Subject: Re: OT: CAPE's Predictive Power
BTW, my phrase "1995 to now is 18 years" is not an arithmetic slip.
I was referring to the fact that it's 18 years to 2013, and 2013 is the last date for which you have data for 10 year forward returns, given that my analysis was done at end of 2023.

Jim's point on averaging is a good one in general, but the scatterplot I showed for the 1995 onwards period is so good w/o averaging that it'd be impossible to improve on it much. The most negative possible correlation is -1 and the scatterplot has correlation -0.926
The full period from 1871 was done both with, and without, averaging of endpoints.

Valuation multiples do change, I go into that in part II. Also trends.

CAEY: Jim, are you defining it as 1/CAPE, or something else?