servants of the “Church of What's Working Now,” sold their holdings in the slower growing companies to buy the shares of faster-growing companies, with little attention to valuation differences.

I remember flipping the chart of the S&P 1500 Supercomposite for P&C Insurers and laying it on top of an index of the dot-com stocks. They looked like twins separated at birth, except one was upside down.

When shares are sold, they don't just disappear. Someone buys them. In this case, P&C firms bought back their own stock, as did industry insiders, and value investors — what few remained. When managed well, P&C insurance is a nice, predictable business that throws of reliable profits, and is just complex enough to scare away a decent number of potential investors. The scare is partially due to the effect that it is not always well-managed, and not everyone can figure out who the good managers are.

So shares migrate. Those that fall in the midst of a rally, despite decent economics, get bought by long-term investors. The hot stocks get bought by shorter-term investors, who follow the momentum. This continues until the gravitational effects of relative valuations gets too great — the cash flows of the hot stocks do not justify the valuations.