Again, "bonds beat stocks" only works if you knew in advance to pick specific bonds (not a bond fund) and hold those bonds for 20 years.
I can also beat equity index fund returns (e.g. the S&P 500) if you let me go back ~20 years and pick individual stocks that turned out to be winners:
Happy 20th IPO anniversary! $1,000 in Amazon 20 years ago is now worth $638,000
+1
I went back and read that thread in a bit more detail. I'm actually surprised that DLDS offered that link up as any sort of support for bonds, because it works against her.
hah - I just deleted two paragraphs of response I was working on, because in writing and trying to support DLDS as far as I could (which still doesn't support what she is saying), I realized that even her narrow view is wrong anyhow.
First, as
ncbill and I have pointed out, a bond portfolio of 100% 20 year bonds, cherry-picked in the rear view mirror, just is not what anyone is going to choose for a 100% bond portfolio. So it is meaningless if that portfolio did better than stocks over some specific time period. If a portfolio never existed, did it make a sound in the forest?
But then I saw the next flaw in her statement. She's saying that since stocks did worse than this phony bond portfolio for even one time period, that stocks would be the riskier investment on the downside, so the 100% bond portfolio is safer, and could provide a higher 100% historically safe WR.
The flaw is, they showed the one or two periods that this phony portfolio did better than stocks. OK. But what was the
worst 30-35 year period for this phony bond portfolio? If that was worse than the worst 30-35 year period of a balanced portfolio, it all falls apart. Even for a phony, cherry-picked bond portfolio.
So DLDS, what was the return on a 100% 20 year bond portfolio in the worst 35 year period in history? That's the point we need to measure against. And realistically, it should be for a bond allocation that someone would actually implement, with a minimum of effort, and no crystal ball or rear view mirror.
A bridge too far.
-ERD50