Hi Don,
I'm not surprised by the profit taking. The latest earnings statistics have been superb relative the the past couple of years and the stock price appreciation in the SCE industry tends to verify the data's implication.
Obviously, since this is a cyclical industry, we won't get nice steady 5% sequential growth quarter after quarter but we'll end up with the usual boom/bust trends that seem to be the inherent nature of this sector. One of the reasons I asked Gottfried for the chart comparing AMAT vs Billings or Bookings (as opposed to BtB) is that it appears we are starting at a lower billings point and that may "justify" more stock price appreciation for AMAT in the intermediate future. It would appear, based on the PEG, that AMAT is a very good buy at this stage unless you subscribe to the theory that 2005 will show negative order growth vs 2004.
One difficulty we have in assessing the situation is that we are inundated with statistics which have very little correlation but which are used to justify certain prognostications. For example, so what if the total IC industry grows at 10% going forward vs 15% in the past if capital equipment investment requirements end up growing faster then the whole IC industry? On this thread, we have been pretty good at differentiating between the different types of data. We've also done a decent job of cutting through some of the BS that's associated with the different implications created by the data and the spin that many want to use based on the same implications.
Not to change the subject but here's an article from Compound Semi that I think might interest some of the folks here:
Strategies In Light Roundup
February 8, 2004...The following report was filed Feb. 6th by our Sr. Technology Editor, Dr. Alan Thompson as a news item. It reappears here as an editorial. Strategies Unlimited held their 5th Annual Strategies in Light Conference last week in San Mateo, California USA. The attendance set a new record for this popular event, drawing more than 300 attendees, and the mood was distinctly upbeat. The reason for this is well recognized by those who have been following the fortunes of the HB-LED industry, which was reinforced by Dr. Robert Steele's traditional leadoff talk covering the HB-LED market and forecast. Compared to most of the semiconductor industry, the HB-LED segment did very well in 2003 with a 47% revenue growth (in US$) over 2002 to $2.7B. This rate matches the average CAGR for this segment over the last 8 years. Bob explained that this rate would moderate over the next 5 years as some applications mature and ASPs continue their inexorable decline, but would still reach $6B in 2008. Our industry, which was nibbling at niches 10 years ago, has become a force to be reckoned with! And as major penetration into large applications such as automotive and solid state lighting occur in the 5 to 10 year time frame, the totals will keep climbing. No wonder the conference participants were upbeat! In 2003, about 50% of the revenues came from mobile appliance applications, such as cell phone keypad lighting and small LCD screen backlighting in phones and cameras. Most of these LEDs were GaN based, and these materials increased at a faster clip than the other materials.
The sharp increase in production capacity in Taiwan, Korea and China was addressed by Dr. Robert Walker, who summarized the results of a new report he co-authored for Strategies Unlimited. The study will be made available later this month (see www.strategies-u.com for details). Bob showed that HB-LED producers in these three countries increased their InGaAlP and GaN chip output by 25 and 700% respectively over 2002, and now generate 80 and 40%, respectively, of the worlds total output. He also pointed out that the region is rapidly transitioning from copying to innovating and that we can expect premium products at low prices as the volumes increase further.
Moving on to applications, there were interesting talks by Hella and Schfenacker Research on automobiles. LEDs are penetrating into all interior and exterior positions, but headlights are not yet realizable (except in show cars that are not street legal). The first use of LEDs for daylight running lights in a production car (the new Audi A6) shows progress but differentiates the needs of lights for noticing (CHMSLs, turn signals and daylight running) vs illuminating (headlights). In the latter case the beam must be precisely shaped and having a small source makes this task much easier. Typical halogen filaments and HID arcs are a few square mm in size and emit kilolumens. To reach the same light output with LEDs currently requires many 1mm sq die occupying a much larger source area. This leads to poor cut-off and large (and often expensive) optics. For this application we need higher brightness (more lumens/area) and need to get the heat out. The speakers who addressed this topic agreed that initial introductions should happen in the 2007-8 timeframe in a few luxury cars or SUVs where cost and size are not so important. Other applications covered included trends in LED use in cell phones by J.J. Jung of LG Innotek, a novel large screen video display by New Millenium Media International, and lighting by OptoLum and City Design Group.
The solid state lighting initiatives in the USA and Japan were outlined, with the former still looking for a budget and the latter starting a second 5 year program. A panel session on High Power LEDs saw representatives from Cree, Lumileds, Nichia, Osram, and Toyoda Gosei discussing their latest advances, particularly for white. Progress continues at a good pace and problems with reliability, color balance, and heat extraction are being tackled successfully. Luminous efficacies are generally in the 30-50l/W range for warm white and multi-watt devices are making their way into production. Specific costs were not given... but all agreed that these have to decrease dramatically to move out of the niche applications that are currently well served by LED technology Steve Johnson of LBL gave a status report on OLED technology, which has the potential to displace LEDs in some current and proposed applications, principally small screens in mobile appliances and for backlighting in some applications. We all need to be aware of this technology's strong points, as well as other competitors such as flexible EL lights for cell phone keypad lighting. The conference concluded with a session covering drivers (by Supertex), reliability prediction (RPI) and trends in LED production equipment (Veeco).
Haitz's law: I'd like to finish this report by covering one of the most entertaining and thought provoking presentations, that of industry veteran Dr. Roland Haitz, who spent many years working on LEDs at HP/Agilent/Lumileds. His topic was the next 20 years of SSL. He examined some of the physics limitations on improvements and pointed out that historical improvements (Haitz's law) will probably continue for some time, leading to his statement that "By 2025, LEDs will be the dominant lighting technology". We have a lot to do before then, but the future is definitely bright! --Alan Thompson, Sr. Technology Editor, CompoundSemi News |
