Paul: From your buddy Albert:
06:22am EST 22-Nov-99 Lehman Brothers (James L. Barlage, CFA (212) 526-60) INTC
Intel Corp: Part I/III. Upside Remains Strong. Buy Rating Reiterated.
Ticker : INTC Rank(Old): 1-Buy Rank(New): 1-Buy
Price : $80 52wk Range: $90-48 Price Target (Old):$100
Today's Date : 11/22/99 Price Target (New):$100
Fiscal Year : DEC 10 Uncommon Values
------------------------------------------------------------------------------
EPS 1998 1999 2000
QTR. Actual Old New Old New
1st: 0.41A 0.57A 0.57A 0.61E 0.61E
2nd: 0.33A 0.51A 0.51A 0.63E 0.63E
3rd: 0.45A 0.55A 0.55A 0.71E 0.71E
4th: 0.60A 0.63E 0.63E 0.85E 0.85E
------------------------------------------------------------------------------
Year: $1.77A $2.26E $2.26E $2.80E $2.80E
Street Est.: $2.26E $2.26E $2.66E $2.66E
------------------------------------------------------------------------------
Price (As of 11/19): $80 Revenue (1999): 29.5 Mil.
Return On Equity (98): 29.2% Proj. 5yr EPS Grth: 25.0 %
Shares Outstanding: 3472.0 Mil. Dividend Yield: 0.1 %
Mkt Capitalization: 277.8 Bil. P/E 1999; 2000: 35.4X; 28.6X
Current Book Value: $8.40/sh Convertible: None
Debt-to-Capital: 10.3 % Disclosure(s): G, C
------------------------------------------------------------------------------
Highlights:
* Recently introduced Coppermine products could intensify competitive
pressure on AMD from a performance and cost perspective.
* As Intel's competitive position strengthens, a multiple enhancement could
develop. This could occur in the first half of the year 2000.
* As Intel's multiple achieves a more appropriate level, its common stock
could drive toward our 12-month target of $100.
* We reiterate our 1-Buy rating.
------------------------------------------------------------------------------
Summary:
Intel's share price performance has been sluggish recently as a result of
what we believe are misplaced concerns raised by several industry observers.
Most recently, the shares were downgraded by one follower who was concerned
about what was perceived to be intensifying competitive pressure from
Advanced Micro Devices. This idea was precipitated by Advanced Micro devices
recent indication that its fourth quarter was developing better than
expected.
In the non-microprocessor area, this strength is coming mainly from flash
memory. On the microprocessor side, AMD's K6 family is showing flat to
slightly higher sequential unit volume, and its average selling price could
trend slightly higher. This is also better than expected performance, but
because this is a maturing product family, it is only of marginal importance.
We also note that this improvement is likely occurring because of a gain in
market share brought about by the temporarily diminished presence of Cyrix
and Centaur because of a change in ownership.
Advanced Micro devices also indicated that its goal of shipping 800,000
Athlon processors was becoming increasingly realistic. Note, however, that
it did not increase its shipment forecast for Athlon products, but only
increased the probability that its prior forecast would be achieved.
Therefore, while the fourth quarter upside surprise is a positive development
at AMD, it appears that it is being driven only marginally by Athlon products
which is the area that is absolutely key to the company's longer-term
survival.
In our view, the competitive pressure on AMD from Intel is increasing, not
the other way around. The vehicle for this, in our opinion, is the recently
announced "Coppermine" family of microprocessors. While AMD may still have
some Athlon momentum into the first quarter of 2000, we believe that sometime
in the first half of the year it will be evident that Athlon growth is
stalling. If this occurs, we expect the multiple on Intel's shares to
improve. Therefore, we believe that Intel could be at the threshold of an
important multiple transition that could drive its common stock to higher
than generally expected levels over the next few months and toward our 12-
month $100 price target.
Intel's recently announced "Coppermine" microprocessor line (introduced on
October 25, 1999) is the most extensive microprocessor introduction in the
company's history. This consisted of 15 products in desktop, mobile and
server/workstation versions. Although much of the performance improvement
offered by these products is evolutionary, we believe that the introduction
is significant.
