GS US Semi Equipment SEMICON West Day
1 - DTja Vu 12 months later
Monday was a busy day with 3 negative preannouncements from Brooks, FEI, and TEL, a
merger announcement, and the beginning of Semicon West. Unfortunately, this July is
eerily reminiscent of last July when SPE companies missed Q2 and guided for a soft Q3.
Like last July, hope springs eternal and managements are again suggesting that the upturn
is ?right around the corner.? Even with 1) mgmts and many sell-siders having wrongly
pounded the table on the upturn for the last 12 mths, 2) the S&P SPE index
underperforming the S&P by 19% Y/Y, and 3) 6 SPE companies having already
negatively preannounced Q2, there are still plenty of bulls and the stocks are trading
strongly despite worse than expected fundamentals. We expect stocks to trade down to
their cycle lows, as we believe hopes for an imminent recovery will be disappointed as
memory orders decline more than foundries increase in H2?05.
SEMICON West, a Semi Equipment industry trade show hosted by Semiconductor
Equipment and Materials International (SEMI), began on Monday with Novellus Systems
hosting an analyst meeting and several companies hosting 1x1 meetings. The show will
run through the end of this week in San Francisco, with the majority of presentations
taking place Tuesday through Thursday.
There was plenty of activity on Monday in addition to the beginning of the show, as three
companies negatively preannounced and Brooks Automation also announced on Monday
that it will acquire Helix Technology (more details below). Brooks negatively
preannounced its CQ2 results in addition to the merger and FEI negatively preannounced
a revenue and earnings shortfall after the close on Monday night. In addition, Tokyo
Electron preannounced a significant order shortfall in semi equipment bookings with SPE
orders down 28% q-o-q as DRAM orders declined and foundry orders remained weak. We
would highlight that 6 semi equipment companies have now negatively preannounced a
Q2 sales/earnings/or a bookings shortfall in the last few weeks. Despite these shortfalls
(remember that when the companies missed their March quarter bookings and guided well
below consensus for June, almost all of the companies insisted once again that the cycle
had bottomed), the majority of companies and sell-siders continue to insist that the upturn
is "right around the corner." We already heard several times at Semicon (as well as on the
Brooks negative preannouncement call) that even though CQ2 came in softer than
expected, and that CQ3 now looks to be softer than originally expected (despite the fact
that all of these same companies insisted that March was the bottom), that the upturn will
now begin in CQ4. After 12 months of calling for an imminent fundamental upturn, we
would have expected some level of capitulation from the equipment companies and/or the
sell-side, but that hasn't come yet.
Below we provide details from the Novellus Systems and Photon Dynamics analyst
meetings, as well as our thoughts on the Brooks Automation acquisition of Helix and
Brooks' negative CQ2 preannouncement.
NOVELLUS KICKED OFF THE SEMICON WEST TRADE SHOW ON MONDAY
NIGHT WITH A PRODUCT-FOCUSED ANALYST MEETING. Novellus kicked off the
SEMICON West (SW) trade show on Monday night. The company provided little to no
commentary regarding the second quarter or current business conditions. Instead, the meeting
highlighted the company's products, as well as its focus on efficiency and driving operational
improvements. In regards to the state of the semiconductor industry, management noted that the
industry is currently undergoing a fundamental shift whereby the industry is becoming
consumer-driven as opposed to industrial-driven. In light of the shift to a consumer-driven business,
management indicated that the industry is being characterized by short and rapid product cycles,
with customers increasing their focus on productivity and cost effectiveness. To that end, Novellus
continues to be focused on growing its market share by increasing customer penetration with new
cost effective and productive tool offerings. Specifically, management indicated that its current goal
is to grow the company's market share significantly (from 25% in 2004) in the next couple of years.
The company estimates that its served available market in 2005 is approximately $4.9 billion,
including the ECD, CMP, PVD, dry strip, HDP, WCVD, and PECVD sub-segments).
The company dedicated a significant portion of the presentation to four of the company's seven
product lines, including Direct Metals, PECVD, Gapfill, and Integrated Metals. In the Direct Metals
segment, the company highlighted its Altus PNL Tungsten tool (where the company believes it has
70% market share at 300mm) and its DirectFill tool (where Novellus expects to have about 10
systems installed in 2005 and over 20 systems installed in 2006). In the PECVD segment, the
company highlighted its VECTOR tool (with over 400 systems shipped in June 2005) and its
CORAL low-k tool (where the company believes its presence is under-estimated by the market). In
the Gapfill segment, the company discussed its SPEEDNExT tool (scalable to 45-nm). In the
Integrated Metals segment, the company discussed its INOVA platform (where the company had 6
300mm customers in the beginning of 2004 and expects to have 18 customers by the end of 2005).
