Semiconductor Equipment . . . Veeco Instruments reported earnings before items of 3 cents per share, ahead of its own outlook for between a breakeven performance and a profit of 2 cents per share. 11 analysts were looking for breakeven results. Including items, the firmposted a loss of $1.7 million, or 6 cents per share, in the first quarter. Sales fell 18 percent in the latest three months to $65.8 million from $80.1 million in the same period a year earlier. Bookings totaled $72.7 million for the quarter, a 4 percent increase from last year. Looking ahead, Veeco forecast earnings before items of 1 to 4 cents per share for the second quarter on sales of $64 million to $67 million.
Applied Materials was cut to Underperform at First Albany. The firm is saying they believe there is a strategic divergence between the company and its customers, as many of AMAT's growth initiatives seem to detrimental to its customers. The firm believes that the company's market share in the overall front-end equipment sector has peaked and is poised to decline, and expects few of the company's strategic alternatives to grow revenues will pan out.
Semiconductors . . . Global semiconductor sales for the first quarter increased 13 percent sequentially, but declined 3.2 percent from year-earlier levels to $36.4 billion, according to the Semiconductor Industry Association. For the month of March, chip sales totaled $12.1 billion, up 2.6 percent from February, the industry tracker said. SIA added that it continues to expect double-digit percentage increases in chip sales for 2003, boosted by consumer PC purchases, a corporate PC upgrade cycle and strength in the wireless sector.
SSB downgraded the chip industry on Friday to Underweight from Marketweight for four reasons: 1) a rise in customers' DOI, which is tangible evidence that chip shipments in 1st quarter were in excess of actual demand, 2) signs of deterioration in end-market demand -- particularly in Asia, 3) valuations cause us to believe that there is more downside than upside at these prices, and 4) the summertime is usually not a great time to own chip stocks.
There was very little pushback against the downgrade, which probably signals that -- although we do not know of any other industry downgrades lately -- the rating change is pretty close to conventional wisdom on the buy side.
Sometimes the conventional wisdom is right, but investors that agree with downgrade should be careful in not being complacent about their (or our) worldview since it may represent greater groupthink than any of us would like.
That said, the fundamental groundwork for the downgrade unveiled a fairly sudden, measurable change in the 'demand pull' upon the supply chain in Taiwan and China. This note contains a summary of some of the observations from our Asian research team -- whose work was a major factor in our downgrade. Much of the recent change appears to be attributable to SARS.
Unfortunately, SARS is probably a lose/lose situation for chips in the near- term. If draconian steps are taken to slow the spread of the disease (as is increasingly the case) such as closing shopping areas, consumer demand in Asia will be negatively impacted. If draconian measures are not taken, the ensuing uncertainty would likely cause consumers in Asia to 'shut-in' anyway.
It should be acknowledged that the concerns of chip suppliers (and analysts) are almost petty in crisis situations such this. To avoid analyzing events from a distance, we are leaning heavily upon the direct reports from our research teams in Asia.
According to data released by the Semiconductor Industry Association yesterday, Semiconductor dollar shipment growth fell to +13% year over year in March from +18% year over year in February. This is the third consecutive decline since peaking in December at +23% year over year. All geographies declined on a year over year basis with the worst performer being the Americas (-8% year over year) followed by Europe (+11% year over year), Asia-Pac (+17% year over year) and finally Japan (+33% year over year).
Unit growth also decelerated to +14% year over year from +21% year over year in February. The silver lining in this report is that pricing is now flat year over year compared to -3% year over year in February. Expect pricing to be about flat in 2003 after sharp declines of 14% and 12% year over year in 2001 and 2002 respectively.
On a month-over-month basis, shipments increased 3% mom after slipping 3% mom in February as Japan (+5% month over month), Asia-Pac (+3% month over month), and Europe (+3% month over month) had positive monthly gains. Only the Americas (flat month over month) did not increase its semiconductor shipments in March after declining 5% in February. Continue to expect shipments and units will weaken through the first half due to tougher year over year comparisons, and as most end-markets remain soft. Units were up 4% month over month in March after declining 2% month over month in February. Pricing was down 1% month over month for the third consecutive month.