With clock speeds up to 733 MHz, it enabled Intel to recapture the
performance leadership from Advanced Micro Devices' 700 MHz Athlon device.
Although it is not clear how well AMD was actually yielding at this speed
grade and, therefore, how real the product was, the image of Intel has
probably been enhanced by its recovery of performance leadership. More
important, we believe, is the broad-based performance improvement that
characterized the Coppermine products as a result of higher clock speeds, on-
die second-level cache memory and an improved frontside bus interface.
Following early next year will be a variable voltage/clock speed feature on
its mobile products called SpeedStep (previously code-named Geyserville).
This feature will enable a mobile computer to run at two different clock
speed/voltage levels -- a high-speed, high-voltage level when the computer is
plugged into a wall socket and a low-voltage, lower-speed mode for power
conservation when operating with a battery. The clock speed differential
between the two modes might approach two-to-three speed grades and will serve
to close the speed gap that currently exists between mobile and desktop
equipment.
The Coppermine products are also the vehicles for the transition of Intel's
manufacturing process from 0.25 micron to 0.18 micron technology. In
addition to the performance enhancements, this transition will bring
significant cost reduction opportunities to the company. The transition of
Intel's entire product line to 0.18 micron Coppermine technology should
largely be completed by the end of 2000. Because of the cost reduction
opportunities of Coppermine, we continue to believe that 2000 will be
characterized by a generally rising gross margin at Intel. We believe that
these broad-based performance and cost enhancements bought on by Coppermine
have significantly increased the barrier that Advanced Micro Devices must
negotiate in order to penetrate into the higher-performance commercial region
of the microprocessor market which we believe is necessary for its survival
as a long-term supplier to the PC market.
We believe that because of these products and other reasons, it will be
difficult for Advanced Micro Devices to achieve substantial penetration into
the commercial PC market. We expect this to become increasingly evident in
the first half of 2000. And, if we are correct in this assessment, the
multiple accorded Intel's common could increase significantly as investors
conclude that AMD's struggle for microprocessor market share is intensifying.
In addition, we expect the worldwide PC market to remain strong in 2000 as
the Asia-Pacific and Japanese regions improve economically and e-commerce
continues to drive PC penetration rates into those and other regions of the
world. Finally, Intel's communication business could achieve sales of about
$3 billion in 2000, almost double 1999 results, which could equal about 9% of
total revenue, up from 6% in 1999. Consequently, the company's
communications business could make a meaningful contribution to growth in
2000.
On balance, then, we expect Intel's revenue growth to accelerate during the
next two years to rates of 17% in 2000 and 20%-25% in 2001. We expect
earnings to grow faster as margins expand and believe that the company's
longer-term earnings growth can approximate 25% annually. If Advanced Micro
Devices does falter early in 2000, as we believe is quite possible, and if
the expected multiple expansion of Intel's common occurs as a result, Intel's
stock could perform well over the intermediate term. Consequently, we
continue to recommend purchase of its shares in aggressive accounts and
maintain our 1-Buy investment rating.
Coppermine Technology
Coppermine is the code name used for the upgrade of Intel's Pentium III
family from the initial Pentium III family code named Katmai. In general,
Coppermine is manufactured using 0.18 micron processing technology (instead
of 0.25 micron technology); has 256 Kbytes of on-chip second-level cache
memory (instead of off-chip SRAM cache memory); is housed in a flip-chip pin
grid array package (instead of a cartridge); and has a 133 MHz frontside bus
(instead of a 100 MHz frontside bus). The net result is substantial cost and
performance improvements over the existing Pentium III products. Because of
these enhancements, Coppermine products will essentially be substituted for
Katmai during the course of 2000.
