INTEL , Reducing Unit and ASP Assumptions, Price Target $24
Author: Scott Randall F01E: $0.49 F02E: $0.55 Price: $21.96
We are reducing our estimates for Intel based on lower unit and ASP assumptions.
With Intel having reconfirmed its guidance on Sept. 7, and with its cancellation window of just a couple weeks, we believe that the September quarter could come in close to our estimates, while the risk to 4Q and 2002 numbers is greater.
Our unit estimates for Intel had been for sequential growth of 5.6% in September and 7.4% in December. While both of these sequential growth estimates are less than the levels typically seen, increasing evidence of continued weak PC sell-through is causing us to further reduce our estimates. Our new estimates call for quarter-over-quarter unit growth of 4.8% and 4.6%.
Our ASP assumptions have also been reduced for both 3Q and 4Q from $169 and $166 to $165 and $162, respectively. Although field contacts suggest that acceptance of P4-based systems continues to accelerate, we believe that the pricing environment is more difficult than we had modeled.
In addition, we believe that Intel's communications business will also incrementally suffer from a somewhat slower recovery in this sector than we have modeled.
From a valuation perspective, Intel continues to trade at a premium to former trough levels on most metrics. Off of our lower estimates, we are reducing our price target from $33 to $24. We are maintaining our buy rating, although PC components continue to be our least favorite sector.
We believe that better times for Intel will require greater traction both from XP as well as high-speed, cost reduced P4s - events that won't really get rolling until sometime in 2002 at the earliest.
We Are Reducing Our Estimates for Intel-based on Lower Unit, ASP and GM Assumptions
With PC unit and market demand continuing to be weak, and with both AMD and Intel continuing to be aggressive in defending market share at almost any cost, we believe that some of Intel's expected market share gains are not materializing as quickly as we had expected.
Our Unit Estimates for Intel Had Been for Sequential Growth of 5.6% in September and 7.4% in December
While both of these sequential growth estimates are less than the levels historically seen, increasing evidence of continued weak PC sell-through is causing us to further reduce our estimates. Our new unit assumptions are for 4.8% and 4.6% quarter-over-quarter growth. Taken in total for all of calendar 2001, our unit estimates had been for a decline of 7%, while our new estimates call for a decline of 8%.
In Addition, We Have Reduced Our ASP Assumptions for the September Quarter From $169 to $165 and December's Assumptions Have Moved From $166 to $162
In combination with lower unit and ASP assumptions, we have also reduced our GM assumptions for F02 from 50.8% to 48.5%. Although we believe that INTC's migration to 0.13-micron technology and a wider speed gap between itself and AMD could help both its cost basis and higher-end ASPs, reduced market demand could serve to keep greater pressure on GMs than we have been modeling.
As We Have Related Previously, Conversations With Field Contacts Over the Last Six Months or So Have Suggested That Intel's Aggressive Pricing Actions on the P4 Have Made This Device More Compelling
However, from a system perspective, Intel's choice of memory technology we believe has somewhat held back the P4 penetration from what otherwise might have been. Recall that Intel's initial choice of memory technology was Rambus RDRAM. Although this choice of memory offers good performance, the pricing premium that this product has commanded has negatively impacted system build costs. To address lower system build points, Intel has now introduced the 845 chipset that supports SDRAM. Unfortunately, while this memory technology does result in lower build costs, the performance of this memory sub-system is less than that which is seen for RDRAM.
Going Forward, Intel's Launch of DDR Memory Support With the 845, Currently Scheduled To Occur in Early 2002, Will, We Believe, Offer a More Compelling System Solution Than Either RDRAM or SDRAM
Simply put, DDR will not have the pricing impact of RDRAM, while system performance should be comparable to that which is seen when using RDRAM depending upon the application.
In Addition, We Believe That Intel's Communication's Business Will Also Incrementally Suffer From a Somewhat Slower Recovery in This Sector Than We Have Modeled
Recall that Intel sells products into a variety of communications applications. Included are chips targeted at both wireline and wireless applications. In wireline, Intel sells a variety of both device and system level products. In devices, Intel continues to participate in the legacy business from its acquisition of Level One, as well as targeting additional applications in areas such as network processors. Although we believe that Intel is having some success in this area, we believe that it has been more difficult than the company would have expected. In part, Intel's continued participation in system level communications products results in some level of customer reluctance to fully embrace Intel as a communications semiconductor partner.
In Wireless, Intel's Current Participation Is Driven Primarily Through Flash Memory
Going forward, Intel is targeting additional exposure into areas such as handsets through the company's efforts with X scale and its PCA architecture. Although we believe that Intel will ultimately have a compelling solution for this area, we believe that customers may show some reluctance to fully embrace Intel as a strategic solutions provider given the fear that Intel may at some point possibly become a competitor.
SoundView research provided through these pages has been prepared by SoundView Corporation, a broker-dealer that is not affiliated with E*TRADE. Please review the important information about SoundView's research reports. |
