Andrew,
I have not yet thought through all the implications
of your insightful post, but I have one question/comment/speculation:
You say that all other things being equal, 10% R&D and 40% COGS
is better than 50% COGS, where equal means equivalent pretax earnings.
Now we have to be very careful:
My understanding is that capital expenditures, etc., are incurred now,
but depreciated later, partly as R&D expenses. So if I look at
*current* accounting earnings, they don't necessarily include
all the current R&D costs?
[That was the question part. An accounting question, I believe.]
The catch is that capital expenditures (like revs and earnings) are
growing very fast. So current reported accounting earnings, not
including all the costs incurred in the current period, are actually
overstated. It's sort of like being in a high-inflation environment, where
reported earnings overstate the true picture. [this is the speculation
part.]
Over time this washes out, but while Intel is in a pseudo-inflationary
(i.e. high-growth) environment, the effect is sort of like
check-kiting: reported earnings appear higher than they "really"
are (vs. FCF) because depreciation hasn't caught up yet.
Thus a lower P/E.
In contrast, KO's advertising is charged as cost-of-sales in the
current period. So if I compare a KO and an Intel, I expect
KO to trade at a higher PE, strictly on accounting effects.
Does this make sense?
--joel |