For the desktop, Intel introduced nine performance (non low-end Celeron)
products. Four of these were new speed grades, ranging from 650 MHz to 733
MHz, extending what had been a top performance level of 600 MHz. Five
overlapping products at 500 MHz, 533 MHz, 550 MHz, and 600 MHz were also
introduced but, in fact, the Coppermine versions outperform the Katmai
versions by as much as 25% because of the on-chip cache memory and advanced
bus structure. Prices have been set to encourage the market to transition to
Coppermine. No Celeron (low performance) Coppermine desktop products were
introduced.
Intel added three products to its mobile line, establishing for the first
time Pentium III-based processors for this market. Previously, mobile
products had been confined to the low-performance Celeron and Pentium II
architectures. Like on the desktop, Coppermine technology has significantly
raised the performance level of Intel's mobile products. Prior to these
introductions, the highest performance mobile processor was a 400 MHz Pentium
II with 256 Kbytes of integrated cache (Celeron mobiles with clock speeds of
433 MHz and 466 MHz had been available but the absence of the integrated
cache lowered their performance below that of the 400 MHz Pentium II). The
mobile Coppermine introductions added processors with 400 MHz, 450 MHz and
500 MHz clock speeds.
06:22am EST 22-Nov-99 Lehman Brothers (James L. Barlage, CFA (212) 526-60) INTC
Intel Corp: Part II/III. Upside Remains Strong. Buy Rating Reiterated.
Ticker : INTC Rank(Old): 1-Buy Rank(New): 1-Buy
Price : $80 52wk Range: $90-48 Price Target (Old):$100
Today's Date : 11/22/99 Price Target (New):$100
Fiscal Year : DEC 10 Uncommon Values
------------------------------------------------------------------------------
EPS 1998 1999 2000
QTR. Actual Old New Old New
1st: 0.41A 0.57A 0.57A 0.61E 0.61E
2nd: 0.33A 0.51A 0.51A 0.63E 0.63E
3rd: 0.45A 0.55A 0.55A 0.71E 0.71E
4th: 0.60A 0.63E 0.63E 0.85E 0.85E
------------------------------------------------------------------------------
Year: $1.77A $2.26E $2.26E $2.80E $2.80E
Street Est.: $2.26E $2.26E $2.66E $2.66E
------------------------------------------------------------------------------
Price (As of 11/19): $80 Revenue (1999): 29.5 Mil.
Return On Equity (98): 29.2% Proj. 5yr EPS Grth: 25.0 %
Shares Outstanding: 3472.0 Mil. Dividend Yield: 0.1 %
Mkt Capitalization: 277.8 Bil. P/E 1999; 2000: 35.4X; 28.6X
Current Book Value: $8.40/sh Convertible: None
Debt-to-Capital: 10.3 % Disclosure(s): G, C
------------------------------------------------------------------------------
Part II of III
An added advantage of 0.18-micron processing technology is low voltage
operation which is especially important for battery power conservation in
mobile personal computers. Although the 400 MHz Coppermine overlaps in speed
grade with 400 MHz Pentium II, it operates at 1.35 V which results in about a
32% drop in active power. In addition to use in normal sized notebook
computers, this voltage level opens the mini-notebook market to Intel. The
higher clock speed processors operates at higher voltage levels but below
those of equivalent 0.25-micron (Katmai) Pentium III processors.
Because higher clock speeds consume more power and dissipate more heat, the
performance of portable personal computers has historically lagged desktop
performance at any point in time. At the moment, Intel's highest performance
desktop processor, with a clock speed of 733 MHz, is seven speed grades, or
almost two years, ahead of its top performance mobile product with a clock
speed of 500 MHz. However, the company has developed a new mobile processor
technology, known as SpeedStep, that could significantly close this gap.
SpeedStep, formerly known by the code name Geyserville, is a technology that
allows the processor to operate in two voltage/clock speed modes. A low
voltage, lower clock speed mode of operation can be used when the computer is
being powered by a battery in order to extend the battery life. When wall-
socket power is available and, therefore, power conservation is not an issue,
the processor can operate at a higher performance level. SpeedStep
technology was not included with the recent Coppermine introductions,
however, it could come in early 2000. It is not known precisely how high the
performance increase could go with SpeedStep, but perhaps two-to-three speed
grades is a reasonable estimate. This could significantly increase the
performance leadership of Intel's mobile products. Moreover, it could foster
increased substitution of the more flexible portable computers for desktop
units which would also be positive for Intel because of its relatively
stronger position in the portable market.