Of the company's products, management indicated that it believes the PVD and CMP (where the
company has yet to gain traction) segments are the most likely to drive growth.
Novellus also discussed its efforts at improving operating efficiencies during the analyst meeting.
The company is attempting to improve its operational performance by reducing cycle times,
optimizing its supply chain by reducing its supplier base, consolidating its facilities from 5 factories
to 2 factories, and driving an "order to cash" cycle time reduction (or reducing how long it takes
from when a customer places an order with Novellus to when Novellus receives a cash payment for
the tool from the customer).
Management also reiterated its target gross margin of 52-54%. We would note that the company
achieved peak gross margins in current cycle of approximately 50% vs. peak gross margins of
approximately 60% in the previous cycle. The company has indicated that it is facing pricing
pressure due to the irrational pricing behavior of Applied Materials. We remain concerned that
pricing pressure is driving secular margin erosion that is unlikely to be fully offset by improvements
in operational efficiencies. Further, the company continues to try to penetrate customers with new
products and management indicated that the company has to "earn its way" into an account and does
not achieve high profitability on evaluation tools. We would expect gross margins to improve
if/when the company achieves greater critical mass in new product segments.
PHOTON DYNAMICS ANALYST MEETING HIGHLIGHTS GROWTH PROSPECTS WITH
THE ADOPTION OF FPD TVS. We attended the Photon Dynamics analyst meeting held on
Monday morning, during which the company provided an overview of its business and potential
drivers of growth. Photon Dynamics makes yield management solutions for the Flat Panel Display
(FPD) industry, specifically test and repair tools used during the FPD manufacturing process to
improve customers' yields. The company's test equipment is used to identify defects early in the
manufacturing process while its repair systems use laser technology to repair defects in flat panel
displays during fabrication. Note that the company's revenue mix between test and repair systems is
about 75% test and 25% repair.
Photon Dynamics' business is driven by LCD maker (i.e., Samsung, LG Philips, Sharp, AUO, Chi
Mei, etc.) investment in LCD equipment capex. Management estimates that LCD equipment capex
will decline to approximately $8.2 billion in 2005 from $11.1 billion in 2004 (this compares to $6.2
billion in 2003, $5.2 billion in 2002, and $3.8 billion in 2001). We walked away from the meeting
believing that the company expects the second half of the year to be stronger than the first half.
Management currently expects LCD capital spending to increase about 10-15% year over year in
2006. To that end, Photon Dynamics is tracking 33 new FPD lines that are expected to be
constructed over the next few years, of which 10 are expected in Korea, 5 in Japan, 16 in Taiwan,
and 2 in China.
Photon Dynamics competes against both Applied Materials and Agilent. Management indicated that
Applied is very active in the space, and in 2005 Applied replaced Agilent as a minority market share
leader.
Management indicated that the primary driver of FPD growth, and in turn its business, has been
desktop PC monitors. However, over the next few years, management believes that the primary
driver of FPD growth will be TVs. The company estimates that about 50% of LCD capex in
2005 will be invested in fabs intended primarily to produce flat panels for TVs.
Consumer adoption of TV LCDs is being driven by reductions in prices. Consequently, Photon
Dynamics' customers are focused on reducing their costs thereby enabling them to pass on price
declines to their customers. The company indicated that FPD capex on a per area basis is no longer
being reduced with each new generation, so customers are forced to look elsewhere to drive down
their costs.
Photon Dynamics believes that it offers its FPD customers a way to reduce costs by improving
yields. According to management, yields in FPD fabs are around 80%, with low yields being driven
by the difficulty of keeping the facility clean, pixel designs becoming more complex, and glass sizes
becoming too large for human inspection. The company estimates that a 2% improvement in its
customers' yields can drive $81 million in incremental revenues and $7 million in incremental gross
margins.
Management indicated that a Gen 5, Gen 6, and Gen 7 FPD line typically requires about 3-4 of
Photon Dynamics' test systems. The ASP on the company's Gen 5 test systems is about $1.7 million.
A Gen 5 FPD line typically requires about 3-4 repair systems per line, which have an ASP of about
$0.5 million. Gen 6 and Gen 7 lines are requiring an increasing number of repair systems, or about
6-8 systems per line. Management indicated that it is seeing a 20-40% increase in ASPs in its test
and repair systems across FPD generations. That said, margins are holding fairly flat as the bill of
materials for the systems is also increasing at a similar rate.
