Beginning in 2009, the LED industry saw aggressive capacity expansion and a series of strategic alliances, and Taiwan-based LED chipmaker Epistar has been one of the most active. Since 2009, it has invested in fellow LED chipmakers Tekcore, Nan Ya Photonics and Huga Optotech, and vertically, it jointly formed production facilities with both upstream and downstream players to secure material supplies and order commitments.
With conglomerates from South Korea leveraging their broad global presence in the LED industry and the rapid rise of LED applications in TV and lighting devices, Epistar chairman Lee Biing-jye believes integration is critical for Taiwan's LED industry to compete globally. Mergers and acquisitions are not the only method for integration, he said, adding that the pan-Epistar alliance will enable each member to find its own segment and at the same time achieve synergy.
Digitimes recently had the opportunity to sit down with Lee to talk about Epistar's future plans and Taiwan's LED industry.
Q: Epistar recently announced that it will purchase a stake in Huga. What are the reasons for such an agreement for both parties?
A: The alliance will enable both companies to share resources, particularly in technological development. Currently, Epistar holds and has filed for approval for 1,165 patents, and Huga also has over 150 patents. In the future, we will take advantage of each others' know how in product development focusing on backlight and lighting applications.
Q: How will the alliance affect Huga's operation?
A: Huga's management team is still responsible for the company's day-to-day operation, and Epistar will provide the necessary support as a strategic partner. We plan to join forces in research and development, marketing end-user applications and branding.
LED is now too big of a market and has too many applications for one single company. For example, there are LED chips of various sizes, prices, brightness and efficiency for a wide array of products. In display applications, standalone monitors are less price sensitive compared with all-in-one systems, which then requires LED at different price points. In the lighting segment as well, outdoor and indoor usages demands LED of varying brightness.
The strategic alliance allows both companies to utilize each other's resources and strengths in a complex market, and of course it will give us more leverage on procurement. Epistar, as a major shareholder, will have a presence in the boardroom, but we will not occupy more than half the seats.
Q: In the past, Epistar had acquired companies such as United Epitaxy and Epitech. Why did Epistar choose to form a strategic alliance and not make another acquisition?
A: United Epitaxy held a patent for red LED, which was key to several applications, so we believed acquiring the company for technological reasons was the best move at the time. As for Epitech, we wanted a stronger presence in China where the company already had a good foundation. Prior to the acquisition, China only represented less than 5% of our sales.
The decision by Epistar and Huga to strategically align is based on branding. The two companies each has its own target segments with very little client overlaps. It's more advantageous to maintain two independent brand names for different segments. Only about 30% of our client mix overlaps currently, and we will continue to restructure and integrate our client lineup in the next 1-2 years for maximum benefit.
Q: With investments in fellow chipmakers including Tekcore, Nan Ya Photonics and Huga since 2009, the pan-Epistar alliance is now equipped with 296 MOCVD sets. From a global perspective, compare Taiwan and South Korea's LED manufacturing industries?
A: There were about 700 MOCVD sets in 2008 worldwide, of which Taiwan LED makers accounted for 200-300 sets and a 40% share of global LED chip supply. Epistar, at the time, represented a 22% share of the worldwide LED chip market. In 2009, aggressive expansion pushed MOCVD capacity to 1,000 sets and 1,700 sets are predicted for 2010.
Taiwan's share has fallen to 30% amid the capacity ramp worldwide, mainly due to the entry of South Korean players. Samsung will be equipped with 170-200 MOCVD machines in 2010, and LG, though slightly slower in expansion, will have 120-130 sets. Furthermore, Samsung recently announced that over the next 10 years, it will invest 23 trillion won US$18.82 billion in its LED business. Although Epistar is still ranked in the top three in capacity, without effective expansion to stay in step with the rise in demand, our market share will continue to trend down. That's why Taiwan's LED industry needs to integrate and work closer together.
A: Upstream LED chip manufacturing will become a battle of scale in the near future. What are your thoughts on big panel companies such as AU Optronics AUO and Chimei Innolux CMI entering the LED scene?
Q: AUO and Chimei Innolux producing LED chips in house is indeed a competitive business model in the current market place with vertically integrated subsidiaries and affliates and timely market information from panel customers. However, the market place and supply chain are evolving. LED chipmakers like ourselves are forming stronger and direct relationships with display and TV vendors, which enable us to be closer to the end market and get a better feel of what's going on.
LED-backlit TV is projected to account for more than 70% of the LCD TV market as soon as 2013, meaning we need to make inroad to the next high-growth LED application, and that is LED lighting devices. In LED lighting, the same vertically integrated strategy and doing everything in house may not be the most advantageous.
Q: What are the benefits of moving upstream LED production to China?
A: The Chinese government offers some attractive support. For each MOCVD machine, the government subsidizes 10 million yuan, about half the cost. However, there are other issues as well such as efficient utilization of production equipment and human resource quality, so migrating production to China requires careful consideration.
Epistar has an epitaxy plant in Changzhou, but the pace of capacity expansion is gradual. For 2010, the plant will have 30 MOCVD sets. The purpose for setting up the plant is not solely for cost management, but also to have a presence in a booming market and be near our clients in China.
Q: Are you worried that the massive capacity expansion could lead to oversupply?
A: Oversupply is unlikely in the short term. Judging from the global LCD TV market, about 1,000 MOCVD sets are needed to completely satisfy the demand for blue LED. In 2010, there will be 500-600 MOCVD machines capable of producing blue LED, and I haven't included demand from display monitors and other applications.
The number of LED used in each LED-backlit TV is on the decline from 320-340 units for 46-inch models in the past to about 240-280 units. If the LED lighting market doesn't take off by the end of 2011, then continuing to blindly expand capacity could lead to oversupply. Integration that allows better communications between manufacturers is a good way to prevent that from happening.
A: What are your thoughts on the growth of the LED market, and when will supply catch up with demand?
Q: The worldwide demand for LED-backlit TV will top 35 million units in 2010, but there are uncertainties with the ongoing credit crisis in Europe and conflict between and North and South Korea. At the moment, global LED capacity is probably capable of producing 32 million units of LED TVs this year.
As I've mentioned earlier, there will be 700 new MOCVD sets going on line, but about 300 will not be ready until the second half. For a new MOCVD machine to contribute for the year, it needs to be set up and ready to go by May.
LED TV sales are projected to jump to 90 million units in 2011, so undersupply will be the norm in the short run.
Q: What are your thoughts on LED lighting?
A: Judging from the current market, the LED lighting market is going to take off. In Japan, LED light bulb already grabbed a 40% share of the overall light bulb market in terms of sales value and more than 10% in volume. As long as the LED lighting market continues its upward trend, oversupply will not be an issue anytime soon.
Pricing is competitive in the Japanese LED lighting market. In 2011, prices for 40W LED light bulbs is projected to slip to US$15-18 and US$10-12 in 2012.
The global light bulb market is about 20 billion units. LED light bulb's market share is predicted to rise to 10% when cost per 1,000 lumens drops below US$10, and further jump to 50% if cost falls under US$5.