At the high end of the processor family, focused on server/workstation
applications, Intel introduced three Coppermine versions of its Xeon line.
These included 600 MHz, 667 MHz and 733 MHz clock speeds, whereas previously
the highest clock speed had been 550 MHz. Like the other Coppermine
products, these devices have 256 Kbytes of on-chip second level cache memory.
Consequently, although their clock speeds are higher than the existing Katmai
products, their performance does not exceed the top end of the existing line
where off-chip cache memory, ranging from 512 Kbytes to 2 Mbytes, can enhance
the performance of a slower clock speed processor. In other words, a 500 MHz
Katmai processor with 2 Mbytes of off-chip cache can outperform a 733 MHz
Coppermine processor with 256 Kbytes of on-chip cache. While processors with
on-chip cache memories of up to 2 Mbytes will be introduced in 2000, for the
moment, the newly introduced Coppermine products will replace the existing
500 MHz and 550 MHz Katmai products with off-chip cache memories of 512
Kbytes. This makes them most appropriate for workstation applications. The
larger on-chip cache versions, to be introduced next year, will be applicable
to the server market and displace those Katmai products with equivalently
sized off-chip cache memories.
Coppermine Provides Cost Benefits
In addition to the performance enhancements offered by the 0.18 micron
Coppermine version of Pentium III, a substantial manufacturing cost advantage
is gained. This occurs because, despite the on-chip cache memory, a 17%
smaller die increases the gross die per wafer by about 30%. This also
enhances yields which increases the net die per wafer by more than 30%. In
addition, the on-chip second-level cache memory of the Coppermine products
enables the device to be packaged in a flip-chip pin grid array rather than
the expensive cartridge currently used in Katmai Pentium III products.
Associated with this is the elimination of separate SRAM memory that is
housed in the cartridge.
By the fourth quarter of 2000, Intel will have largely converted its
microprocessor output to 0.18 micron process technology and eliminated
cartridge packaging. At that time, the company will have achieved
approximately a 50% reduction in the unit cost of a Pentium III processor
from current levels and a 20% decrease in Celeron unit costs. Intel
obviously has a choice of passing these cost reductions on to its customers
in the form of lower prices or keeping them and allowing its gross margin to
rise. In fact, a combination of these options will likely occur and we,
generally, expect Intel's gross margin to increase during the course of 2000.
Coppermine Extends Product Leadership
Intel has typically increased the performance of its products by about one
speed grade per quarter. However, the Coppermine introduction greatly
exceeded this pattern as the desktop performance was enhanced by four speed
grades and the mobile line by three. This puts Intel back in a desktop
leadership position relative to Advanced Micro Devices that had announced a
700 MHz Athlon processor and extends its leadership in mobile products.
Announcing a device and shipping it in volume are two different things and it
is not clear how many 700 MHz parts AMD was actually delivering and/or if
they were being manufactured cost effectively. However, AMD is also in the
process of bringing on 0.18 micron manufacturing technology and, therefore,
should reap some of the same cost/performance benefits as Intel. However,
when considering the production problems Advanced Micro Devices has
historically experienced when attempting to ramp a new technology, our
assessment is that AMD has likely returned to a mode of trying to catch up to
Intel's rapidly moving technology target rather than leading it. This could
impair its ability to substantially penetrate the performance segment of the
microprocessor market and push its accompanying average selling price down to
the low end of the range. We also believe that Athlon's lack of pin
compatibility with Intel processors could negatively influence demand and it
clearly puts increased technological pressure on AMD.
