> Because only people who really, really, really want to make bad
> mistakes do things in the un-debuggable Excel (or Numbers)…
and then proceeds with an extended analysis in R. He may have used
Excel in the earlier posts; the comment is certainly a reference to an
infamous series of errors in a spreadsheet behind a much publicized
claim that it was disastrous for countries to have debt > 90% GDP. See
http://en.wikipedia.org/wiki/Growth_in_a_Time_of_Debt for details.
The background question to the blog post is whether current stock market
valuations indicate stocks are likely to be a poor medium-term
investment (10 years).
A quick look does not reveal to me what, if any, substantive changes to
the conclusions resulted from using R not Excel (or even if is previous
conclusions were based on Excel). In fact, it's kind of hard to tell
what the conclusion is!
A couple of more technical comments:
DeLong (the post's author) uses simple regression (lm) and then observes
> The significance levels that R reports are wrong: its naive regression
> package assumes that each of the 1482 observed 10-year returns is
> independent of each of the others. They are not.
I assume R has some tools that can do proper time series analyses; I'm
not sure why he didn't use them. DeLong is an economic historian, not
One important observation stems from something even simpler than
> Basically what we know about expected returns is that on the one
> occasion when CAPE rose above 30, the dot-com crash of 2000 was in the
> near future and the housing crash of 2008 came into the ten-year
> return window. That is not much information on which to base a
> long-run "sell" decision.