ML:Comm Rev is a grab bag of biz fr Shiva and Dialogic hw to HN boxes to Level One PHYs
Excerpts from ML 11/2/00
12 Month Price Objective: $50
Estimates (Dec) 1999A 2000E 2001E
EPS: $1.17 $1.69 $1.64
P/E: 38.4x 26.6x 27.4x
Investment Highlights:
• Intel’s executive webcast contained no
financial surprises – comments on the fourth
quarter outlook were unchanged.
• More interesting were comments on the
upcoming P4 ramp, the move to 0.13 micron
manufacturing and the growth target for
Intel’s core MPU business.
• We reiterate our intermediate-term and long-term
accumulate ratings – there is no reason
to be more aggressive at this point.
Fundamental Highlights:
• We continue to believe that Intel’s major
intermediate-term challenge will be ramping
P4 without impacting gross margin. Although
yesterday’s presentation included some
interesting cost data, the gross margin
question remains unanswered.
• CEO Craig Barrett discussed a 10% growth
target for the Intel Architecture Business
Group, and a much higher 50% target for the
communications business. Hitting that 50%
number will be a challenge, even for Intel.
No earnings surprise . . .
Expectations for another dramatic financial announcement
from Intel during the company’s executive webcast
yesterday were dashed when the company offered
unchanged observations about the state of PC demand for
the quarter. More interesting were Intel’s observations
about the ramp of the P4, and the rates of growth that the
company is targeting for the PC and non-PC businesses.
We reiterate our intermediate-term and long-term
accumulate ratings – Intel is doing a good job of trying to
move its focus away from the PC, but it faces enormous
challenges in growing its newer businesses at a rate
necessary to keep the overall top line moving. We also
heard nothing yesterday to alleviate our concerns regarding
gross margin pressures during 2001.
P4 ramp is steep
Intel still is committed to ramping the P4 in volume during
2001, despite the fact that the company’s 0.13 micron six-layer
copper process is not going to ramp until the fourth
quarter – that is slightly later than we had expected. Intel
is talking about P4 unit volume exceeding PIII volume by
the first quarter of 2002, which is a little later than our Q4
2001 estimate. That ramp schedule implies substantial P4
volumes by the end of 2001, perhaps exceeding 10m units
by the fourth quarter, with the vast majority of those parts
on the older 0.18 micron process. Combined with
depreciation that will be going up next year in dollar terms,
we cannot imagine how Intel will steer clear of pressure on
its gross profit margin. The planned ramp of 0.13 micron
manufacturing in late 2001 and 2002 is steep, with 8 fabs
coming up on the process – we think that the result has to
be capital spending of at least $6.5 billion in 2001.
The depreciation and cost data that Intel showed yesterday
were interesting, but were not sufficient to change our
thinking. Intel pointed out that depreciation will comprise
roughly the same percentage of total unit cost in 2001 as it
has in 2000, without talking about the unit assumptions
that underlie that calculation. Also shown were data
depicting total weighted average unit costs, which Intel
believes will fall slightly over the next few quarters before
beginning to rise again. The crucial questions – what set
of unit and pricing assumptions it takes to get to Intel’s
numbers – remain unanswered.
Interestingly, one approach that we expected Intel to
discuss – drawing down Celeron manufacturing volumes –
was not discussed yesterday. Reducing exposure to
Celeron, which carries a lower gross average margin than
the rest of Intel’s businesses anyway, would be an easy
way to free up capacity for P4 and support overall gross
margins at the same time. Paul Otellini insisted, however,
that Intel will continue to refresh its Celeron product line.
50% growth for the networking business?
We were pleased to see that the revenue breakdown that
the company published closely aligned with our own
model. We estimate that the most recent quarter’s
revenues were 74% microprocessor, 6% chipsets and
related products, 6% flash, 3% non-x86 processors, and
11% “other”, mostly communications products and
hardware. Characteristically, Intel did not furnish
investors with actual numbers, but a look at the pie graph
reveals that our estimates are reasonable.
Intel CEO Craig Barrett observed that he has set a target of
10% YoY growth for the Intel Architecture Business
Group going forward, which may perhaps be lower than
some investors have expected. The strategy for holding
overall top line growth in the mid-teens thus centers
around the newer businesses, which Barrett said he has
targeted for growth in the 50% YoY range.
That will be a mighty challenge, even for Intel. One first
needs to back out the flash business, which is growing at
100% YoY currently but which is unlikely to sustain that
rate for the next two years. Intel’s embedded processor
business, at 3% of sales, does not grow at 50% either.
There are some hot spots in the networking
communications business, to be sure, but the roughly
$1billion per quarter in “other” revenues is not pure, high-octane
communications semiconductor revenue
comparable to Broadcom or PMC Sierra. It is a grab bag
of businesses ranging from Shiva and Dialogic hardware to
home networking boxes to Level One PHYs. It is also big,
at $3.5 billion in revenue this year.
We like our $1.64 EPS estimate for next year, and we like
our accumulate rating. Intel is doing as good a job as any
company could possibly do in dealing with the challenges
that it faces. However, we have a difficult time
understanding how the earnings multiple can re-expand
when consensus estimates for 2001 are too high, and when
Intel’s ability to grow its large communications business is
unclear. |
