CompoundSemi Online - The Original Compound Semiconductor Industry Newspaper

February 11, 2005...The following report has been filed by our Senior Technology Editor, Dr. Alan Thompson.

Once again the folks at Strategies Unlimited are to be congratulated for putting on an excellent LED conference. Attendance exceeded expectations and beat last year’s record handily (in spite of coinciding with the Chinese New Year). The program was well thought out, speaker quality was high, and the move to a new location gave more room for all the functions

My main impression is that business is very good on the whole with high power LEDs in particular penetrating new applications and enjoying high growth rates. However, a closer look at all the numbers reveals there may be some storm clouds looming on the horizon. I invite you to read on for my very personal (and inevitably biased) view on the story unfolding at Strategies In Light 2005

Bob Steele’s traditional kick-off looked at the HB-LED markets and forecast. He reported a 37% revenue gain for the overall market for 2004 over 2003 to $3.7 billion. While this percentage is skewed by the weak US dollar, it still represents about a 32% gain for the year. This very healthy increase is reasonably in line with the long-term CAGR of 46%. Bob did point out that as the overall market grows and matures he is expecting a decline in the growth rate to about 14%, resulting in a total market of $7 billion in 2009 (over 10X growth in a decade!). Some interesting points to emerge from his very detailed analysis are that mobile appliances grew to 58% of the total HB-LED market (up from 28% just 3 years earlier), that GaN based LEDs captured more than 80% (but "only" 62% by units), and that signs and automobile applications were tied for second at 13% each by revenues

You will find a more detailed description about individual presentations in the main body of this newsletter over the weeks ahead, by our incoming editor, Scott McMahan, who also attended SIL, so I will move on to what I see as those gathering storm clouds. First some facts and opinions based on industry trends and conversations with speakers and vendors at the conference:

The cell phone market grew unexpectedly to 670 million units in "04 (most people were expecting little or no growth at the 525 million level at this time last year). This drove a large fraction of the LED market growth in ’04. Is this sustainable or a one-time pop?

The overcapacity in Taiwan resulted in a doubling of units with flat revenues (significant price erosion), and a sudden slowdown in capital equipment sales

Korea and China have also added significant capacity recently and are aggressively pursuing market share

The rapidly growing markets for HB-LEDs are mostly a small percentage of the total at present

I have been in the semiconductor industry for four decades and have experienced many of the infamous cycles. When business is good everyone is happy but trying to gain market share. This leads to shortages, double ordering, and frantic attempts to add capacity (which takes a long time). Eventually the supply catches up and passes the true demand, resulting in cancellation of the double orders and an overcapacity situation. Then the suppliers have to lower prices drastically just to maintain volumes out of their expensive fabs, and profits quickly turn to losses. Some of the weaker companies go under, the stronger ones survive, and the cycle starts all over again. The main lesson is to question any sudden increase in a market – recent examples include the telecom market in 2000-01 – is it real or are too many companies chasing (and expecting to get) the same piece of business?

When we look at the HB-LED market, we definitely have overcapacity. As such, ASPs have fallen faster than would be expected from increasing yields and efficiencies alone; Cree has reported unexpectedly flat sales; Taiwan is hurting, and epi tool sales have plummeted. Do we, as one colleague remarked, have 12 companies each expecting to get 25% of the total business? And more important, how is our major end market doing? What if cell phone sales fall to 600 million units or even 500 million units in ’05? Unfortunately, the emergence of 3G is not going to happen nearly quickly enough to significantly impact the market volume for at least two years, in my opinion. If the cell phone market stutters, you will see tremendous pricing pressure and the revenues will decrease substantially. The other applications for HB-LEDs are growing nicely but do not have a big enough piece of the pie to overcome such a drop. This could lead to a flatter ’05 (later in the year) and a real decrease in revenues for ’06

Does all of this sound like a "perfect storm" to you? Or am I being waaaay to pessimistic? I’m sure we are all going to be watching the end market for cell phones carefully, since this has such a huge impact on our current HB-LED market, while pursuing all the other business that is definitely out there

Disclaimer: This is my opinion only and does not reflect the thinking of CompoundSemi Online, Strategies Unlimited, or anyone else mentioned in our reports. Comments are definitely welcome, because through dialog we can all understand the opportunities and pitfalls better -- Alan Thompson, Sr. Technology Editor
(You can contact Alan directly at alant@compoundsemi.com).

Turmoil in Veeco's TurboDisc Division Over Internal Accounting
Jo Ann McDonald

Kopin's Unnamed Partner in Blue Spectrum LED JV in China is Bright LED

Cree and Nichia Agree to New Cross License
Jo Ann McDonald

Kopin's Blue Spectrum LEDs Find a Home in China
Jo Ann McDonald

Bob Steele Kicks Off SIL with Forecast of $7 Billion Market by 2009

Oriol White LED IP Auction Brings $1.5 Million

Emcore Rescues the "Stranded" by Extending the Reach of XENPAK

RFMD Opens Design Center in France & Introduces PowerStar PA Modules

Alan Thompson Reports on Strategies in Light 2005

Turmoil in Veeco's TurboDisc Division Over Internal AccountingJo Ann McDonald