Semiconductor Equipment . . . According to Fulcrum, Novellus's mid-quarter update appears to have been in line with consensus expectations in every metric. Nevertheless, firm believes that management's conservative tone during the call as well as its expressed expectations for a "muted recovery" could weigh on NVLS and other equipment stocks. Given firm's longer-term outlook for the group, Fulcrum reiterates its Buy rating on NVLS despite possible profit taking in NVLS and others in the space. Firm sees downside risk in NVLS in the $36-$37 range and would recommend aggressive buying in the stock price should it reach these levels. Fulcrum is raising its long-term, 12-month target from $40 to $50, which represents a P/E multiple of 31x estimated EPS for the 12-month period ending June 2005.
While tone of the mid-quarter conference call was lukewarm, ThinkEquity believes several end market indicators have turned positive in recent months and that Novellus should be one of the beneficiaries. However, while NVLS offers an attractive risk to reward profile, believes Applied Materials offers better near term returns. The firm notes that Taiwan is showing some signs of a pick up but not yet strong enough to benefit Novellus.
Asyst upgraded to Neutral at JP Morgan to Neutral from Underweight. The upgrade is based on their assumption that photolithography and automation equipment suppliers will likely experience higher than group average sequential bookings growth in 3rd quarter-4th quarter, which should drive superior stock multiple expansion. While firm believes ASYT is ripe for new money, Brooka Automation remains their top pick in automation due to its margin profile.
Novellus maintained its net order guidance of $220 million during its mid-quarter update, along with revenues of $215 million-$220 million and EPS of breakeven, in line with our preview. While the company is positive on order trends, it was a bit cautious on the rate of growth in the near term. Novellus is not a major supplier to Powerchip, which could be very active placing orders in 3rd quarter 2003/4th quarter 2003. Also, the company noted it was lowering its operating expenses through several measures (details to follow after the end of the September quarter), which could include headcount reductions, streamlining of R&D expenses, lower cycle time of components, manufacturing efficiencies and others, and are expected to result in increased operating leverage going forward. It is targeting gross margins in the mid-50% range on revenues in the $320 million range. Estimate cost savings could add $0.10-$0.20 per share in 2004. NVLS is also taking a charge of about $70 million in the quarter for inventory writeoffs of 200mm systems as its 300mm systems are ramping faster than expected as well as $5M in charges for the writeoff of its joint development project on advanced clean operations.
Semiconductors . . . Intersil target raised to $34 at Piper Jaffray. Yesterday, company completed the sale of its wireless networking business to GlobespanVirata for $250 million in cash and roughly $114 million in GSPN stock. The company's new price target is based on 40x the firm's 2004 EPS estimate of $0.86 as well as its DCF analysis.
Goldman Sachs reits Outperform on ATI Tech. Checks suggest that the company may be gaining market share in desktops driven by new add−in board relationships in discrete graphics and traction with the new Radeon 9100 in integrated graphics; also, firm thinks ATYT is attractively valued versus the group, at 2.0x estimated CY04 EV/sales vs. the group average of 4.9x.
PMC-Sierra estimates raised at CIBC. 3rd quarter revenue estimates go to $63.4 million from $61.0 million and the firm thinks co can break-even in 3rd quarter. Management has not changed their guidance, but CIBC feels the "flattish" guidance given on July 18th was too conservative. Firm calls PMCS a core holding in the semi sector although on an EV to 2003E and 2004E sales basis, the stock is not cheap at 8.7x and 7.1x.
ChipPAC estimates cut at Merrill Lynch 2003-04 EPS estimates below consensus. The firm believes the factors leading to muted 3rd quarter growth have been more severe than originally estimated; issues include greater than normal price erosion due to aggressive competition at specific accounts, temporary mix shift to lower margin computing packages, and inventory correction in the higher margin communications segment following very strong growth in 2nd quarter. Firm maintains Buy rating, but cuts their target to $8.50 from $10.
Smith Barney initiates 2005 EPS estimates generally above consensus for Texas Instruments, Analog Devices, International Rectifier, and Cypress. The firm also raises their targets for TXN (to $25 from $17), ADI (to $44 from $35), IRF (to $44 from $27), and CY (to $21 from $16). In addition, firm says Intel remains their favorite chip stock going into the end of the year due to EPS upside potential through further margin leverage, and says Broadcom, LSI, Cypress, Fairchild, and Integrated Device appear to be among the better stocks offering good near-term growth prospects and/or more palatable valuation than most chip stocks.
Intel has the kind of business model that analysts look for in a company. First, the company has been consistently profitable. Intel has been profitable in each quarter over the past 10 years, most of the time with operating margins that exceeded 30%. Second, the PC microprocessor (MPU) market has evolved to a point where one competitor dominates the market. Intel has about 90% of the $23 billion PC MPU market. The other 10% of the market is held by AMD, Via, and Transmeta.
