Subject: Re: Just to be provocative
I'm just kidding about the K-wave, but Shiller's CAPE plot does look like a long, saw-tooth cycle.

I once segmented historical P/E data and the corresponding forward returns into long periods of rising P/E and falling P/E, and then plotted forward return versus P/E. The result was two, 2nd order, curves, one for periods of rising P/E that was curved downward, and one for periods of falling P/E that was curved upward. The two curves met at the extremes of P/E with the highest forward returns at the lowest P/E and the lowest forward returns at the highest P/E. The curves diverged at the median P/E, forming sort of a tilted football. I think this is why linear fits of return versus P/E are fat in the middle. But I digress.