Advanced Micro Devices Challenges Have Increased
Previous to the Athlon, AMD's microprocessor products were pin-compatible
with Intel's. Consequently, the same PC platform (i.e., motherboard,
chipsets, etc.) could be used to support the microprocessor products of both
companies. This made it easy on the user since the only decision necessary
was whether Intel's or AMD's processor would plug into the empty socket on
the platform. However, the Athlon can no longer be attached to an Intel-
based platform and, therefore, AMD and/or its partners have been required to
develop their own platform solutions. This requires AMD to achieve
performance parity at the systems not just the chip level. It also increases
AMD's capital, technological and time-to-market risk from what it has
historically assumed. Moreover, Athlon users must now inventory both AMD and
Intel-based platforms which increases Athlon associated carrying costs and
obsolescence risks from those previously assumed. In addition, should AMD
fail to deliver the desired volume of microprocessors, as has occurred in the
past, the user could lose a sale rather than being able to stuff the board
with an Intel processor. These disadvantages will reduce the incentive to
use Athlon microprocessors and, therefore, increase AMD's already difficult
marketing challenge. Lastly, commercial PC customers will be more resistant
to mixing different platforms in their networks because of potential
compatibility issues.
06:22am EST 22-Nov-99 Lehman Brothers (James L. Barlage, CFA (212) 526-60) INTC
Intel Corp: Part III/III. Upside Remains Strong. Buy Rating Reiterated.
Ticker : INTC Rank(Old): 1-Buy Rank(New): 1-Buy
Price : $80 52wk Range: $90-48 Price Target (Old):$100
Today's Date : 11/22/99 Price Target (New):$100
Fiscal Year : DEC 10 Uncommon Values
------------------------------------------------------------------------------
EPS 1998 1999 2000
QTR. Actual Old New Old New
1st: 0.41A 0.57A 0.57A 0.61E 0.61E
2nd: 0.33A 0.51A 0.51A 0.63E 0.63E
3rd: 0.45A 0.55A 0.55A 0.71E 0.71E
4th: 0.60A 0.63E 0.63E 0.85E 0.85E
------------------------------------------------------------------------------
Year: $1.77A $2.26E $2.26E $2.80E $2.80E
Street Est.: $2.26E $2.26E $2.66E $2.66E
------------------------------------------------------------------------------
Price (As of 11/19): $80 Revenue (1999): 29.5 Mil.
Return On Equity (98): 29.2% Proj. 5yr EPS Grth: 25.0 %
Shares Outstanding: 3472.0 Mil. Dividend Yield: 0.1 %
Mkt Capitalization: 277.8 Bil. P/E 1999; 2000: 35.4X; 28.6X
Current Book Value: $8.40/sh Convertible: None
Debt-to-Capital: 10.3 % Disclosure(s): G, C
------------------------------------------------------------------------------
Part III of III
On balance, then, we believe that the Coppermine introduction has intensified
an already difficult competitive situation for AMD. We believe that the net
effect will be limited penetration into the commercial PC market and,
therefore, a PC market presence that will remain largely confined to the
consumer sector. If so, AMD's sales and earnings would remain under pressure
and the company would fail to achieve its stated volume goals. If these
conditions do, indeed, develop, we believe that it could be evident sometime
in the first half of 2000.
Many observers view AMD's survival as being dependent on the success of its
Athlon product line. As such, if it appears that it is faltering in the
market place, a higher multiple could be accorded Intel's shares since the
company's business model would be significantly enhanced if AMD's market
presence weakened. We think the probability is quite high that this
condition could develop in the first half of 2000. Therefore, we believe
that Intel's common could be driven higher than generally expected levels
during the next few months by multiple expansion.
The Business Environment Remains Strong
The underlying business environment remains strong for Intel. The personal
computer market is showing the strongest unit growth since 1995 with a 23%
year-over-year gain for the first nine months of the year. We expect this
growth rate to be maintained in the fourth quarter and for the year estimate
a 23% worldwide growth rate to 114.6 million units. This would compare to a
16% worldwide unit growth rate in 1998.
