The Chip Business: Bouncing Back…Or Not?
Roughly two weeks ago, the Financial Times was
trumpeting that
the chip business is bouncing back, in an article titled
“Chip
sales expected to reach new highâ€.
Yes, the bandwagon
of longer-term semiconductor bulls is getting pretty full. Almost everybody is
on-board. Driving top-line growth is nice, but profits need to be part of
the process also.
I have concerns
with comments like the following:
“…driven by
better-than-expected demand in a number of important end markets, including
personal computers and wireless handsets” — I’m hearing cell phones are
piling up
“Excess
inventories had largely been worked off in the first quarter“. A lot of
people are parroting this line but, as discussed by Denali a several months ago,
inventory can now hide in so many places it is impossible to truly track
globally. I might start to believe the inventory claims if they cited examples
of exactly who got rid of inventory…
“… chip
suppliers Renesas and Toshiba bucked the trend, posting revenue rises of 10 per
cent and 8.6 per cent respectively“. How is growth bucking the trend? The
growth part of their sales is NAND; reports the last two weeks of NAND price
weakness imply the good times may go on pause; iPod Shuffles are starting to
pile up, the
regular iPod is on sale at Fry’s, which previously only sold them at full
retail/list price It will be really(!) “interesting” to see if Apple’s sales
slow down and, if so, it’s dismissed as due to the Intel conversion rather than
iPod declines. (I would honestly be impressed with Apple’s management if/when iPod
slows and it’s admitted. The Intel conversion might even be thinking
ahead. Actually taking their medicine all at once rather than denying the
truth…)
Missing from this
is TSMC/UMC lead times “slipping” to 8 to 10 weeks. First, this is still just
standard lead time, or less, thru a fab; second, I’m convinced this is insurance
production in case there really is a second half 2005 demand spike; not proof
there will be one.
Also missing is
any mention of April sales being lower than January. Although I just reviewed
the numbers since 1976 and the April average is often lower than the March
average, even in good years. On the other hand, the April average has been lower
than January’s average only 4 times since 1976 –1985, 1996, 1998 and 2001.)
Everything still
predicated on the second half “recoveryâ€, something the pundits have been
predicting each year since the bubble burst .I guess we’ll see.
Just my opinion. I
could be wrong.
I think we get one
more sharp rally / spike of +20% to 25% in semis, then a major sell-off again in
2006. In terms of industry outlook, it’s hard to quantify much above 10% unit
growth in the areas that consume 70% of semi content (PC’s and handsets), and
ASPs are not going up. So the best case industry growth looks like around 10% in
my view. The reason I think we get a rally: I think we’ll see inventory
snapback to normalized levels at OEMs, EMS, and distributors for fourth quarter
build.Â
I’d expect a
sell-off in early 2006 with no material upward revisions. Current Street numbers
bake in near peak margins for 2006, so it’s hard to see any material upside for
semiconductor stocks as a group once we move into early 2006.
Melanie
Hollands
melanieholldands@yahoo.com