Sustainability -- Energy & Power Technologies
The Second Cycle
Worldwide LED supply/demand analysis
and technology primer
• Our base case supply/demand scenario highlights that LED manufacturing capacity is
quickly heading for a significant shortage around 2010. We conclude that meeting
the projected demand will require between 300 and 650 high-capacity MOCVD
reactors worth $650 million to $1.4 billion. As such, we see the current cycle as a
bullish indicator for both of the incumbent equipment suppliers.
• Our prediction of a capacity shortage is driven by a 20% CAGR forecast in total LED
demand through 2012, resulting in projected revenue ranging from $15.8 billion to
$24 billion. The largest drivers of this demand forecast are notebook and LCD TV
backlighting, which we expect to go to 95-100% penetration and to 20%-50%
penetration by 2012, respectively. We predict general lighting will represent the next
wave of LED growth, at only 10%-13% of total LED revenues in 2012.
• Our analysis concludes yield losses at the reactor and packaging levels are well below
optimal, suggesting further vertical integration between epi, chip, and package
companies in an attempt to bring up yields, as well as a renewed focus on
manufacturing techniques in an effort to drive down costs.
• We see the Chinese market as very large, yet almost impossible to quantify in terms
of actual demand. China presents a critical sales channel for many non-Chinese LED
suppliers as an outlet for the large quantities of “off spec” die from the poor yields
mentioned above. This market is likely to be increasingly served by domestic Chinese
manufacturers as they climb the learning curve in LED capacity and manufacturing
prowess, placing greater emphasis on distribution schemes and pressuring margins.
• New entrants such as Samsung and LG Electronics should bring greater scale and
manufacturing expertise from traditional semiconductor markets. However, it is too
soon to be certain how successful and what yields these companies will achieve, and
what the implications are for the industry‘s supply/demand balance.
FOREWORD
It has been over three decades since we saw the first commercial application for LEDs:
remote controls for televisions. Since this first introduction, LEDs have become
ubiquitous across numerous consumer, automotive, and commercial applications. We
attribute this adoption curve and market acceptance of LEDs to the spectral efficiency
that differentiates LEDs from conventional incandescent and florescent-based lighting
technologies. As the underlying material sciences have improved over the years, so has
the efficacy of the LEDs, creating a more compelling value proposition to displace
incumbent lighting technologies in many markets.
As LEDs prepare to enter some of the higher-volume mainstream markets and transform
the way in which we light the world, we thought a thorough primer – as well as the first
published worldwide supply/demand analysis that we know of – would be a useful tool
for investors and companies alike. The following report combines a decade of research
on this sector, as well as over a year and half of diligence meetings specific to this
project, that has resulted in a proprietary bottom-up, supply-based model. This model is
based on a worldwide reactor count (the only one that we know of) and includes many
assumptions such as epi, chip, and package yields, which we outline. Our demand
analysis is a top-down approach, forecasting penetration rates and adoption scenarios in
over 15 individual market applications. Our goal is not necessarily to stimulate
discussion about our individual assumptions, but rather to provide a framework to foster
a universal understanding of how key assumptions affect the supply/demand imbalances
in the market.
On a personal note, we would like to thank all of the numerous individuals and contacts
that have shared this data and enabled us to produce this analysis. We are only as good
as our contacts. You have allowed us to develop our market understanding over the past
decade and sincerely appreciate the candid sharing of information. It is as a result of this
candor that we will not be sharing any individual data but rather the aggregate
information, which frankly we feel is the key to this analysis.

Inside
Foreword 2
Executive summary 3
Supply-demand 6
Supply analysis 8
Demand analysis 11
Materials usage & constraints 60
Intellectual property landscape 62
Technology primer 66
Emerging technologies 76