Last weekend I ran the dividend contenders list through a
set of quantitative filters, and after it was all said and done 13
stocks piqued my interest.
Well I screwed up the formula in the cell that calculated
cushion ratio. I accidentally inverted the numerator and denominator in the
final ratio calculation, so I was dividing the projected dividend total by
future “projected cash”. So stocks that
had a ratio over 1.0 were actually projecting their dividends to exceed the
cash generated over the next 5 years. That’s the opposite of what I wanted to
There were several clues that something was wrong.
cash flow position hasn’t deteriorated that much in just a few weeks. But I decided I the error had to have been made before rather than this time around:
I think I screwed up the dividend cushion ratio calculation last month, because that’s the one metric Cummins missed on this time around, and I can’t imagine it changed that much.
In discussing MSM, I said the following:
Their normal quarterly dividend costs them
~$25M in cash. That quarter they paid out $208M.That was very nice of them. It
might be nice if they had that cash now, but with a dividend cushion ratio over
2 (even if the special dividend became permanent), I doubt they’ll miss it.
To maintain a cushion over 2 even after the special
dividend hypothetically became permanent would be insane.
What was I thinking?
IDK. The calculation was obviously screwed up though. There are at
least two lessons to be gleaned here.
easy to justify or explain any kind of anomalous data.
believe everything you read on the internet.
I’ve corrected the googlesheet, which is still published
The list of 13 stocks isn’t any different, but their
relative rankings to one another are shifted quite a bit in light of their
actual dividend cushions.
I think TXN, XLNX and PII deserve much closer looks than I
gave them the first time around.