Ken,
The point I have tried to make is that you can't simply look at the percentage yield of a stock and determine whether or not it is risky. Just because it's HY. You have to look at those other factors. I call that the quality of a dividend. Things like it's payout ratio, earnings, earnings growth rate, P/E, P/B, ROE, ROA, sector, moat, market cap, . . . . .
And you have to look at all of those quality conditions, regardless of whether the percentage yield of the stock is 10%, or 2%. That is, a 2% stock could be riskier then a 10% yield stock, if all, or most, of those other factors were present.
My ONLY point in writing any of this is whenever I see anybody catagorically say that HY is riskier then LY, or no yield. Say that high payout ratio is riskier, or no/negative earnings growth rates are riskier, or high P/E and P/B stocks are riskier, or small cap is riskier then large cap. Say that HY bonds are riskier then LY.
"(JP sets a threshold around 8% I believe)"
The 8% was a 'target' that he used, and was the sum of the current yield plus the dividend growth rate. He had a minimum yield of 2% or so, and the higher the current yield, the less dividend growth rate he needed. Likewise, the 8% included, as I recall, a percent or two of what I call slop.
For example, in my case, I designed my retirement portfolio to support a real, inflation adjusted 6% (of the initial portfolio value) rate of withdrawal. I felt that 4% was a slam dunk, using almost any yield based withdrawal strategy.
Applying JP's methodology, needing 6%, I would include 2% of slop, so would choose assets (individual stocks) where the sum of the current dividend plus the dividend growth rate was 8%.
That 8% could come from a stock that had a 4% yield plus 4% growth, 6% yield plus 2% growth, 8% yield plus 0 growth, 10% yield with a MINUS 2% growth rate! That is, 10% yield where you expected earnings, and dividends, to decline. All of these add up to the magic 8%, which is 33% more expected return then the 6% that I need.