Final count is in: there are 108 LED manufacturers: majority in greater China
During the time we’ve worked on our comprehensive report covering LED manufacturers, we’ve seen a few purchase reactors but confirm they are exiting the space and found out about several others we never knew existed. We’ve visited LED manufacturers that have no website, or have a half completed factory and are unheard of outside China, investigated the progress of start-ups in various regions, and expanded our knowledge of top players.
We’ve kept a list of all LED manufacturers that went through continual editing during the report as companies entered and left the space, and as we determined which reactors were purchased for LEDs, which for power, which for mass production, which purely for research, and so on. After much expansion in 2010 and 2011, we can now confirm that there were an estimated 108 LED manufacturers at the end of 2011. It is striking (although won’t be a huge shock to some LED industry insiders) to note that China (45) and Taiwan (27) account for exactly two-thirds of the market, (followed by 11 in Korea, 10 in Japan 8 in the US, 6 in Europe and 1 in Malaysia). While many people in the industry can easily name all the top suppliers, we aren’t aware of such a comprehensive listing and profiling of all manufacturers being done before.
Of course, there is no absolute number of LED manufacturers since any definition of whether a company is or isn’t an “LED manufacturer” will inevitably be somewhat arbitrary for a number of reasons (what exactly constitutes leaving a market? How much exactly is mass production? Etc). It would be pointless trying to come up with a precise definition but what we tried to do was cover several different company types. Firstly, any company that is mass producing LED die, even smaller third tier players. Secondly, any company that we believe is recognized as a leading packager within the industry was counted and included, even if they make little or no die, e.g. Everlight and Kingbright. In practice there are around 30 companies in this category, typically with revenues of $100 million+ per year. Thirdly there are a small number of companies that may focus more on wafer production than die production, but have the capability for both. Finally, there are a few startups that are not quite at mass production stage but worth including, such as TSMC.
Some organisations we didn’t see as worth including, such as pure R and D companies, perhaps affiliated with universities, companies almost entirely focusing on power or other products rather than LEDs, and smaller packagers. There are many hundreds of smaller packagers and we didn’t see any value in trying to track all of these. Only component LED companies are included, not companies that produce only LED lighting finished goods like GE or Zumtobel. Also, we only looked at companies with GaN and AlInGaP production, which are the modern, brighter types. Companies that produce only other LED types, defined as “standard” by IMS Research (rather than “high brightness”) like AsP, GaAs, GaP, AlGaAs are not included in the report. These are older, dimmer types not used in backlighting or lighting and only really used in products like indicator status lamps, industrial markets and toys.
To comprehensively track all players in the main category of die producers, we have tracked almost all MOCVD reactors sold by Veeco, Aixtron and others and identified all of these purchased for LED production. Our global network of analysts have used multiple local sources to verify this information where possible.
Although we didn’t assess the supplier base in the same detail in 2009 and 2010, we believe the number of players was approximately 65 in 2009 and about 90 in 2010. From 65 to 108 in 2 years is an increase of 66%. While new players will continue to enter the market, this number will not change much in 2012 and we can expect to see a few companies merging or going out of business. It’s unlikely to be the established players that suffer. Most of these seem healthy and it will probably be newer entrants without an established customer base that suffer due to tough market conditions. These conditions relate to lighting not yet reaching huge volume, backlighting weakness vs. expectation due to fewer LEDs per display and pricing declines, low capacity utilization (40s to 60s at the moment at many companies) and general economic concerns.
So given this we think the number of suppliers is probably at or near its peak and don’t anticipate much change from this number in 2012. The subsidies that drove growth in Chinese startups are not expected to be as widely available in 2012, nor in 2013.
Of the relatively younger players, so far it has been the Korean players of Seoul Semiconductor, Samsung and LG Semiconductor that have managed to take most market share from established suppliers in Japan and the West. Chinese players have not yet entered the top 10 rankings or captured much market share individually; however, this is likely to steadily change over time. Having said that, the big companies like Nichia, Osram, Cree and Lumileds, continue to wield considerable market power and influence, and this will remain so in the short term at least, as they continue to defend their position.