Jeff Paul published good "Finding High Yield, Low-Payout Outperformers" article (seekingalpha.com/article/357811-finding-... analysis of 2009 report by Credit Suisse (NYSE:CS) - see link in his article. At page 25 of this excellent report esp. for dividend zealots (seekingalpha.com/instablog/725729-sds-se...) I found interesting table I reproduce here for convinience:
Jeff Paul used the most right column and corretly concluded that over an 18-year period, from 1990-2008 the "High-Yield, Low-Payout" (HY-LP) split shown in pink in the table above offered the best return.
I ploted HY-LP position for each year from the table (+5 than it is the best performer and -5 than it is the worst performer, +1 than it is just a level above and -1 han it is just a level below average) and got the following graph:
In my mind this is quite a rollercoaster because of often (quarterly) rebalance and I'd stay away from this. I think a dividend investor should not chase most promissing approach and should keep long-term perspective. I prefer to have a broad portfolio seekingalpha.com/instablog/725729-sds-se...) of different dividend stocks and ignore noise.