No. of Recommendations: 27
I was intrigued by Jim's work on CAPE's accuracy to predict future long-term S&P500 returns:
AdrianC writes on CAPE and BRK post #448 in thread
So I dug into the issue a bit myself using Shiller's data back to 1871 (Jim considered 1995-onwards). I basically reproduced Jim's result, then looked further back in time. Below is my 5 page report. You can get the gist by just looking at the pictures and reading text immediately above and below, or read on more for more detail
dropbox.com: PDF documentThe Mr. Finke that I refer to is a finance person who wrote a piece on CAPE, "The Remarkable Accuracy of CAPE as a Predictor of Returns”, claiming that it's remarkably accurate well before the 1995-onwards time-frame that e.g. Jim used. But I don't reproduce his result. Instead I find that 'Rsqr', the square of the correlation of CAPE and Shiller's future ten year returns, plotted over time using a rolling 15 year window of data, has great periods like 1995-onwards and also previous great periods, however it also has periods where Rsqr falls off a cliff. I tried to avoid making an error in the analysis, but it's possible.