We do not expect this growth rate to be maintained in 2000 but do expect a PC
unit market growth rate of about 18% and believe that growth at the high end
of a 15%-20% range can continue in the following years. In our opinion, the
basis for this growth is under-penetration of the worldwide population with
personal computers relative to the U.S. with roughly 500 PCs per 1000 people
in the U.S. versus 65 PCs per 1000 people worldwide. While some of the third
world regions will not be large consumers of PCs in the foreseeable future,
many advanced economies like Europe and Japan have only about 50% of the
penetration rate of the U.S. And, the driving force stimulating further
penetration is the Internet and e-business. In addition, Intel's penetration
of the server/workstation market with microprocessors will slightly add to
its unit volume and more substantially to sales because of high average
selling prices. If total microprocessor unit growth is maintained at the
high end of a 15%-20% growth rate, average selling prices could drift
slightly lower. However, microprocessor revenue growth could still be
maintained well within a 15%-20% growth rate for 1999 and 2000. As Itanium
(previously code-named Merced) volume builds in 2001, Intel's microprocessor
sales growth rate could increase to the high teens or low twenties in
percentage terms.
Communications Products Will Add To Growth
Intel has expanded its business model to focus its attention on becoming a
building block supplier to the Internet as contrasted to its former mission
of supplying only the PC market. Among other things, this has intensified
Intel's interest in the networking/communications market for semiconductors
and systems. To date, sales generated by networking and communications
products have been small relative to Intel's total and, therefore, relatively
unimportant in driving its growth. However, beginning in 2000, the sales of
these products could reach a significant enough size to begin having a
measurable impact on Intel's growth rate. We estimate that in 1998,
communications/networking sales accounted for about $900 million, or 3% of
total company revenue. Because of acquisitions and internal growth, these
sales could total about $1.7 billion, or 6% of total company revenue in 1999.
These could reach about $3.0 billion, or 9% of total sales in 2000 and,
therefore, add 3%-4% to the company's overall sales growth rate.
In total, then, we expect Intel's revenue growth to accelerate during the
next two-to-three years and stabilize within a 20%-25% range beyond that
time. We expect 12% revenue growth in 1999 to $29.5 billion, 17% in 2000 to
$34.5 billion and a 20%-25% growth rate in 2001. We are estimating that
earnings per share (before unusual items) will expand 28% in 1999 from a
depressed $1.77 per share to $2.26 per share. We are estimating a 24%
earnings growth in 2000 (not from a depressed level) to $2.80 per share.
Recall that revenue is expected to grow 17% in 2000 as expanding margins
drive earnings to a higher growth rate. Longer term, we expect earnings per
share to grow at an average annual rate of about 25%.
A Multiple Expansion Could Occur
Intel's multiple has, historically, traded at a discount to a theoretically
more appropriate value based on its growth rate and stature in the industry.
In general, this has been the case because Intel's position has been so
strong that there has been no place to go but down. Therefore, over the
years, fears regarding circumstances that could weaken this position such as
litigation issues, technological substitution issues or microprocessor
competitors have put pressure on the valuation of Intel's common.
Most recently, the presence of AMD as a potentially more significant
competitor, has played a part in this pressure. However, if, as expected in
the first half of 2000, it becomes increasingly apparent that AMD's Athlon
challenge is weakening, some of this multiple pressure could lift.
Currently, Intel's common is selling at 28.6x 2000 earnings vs. a multiple on
the S&P 500 of 24.9x. This translates into a multiple premium relative to
the S&P 500 of only 13% when its longer-term growth rate may be triple that
of the market. Viewed another way, Intel's common is selling at 1.13x its
longer-term growth rate vs. about 3.1x the growth rate for the S&P 500.
Other high quality semiconductor companies are currently selling at
valuations in excess of 2.0x their growth rates on 2000 earnings. Were Intel
to achieve a valuation along these lines, a multiple of 50.0x earnings would
result, which would indicate a current stock price of about $140. This is
significantly higher than our 12-month price target of $100 and underscores
the appreciation potential that could exist if AMD's bid for significant
market penetration by its Athlon chip falters. |