On a relative basis, the steadiest segments within the semiconductor industry are DSPs and standard linear, and Flash memory to a lesser degree. Other segment details include:
* Discretes declined to +15% year over year in March from +23% year over year in February, with a 3% month over month increase in March shipments. Discrete pricing has been relatively stable for much of the past year after deep declines in much of 2000 and 2001. Competitors under coverage in this segment include Fairchild Semi and International Rectifier.
* Analog overall declined to +16% year over year in March from +21% year over year in February, with only a 1% mom increase in March shipments. Pricing has improved to -3% year over year while units have decelerated to +16% year over year. Unit performance in analog continues to be strong. The standard linear portion of the analog market was up 21% year over year, within the 20-23% range of the last several months. Competitors under coverage in the analog segment include Analog Devices, Fairchild Semi, International Rectifier, Linear Tech, Maxim, National Semi, and Texas Instruments.
* Microprocessor (MPU) shipments were -3% year over year in March, flat from February levels, following a strong 6% mom increase in March shipments. Competitors under coverage in this segment include Advanced Micro Devices and Intel Corporation.
* Digital Signal Processor (DSP) shipments increased to +48% year over year in March from +45% year over year in February, with a 2% mom increase in March shipments. DSP pricing is down about 5% month over month and 6% year over year, though units are up a strong 58% year over year. Believe a cyclical deceleration in DSP units is due later this year given the segments very high growth rate and tougher comparisons beginning in the summer months. Competitors under coverage in this segment include Analog Devices and Texas Instruments, and to a lesser extent Agere, and National Semi.
* Programmable Logic Device (PLD) shipments increased to +28% year over year in March from +27% year over year in February, with the SIA reporting a +19% mom increase in March shipments. PLD units are up 31% year over year, in-line with shipments, and pricing remains stable at -1% quarter over quarter and -2% year over year. Competitors under coverage in this segment include Altera and Xilinx. Note that PLD data is highly volatile on a month-by-month basis, reducing the utility of the data.
* Dynamic Random Access Memory (DRAM) shipments declined to -11% year over year in March from +11% year over year in February, with a -5% month over month decline in March shipments due to weak ASPs. Pricing remains weak in the DRAM segment with ASPs -54% year over year and -7% month over month. ASPs remain firmly below fully loaded production costs causing many DRAM competitors to begin producing NAND and NOR flash production in order to stem operating losses. DRAM bit growth increased to +49% year over year in March from +28% year over year in February, which supports the argument that PCs are including greater bit content per box. DRAM bit growth is now averaging 50% per year, down from its long-term growth rate of 70% since 1985. The lone competitor under coverage in this segment is Micron Technology.
* Flash memory shipments declined to +54% in March from +59% in February, with a 2% month over month increase in March shipments. Price per bit also remains weak with March pricing down 28% year over year and 3% month over month. Flash bit growth is strong at +118% year over year and +5% month over month. As more and more competitors enter Flash production, excess capacity and weak pricing will hamper this segment. Competitors under coverage in this segment include Advanced Micro Devices and Intel Corporation.
The analog chip market is arguably the best segment in the chip industry due to its diversification and relative pricing power. It is also among the most eclectic group of chip companies we can think of -- which makes generalizations particularly dangerous in this area.
One of the issues that complicate the analysis of the group is how to cleanly categorize sub-segments since definitions are fairly subjective. Each company takes a different approach in describing its business, with particular vagueness about the line dividing standard versus application specific analog/mixed signal products. Several companies resist discussing specifics for valid competitive reasons, so it is difficult verify data with any great certainty.
In order to employ presumably impartial third-party data, we are using recent reports from the Semiconductor Industry Association (SIA) and industry research firm Gartner Group for the baseline of our analysis. 2003 forecasts are based upon the individual models for the companies and presume that the mix of standard analog products does not change as a percentage of sales.
It is also appropriate to caution that it is a little artificial using calendar year-ends as strict starting points and ending points for market share analyses. Of course, we have to pick a particular period in which to measure, but problems are created when particular analog sub-segments of the market decelerate or reaccelerate sooner than another sub-segment.
For example, in the downturn of 2000 and 2001, the amplifier sub-segment bottomed about a full year after the voltage regulator market. Thus, Analog Devices (high exposure to amplifiers) had market share gains in 2001 as the amplifier segment stayed strong, compared to a company like National Semi (higher exposure to voltage regulators) which showed market share gains in 2002 as the voltage regulator segment recovered earlier than amplifiers.