BROOKS' ACQUISITION OF HELIX SUPPORTS OUR VIEW THAT THERE CONTINUES TO
BE A NEED FOR CONSOLIDATION IN THE SEMI EQUIPMENT INDUSTRY; BROOKS'
NEGATIVE PREANNOUNCEMENT UNDERSCORES DIFFICULT ENVIRONMENT IN SPE.
Brooks Automation announced an agreement to acquire Helix Technology on Monday morning.
The deal is a stock-for-stock transaction with about a 25% premium paid to Helix shareholders
based on the company's closing stock price on Friday. Helix shareholders will receive 1.11 shares of
Brooks' common stock for every share of Helix common stock (the transaction value of Helix is
approximately $454 million). Post the transaction, Brooks stockholders will own approximately
61% of the combined entity and Helix stockholders will own approximately 39% of the combined
entity on a fully diluted basis.
The deal is anticipated to close in CQ4 2005 and is expected to be significantly accretive to Brooks'
earnings in the first year. The combined company will have trailing annual sales of about $720
million. Additionally, management indicated it expects to realize approximately $10 million in cost
synergies (equally split between fixed costs and COGS) as well as $6 million to $8 million in tax
synergies as a result of the acquisition. The combined entity will be called Brooks Automation and
will be headquartered in Chelmsford, Massachusetts. Ed Grady, president and CEO of Brooks, will
be president and CEO of the combined entity.
Recall that Helix's business is divided into two primary segments, services and hardware. The
services business, which accounts for between 30% and 40% of total sales and is focused on the
end-user (or semi device maker), includes spares, repairs, upgrades, e-diagnostics, and support
agreements. The hardware business focuses primarily on cryopump vacuum systems that are used in
applications such as ion implant, PVD, CVD, and etch, with PVD and ion implant the two largest
market segments into which the company sells its cryopump technology.
Brook's decision to acquire Helix supports our long-held view that there continues to be a need for
consolidation in the semi equipment industry, particularly given the industry's slowing long-term
growth prospects. Recall that the industry has grown at a mere 1% CAGR over the last two cycles.
Additionally, semi equipment customers continue to express a preference for a more consolidated
supplier base. To that end, Brooks' management noted that customers are making it clear that they
want fewer suppliers with more critical mass. As a result, the new Brooks hopes to offer customers
a more comprehensive product offering by supplying customers with vacuum based integrated
subsystems vs. supplying individual vacuum tools separately. The combined company will aim to
leverage Brooks' expertise in vacuum tool automation systems/modules as well as Helix's expertise
in process vacuum technology, thermal management and global services to create a strong supplier
of vacuum based integrated subsystems. Management also indicated that it hopes to transition
Brooks' products to the Helix service model due to Helix's superior service infrastructure, while
Helix is expected to benefit from Brooks' service networks in Eastern Europe and Asia.
We believe the deal makes good sense from a cost synergy and critical mass perspective as the
subcomponent suppliers need to get larger in order to withstand the pricing pressure they are getting
from their OEM customers. However, we believe the revenue synergies are less obvious and we will
need to see the company make progress on selling customers subsystems (as opposed to
subcomponents) before we believe that the deal is beneficial beyond the cost savings the combined
company should recognize.
In addition to the acquisition announcement, Brooks has also revised its CQ2'05 revenue guidance
to $111-$113 million from $115-$120 million. The revenue shortfall during CQ2 was attributed to
push-outs of two key software projects during the quarter. Specifically, management commented
that two of the company's customers put a freeze on their capital investments for one quarter and
two quarters, respectively, which caused the project delays.
Recall that Brooks supplies automation equipment to all of the major front-end semi equipment
companies including Applied Materials, Lam Research, Novellus, etc. While management did
indicate that the revenue miss was solely software related and the company expects software
revenues to be on track again in CQ3, we believe that the preannouncement underscores the
difficult environment in the SPE segment.
To that end, we continue to expect Brooks' OEM business to be weak in H2'05, driven by
deteriorating semi OEM shipments (approximately 50% of Brooks' business is driven by semi OEM
shipments). We would highlight that semi capex budgets appear to be significantly front-half loaded
with semiconductor companies likely spending 60% to 65% of their capex in H1'05. As a result, we
believe semi equipment shipments are very likely to worsen in H2'05.
Following the negative preannouncement, we are reducing our already below-consensus CY2005
EPS estimate to $0.03 from $0.05 (FY'05 goes to $0.17 from $0.19). We would note that the Street
CY2005 EPS estimate is $0.45.
Each of the analysts named below hereby certifies |
