A picture prepared by J.P. Morgan and published in a recent SA article /
seekingalpha.com/article/319663-on-divid... attracted my attention. I reduced the picture to only US market to illustrate the pattern I observed.
Original of the picture: www.jpmorganfunds.com/cm/BlobServer/jp-l...;filename=jp-littlebook.pdf
The pattern is:
If dividend yield INCREASED in the decade K to compare with the previous decade K-1, then capital appreciation in the next decade K+1 will be HIGHER than in the decade K.
Or the same: If dividend yield DECREASED in the decade K to compare with the previous decade K-1, then capital appreciation in the next decade K+1 will be LOWER than in the decade K.
I used green and red lines to outline the pattern.
Of course data set is limited and I beleive prof. Siegel (?) or prof. Shiller or somebody also provided extended dataset that should be analyzed (sorry I don't remember exactly the book author and title but the author reconstructed data for US market from ~ 1870). Moreover might be the author of this book or somebody also already recognized the pattern, so I use "new (?)" with question mark in the title.
Of course I know that human beings tend to see pattern in random numbers but I think I can explain this pattern by reverse to average postulate for equity markets.
Anyway, if this pattern will continue I think we had slightly higher yield in recent previous decade to compare with 1.8% (S&P suppose to have excact number) and should expect at least positive capital appreciation in this decate. Let's see.
Added 3 March 2012:
The pattern was already described:
The case is closed - it correct.