The overarching theme for the analog market in 2002 was that the high performance (and higher ASP) companies lost overall analog market share while manufacturers of more commodity-oriented power management analog and lower- end standard analog gained overall analog share.
Generally, companies like Linear Technology, Maxim, and Analog Devices had higher exposure to networking and telecom infrastructure, which continued to deeply correct in 2002. Products sold into networking and telecom infrastructure are generally high ASP sales. As such, the high-end analog players lost a disproportionate amount of business relative to the analog market overall.
Because of the highly fragmented nature of the analog space, the shifts in market share do not necessarily represent the results of head-to-head competition. In other words, the results are at least partly the result of the particular mix of niches that each company served.
As a case in point, SIA data confirms that the product segments that companies like ADI and Maxim have the greatest exposure had lower growth compared to the lower-end power management markets. Data converters (-18% year over year, where ADI, Texas Instruments and Maxim have the greatest presence) and amplifiers (-14% year over year) lost share in 2002. While voltage regulators and references (+22% year over year, important for Linear Tech and National) grew fastest.
By company, the winners in 2002 include:
* Fairchild Semi increased its market share to 4.2% in 2002 from 3.7% in 2001, primarily driven by the company's power management expertise. For 2003, expect Fairchild's market share to decline a bit to 4.0% as the company is growing more slowly than the overall analog market this year, by our estimates. Fairchild most likely took a bit of share from high-end competitors in 2002.
* International Rectifier has been steadily increasing analog market share for the past several years, with its 2002 share up 50bps to 1.7% of the market. Like its primary competitor Fairchild, International Rectifier has a solid power management franchise, though the company seems to be transitioning to proprietary analog products faster than Fairchild (albeit from a lower base). As such, expect the company's Analog market share to hold flat at about 1.7% in 2003.
* National Semiconductor's market share increased 40bps to 8.2% of the analog market after a steep decline in 2001. For 2003, expect National's market share to tick back down to 7.8% as the high-performance analog group takes back some of the share it lost in 2002, and the power management side remains healthy.
* Cypress increased its market share from 1.3% in 2001 to 1.7% in 2002. This is mainly due to the company's competitive positioning in USB, frequency timing generators (FTGs) and other mixed signal ICs. For 2003 expect the company's share of analog to downtick to 1.6% as the company has slightly slower overall growth prospects than the market in total. This may prove conservative if Cypress regains market share in the FTG market.
By company, the market share losers in 2002 include:
* Texas Instruments' market share declined to 17.6% in 2002 from 18.9% in 2001, primarily due to increased competition in analog baseband chipsets, and due to a shrinking market for high speed analog interface, in which the company has a large portion of that market. For 2003 expect the company to gain back about 30-bps of share to 17.9% of the analog market, primarily due to a richer mix of products sold to higher ASP GPRS cellular handsets.
* Analog Devices lost 130bps of share in 2002 to 10.9% primarily due to its high exposure to cellular handset and telecom infrastructure end-markets. The amplifier and converter markets, both weak in 2002, are expected to recover some going forward. As such, for 2003, expect the company's share to increase back up to 11.6% due to a partial recovery in demand for those products. As mentioned before, the amplifier market did not bottom until Q1 2002, and thus is expected to recover some in 2003.
* Linear Technology lost a good amount of total analog share in 2002, declining to 5.5% from 7.1% in 2001. For 2003, expect Linear's share to rebound 40bps to 5.9%, largely linked to the company's exposure to the power management area. However, cannot rule out the possibility that Linear is losing share in its regulator business to lower-end competitors like Fairchild Semi or National Semiconductor.
* Maxim also lost a fair amount of total analog share in 2002, declining to 8.3% of the analog market in 2002 from 9.5% of the market in 2001. For 2003 expect Maxim's share to essentially stay flat at 8.3% of the market, at least in part because the converter market (which is estimate is 20-25% of Maxim's sales) should do a little better this year.
2020insight.com
Lets assume you are right. BPNDX was up 3 today to 68. We are only 7 buy signals away from 75. Another 14 buy signals from the 82 reading seen at the previous top in December.
investorshub.com
Would you short something in this market if the SOX reached 393 again?
RtS |