This dominance has enabled Intel to create barriers to entry in its markets. Intel’s size means that no other company can come close to it in terms of investing in development of new technologies. For example, Intel’s research and development (R&D) budget in 2002 was $4 billion, which was four times the R&D budget of its three largest competitors combined. The company’s size and market share also give it economies of scale that enable it to have a manufacturing advantage. For example, with 10% of the PC MPU market left for its competitors, few can afford the $2-$3 billion price tag that comes with building a new semiconductor fabrication (fab) facility. This, of course, creates a vicious cycle—only the highly profitable companies can afford to buy fabs that will make them more profitable. Intel’s successful execution has resulted in an industry that is largely dependent on it. Peripheral chip suppliers and even PC original equipment manufacturers (OEMs) themselves depend on innovations and specifications from Intel to help generate product and replacement cycles. Indeed, Intel, along with its competitors, sells into systems that are referred to as “Intel architecture” PCs.
Most investors know Intel as a supplier of PC MPUs. Not as many understand that it also sells a number of other chips that go into the PC, such as the memory controller hub and I/O controller hub (also known as the PC chipset or northbridge and southbridge chips), graphics processor, and increasingly interface chips such as Ethernet and wireless LAN (WLAN). While Intel has been selling a number of the chips around the MPU for a while, it recently has been more aggressive in trying to bundle them together. For example, Intel recently launched its “Centrino” branded mobile platform, which bundles its “Pentium M” mobile processor with its chipset and wireless LAN solution. PC OEMs that sell mobile PCs with Intel’s Centrino platform and use the Centrino brand name get access to a marketing budget from Intel. Intel’s marketing budget for Centrino is $300 million. While PC OEMs may have used Intel’s MPU and another vendor’s chipset or WLAN solution, bundling gives Intel a higher total revenue per unit.
The logical extension of bundling is to integrate separate components onto a single chip—not surprisingly, Intel has been successful in doing that also. For example, Intel has integrated the GPU into the memory controller hub, and has become the largest supplier of graphics processors. Longer term, we would expect Intel to integrate other PC components such as Ethernet and WLAN into the I/O controller hub, effectively capturing more of the silicon value in the PC.
While Intel’s long-term strategy appears sound, near-term stock performance for technology stocks is driven by product and market cycles. In Intel’s case, expect it to benefit from four over the next 12-18 months:
* Higher value notebook PCs cannibalizing desktop PCs. Mobile MPUs have a 15% premium to Desktop MPUs.
* A new high-end server MPU that expands the company’s addressable market. Intel has been highly successful with its Itanium II product addressing the server market.
* The Asian market just starting to get penetrated. Asia/Pacific accounts for over 50% of the population and is roughly 2% penetrated by PCs.
* An ongoing replacement cycle on an aging PC installed base. Intel estimates that a total of 150 million PCs are over three years old.
The valuation method used to determine a $34 price target is based on a 34 times multiple of 2004 EPS estimate, which translates to a 20% premium to the pre-bubble P/E multiple of 28 times. This is warranted given the low-interest-rate environment and prospect for improving fundamentals. These results are consistent with our discounted cash flow (DCF) analysis, which suggests a share price of $30-$34 for a weighted-average cost of capital (WACC) of 10% and a terminal growth rate of 3%-4%. For our analysis, we assumed a top-line CAGR of 8% for the company between 2003 and 2008. We believe these growth assumptions are reasonable given the growth prospects in the PC market and that Intel has a monopoly on MPUs sold into PCs.
As emerging markets for PCs grow, expect the lower-priced “white box” PCs that are sold into those markets to also grow. White box computers represented 45% of total PC shipments in 2002. If white box PC penetration accelerated, it could potentially result in lower ASP and margin assumptions for Intel.
S&P 500 earnings are another key indicator of corporate IT spending. A positive investment thesis on Intel is based in part on positive growth in corporate earnings. If that growth fails to materialize, then our revenue and earnings forecasts would be at risk.
According to Clay Christensen’s disruptive technology theory, existing technology is at risk of being replaced by lower-end technology when improvements in the existing technology exceed the ability of users to absorb those improvements. Lengthening PC replacement cycles suggest that the PC market is at risk of being disrupted by lower-end technologies.
INTC is trading at over a 20%-100% premium to its average historical valuation metrics. The premium is justified due to the low interest rates and the timing of the semiconductor cycle. There is a risk that the valuation multiples could contract by over 50% if interest rates rise or the semiconductor sector’s fundamentals deteriorate.
RobBlack.com MarketWrap
robblack.com
Thanks Michael!
RtS |
