This Upturn Has Legs
Q4 Financial Preview
By Dan Scovel, Needham & Co.
Dan Scovel is an analyst with Needham & Co., based in New York. Scovel tracks the semiconductor market.
We believe the semiconductor industry remains in the midst of a robust upturn that is likely to persist until sometime during 2002 and that the recent weakness in equity share prices has been overdone and opened up a number of attractive valuations on a fundamental basis.
The primary growth-drivers for this cycle are communications markets—all things Internet and cell phones—and consumer electronics including set-top boxes, DVDs, digital cameras, MP3 players and PDAs, among many others. At the same time, we believe we are currently in the midst of a midcycle plateau caused by the correction of overly enthusiastic expectations for cell phone handset and PC unit growth. In the very near term, we believe fourth-quarter business will be boosted by strong growth from consumer electronics and networking/telecom end-markets and supported by a lackluster seasonal expansion from PCs, whose combined effect is likely to offset an inventory correction in the weak cell phone space.
This mixed bag of near- and long-term dynamics has lead to fear and uncertainty in the sector's equity market. The Philadelphia Semiconductor Index Option index fell nearly 50 percent in the three months ending with November to a level in the mid-600s. We agree many equities were priced to perfection at the beginning of this period and a correction was warranted. However, we also think much of this correction has been overdone given persistently strong business conditions and single-digit P/E multiples, in many cases.
So, what happens when fourth-quarter financials are announced during the January earnings season? We believe the bad news has already been more than reflected in current share prices and the sector is likely to rally as investors seek value in a discriminating market environment.
End-Market Strength and Tribulations
Consumer markets account for approximately 15 percent of worldwide chip sales and are one of the two primary growth-drivers of the current business cycle, in our opinion. The semiconductor industry benefits from this segment in the near-term due to peak holiday seasonality, and more generally, as the emerging market nature of these applications tempers seasonal softness over the next couple of years. Set-top boxes, digital cameras, MP3 players, video games, PDAs, and a variety of handheld gadgets are proliferating as expected, and early reports indicate these applications, as a group, may be a big winner during this year's peak holiday buying season.
Communications account for approximately 25 percent of total semiconductor revenues and are expected to be the single fastest-growing market for the industry over the next few years. Networking and telecommunications applications within this segment account for less than 20 percent of total industry sales and are expected to be driven primarily by the build-out of the Internet infrastructure. These markets remain healthy at the current time. However, component inventory building in the networking channel and a potential slowdown in the telecom space due to credit financing of equipment sales to struggling service providers—including staggering CLECs—have surfaced in recent weeks. We would not be surprised to see these segments plateau for a few weeks following heady growth rates in recent quarters. While fourth-quarter business conditions remain healthy, we will be paying close attention to first-quarter 2001 outlooks provided by chip companies with disproportionate exposure to these areas.
Cell phone handsets are one of the fastest-growing subsegments within communications markets and account for less than 10 percent of total chip industry sales. The big picture is one of persistently high growth as worldwide unit shipments of 280 million in 1999 are followed by more than 400 million this year and 500 million next year. However, near-term the space is suffering from a correction of overly optimistic expectations and a slight build-up of component inventory in the channel. We believe this situation will correct itself with the coming of the new year.
Cell phone unit shipments this year were originally forecast in a range around 435 million units, but expectations inflated to 500 million, after a strong first quarter posted shipments of 125 million. However, chip industry component limitations—primarily of flash memories—were never going to allow cell phone units to exceed 435 million. The first disappointment to equity investors was the realization that 500 million was unattainable. The second fear resulted from uncertainty associated with market-share shifts among major players. The net impact is a slowdown as lagging handset makers burn-off accumulated inventories while adjusting for slower-than-expected growth rates. However, we expect the impact of this correction is limited to three to four months around fourth quarter—and uneven by product type and end-customer exposure. For example, we do not expect flash memory suppliers to be adversely impacted due to significant industry-capacity constraints. Nor do we expect chip suppliers to Nokia, Samsung and Siemens to experience any meaningful slowdown, given recent trends of market-share gains. On the other hand, suppliers to Motorola and Ericsson are more likely to feel the pain associated with a correction. Nevertheless, we expect the combined effects of inventory depletion, continuing industry expansion, and correcting expectations to result in a significantly more favorable business and investing environment with the arrival of the new year.
The days of semiconductor growth being driven by PC markets have passed and PC expansion now lags that of its peers. However, the chip industry remains dependent on PCs for its overall health: PCs and peripherals are the largest end-market for semiconductors, accounting for approximately 40 percent of total revenues. We believe price erosion will largely offset PC unit growth in the mid-teens percent, leading to tepid revenue growth for the foreseeable future. The chip industry needs PCs to not do poorly, as a minimum, to ensure industry-wide growth in the aggregate.
PC market growth expectations have been relatively modest this year, but inflated after a stronger-than-seasonally-normal second quarter—which is typically the weakest period of the year. Through the benefit of hindsight, we believe the strong first half 2000 resulted from a conscious effort by PC makers to accumulate buffer stocks to support peak-season sales at year-end, after component constraints limited their ability to enjoy such fruits last year at this time. In our view, this explains the persistently modest shortage of microprocessors and relatively robust market prices for DRAMs throughout much of the year. However, both of these product areas are now experiencing a growth hangover. We believe PC markets will remain solid, albeit lackluster, and not adversely impact chip industry growth.
Product Area Dynamics
Flash and other nonvolatile memories remain in extremely short supply as do SRAMs for communications applications. We expect such conditions to persist throughout 2001 due to limited suppliers and increasing content per application—despite fits and starts from various end-markets. Flash memories are the semiconductor industry's fastest-growing product area at more than 150 percent through September 2000. Major flash suppliers include Intel, Advanced Micro Devices Inc. (AMD), Atmel and STMicroelectronics.
PLDs are the second fastest-growing product segment at more than 80 percent and benefit from disproportionate exposure to networking and telecommunications markets. However, PLD makers recently announced slower near-term growth as contract manufacturers work off accumulated inventory. Pure-play suppliers include Xilinx, Altera, Lattice and Actel.
Total DRAM revenues are up more than 60 percent so far in 2000, but remain a wildcard for the fourth quarter, in our opinion, as market prices crashed by half during the first half of the period and now appear to be stabilizing and may recover. DRAMs have disproportionate exposure to PC and peripheral markets at approximately two-thirds of sales, but also have their own cycle of boom and bust with industry under- and oversupply. At 14 percent of total chip industry revenues, these products may single-handedly trigger industry-wide weakness as we exit the year. Major suppliers include Samsung, Hyundai, Micron Technology, Elipida (NEC and Hitachi) and Infineon.
Microprocessors are also largely dependent of PC markets. The product group is underperforming to the industry average this year and is forecast to remain that way for the foreseeable future. We believe Intel is unlikely to escape the vagaries of the product sector or PC markets—despite urgent diversification into communications. On the other hand, we are bullish on AMD given its attractive competitive offerings and inferior market-share position.
Investment Opportunity Awaits
We believe that virtually all the bad news and uncertainty in the semiconductor sector over the next year to year-and-a-half has already been factored into current share prices. In our view, a correction was warranted in early September, given high fundamental valuations based on longer-term historical averages. However, we now believe the correction has been overdone: We are in the midst of a strong cyclical upturn and many shares are trading at levels anticipating an imminent downturn. We view this as an opportunity for investors. We believe fourth-quarter earnings announcements during January 2001 will largely exceed corrected expectations and are likely to spark a rally in the sector as investors seeking fundamental value realize that the industry upturn remains intact and that the recent share price correction was overdone.
electronicnews.com |
