|
January 27, 03
A Bright, Flat Future
by Matthew Smith, Taipei
Taiwan is fighting it out for the number one spot in a major
high tech manufacturing industry that is poised for extremely rapid
growth.
"It's shaping up to be a real horse race. This is one foreign
observer's assessment of the fierce competition between Taiwan and
South Korea in the flat panel display manufacturing industry, which
is poised for rapid growth -- and tremendous change -- over the
next decade. With potential applications for almost every consumer
IT product sold, color flat-panel displays (FPDs) -- mainly the
thin-film transistor, liquid crystal display type (TFT-LCDs) --
are expected to become a standard feature of products such as digital
still cameras, personal digital assistants, and mobile phones over
the next few years. In the meantime, large-sized applications, notably
for notebook and PC monitors, have been rapidly gaining in popularity,
and a global flat-TV market in the millions of viewers beckons a
few years down the road.
South Korea and Taiwan, the two nations best positioned to supply
the world with flat screens, currently account for roughly 70% of
the global market, with Taiwan slightly in the lead. Japan is an
increasingly distant third, and China, the only other country with
commercial TFT-LCD production capabilities, accounts for less than
1% of market revenue.
For Taiwan, which entered the market for TFT-LCDs just three years
ago with help from Japanese technology transfers, the economic implications
are enormous. According to DisplaySearch, a U.S.-based research
firm that focuses on FPDs, the global TFT-LCD market will grow by
55% this year alone, and industry epresentatives insist that even
that impressive figure will be dwarfed after 2004, when TFT-LCD
televisions are expected to start becoming a popular consumer item.
But makers here face a tough challenge from the major South Korean
players, who are in far better financial shape for expanding their
operations in this capital-intensive industry -- and who, according
to some analysts, are now working to drive down the prices of large
FPDs (those wider than 10.1 inches, measured diagonally) in a bid
to limit the financing abilities of their Taiwanese competitors.
To take advantage of this fastest-growing segment of the market,
most Taiwanese makers feel that they must first build expensive
new fabrication facilities, which will require raising funds.
To be sure, Taiwan's TFT-LCD manufacturers are not without strengths
and resources of their own. According to the Market Intelligence
Center (MIC), a unit of the Institute for Information Industry,
Taiwan accounted for 28.3% of global large-sized FPD production
in 2001. Overall, the nation's manufacturers generated revenues
of US$5.4 billion last year, an annual sales increase of 42% despite
the global technology slump, according to the Photonics Industry
and Technology Development Association (PIDA). With such large potential,
the government is heavily promoting the industry in Taiwan. The
Ministry of Economic Affairs has selected FPD manufacturing as a
key industry for national development (along with semiconductors,
biotechnology, and digital media). And the Council for Economic
Planning and Development (CEPD) has set a target of US$39 billion
in combined production value for flat panels and FPD monitors (with
most assembled in China) for 2006.
However, despite the industry's rapid rise and the government's
rosy forecasts, it would be premature to declare that the FPD industry
is a Taiwanese success story. In fact, many recent international
media reports have warned of the opposite outcome. Critics believe
that, as so often happens in Taiwan, the domestic FPD industry is
fragmented, with too many makers offering too little product differentiation.
This situation raises the specter of over-competition and poor profitability,
fears that have not been eased by the imminent prospect of a massive
increase in global manufacturing capacity.
This year, the world's two largest TFT-LCD panel manufacturers,
Samsung Electronics and LG-Philips, are coming on line with new,
more efficient fifth generation (Gen 5) fabrication plants that
will allow them to significantly boost production output while continuing
to cut the prices they charge for large-sized flat panel displays.
A Gen 5 plant can produce a single 15-inch PC monitor panel, an
industry benchmark, for about US$20 less than a Gen 4 plant. That
figure currently represents a cost savings of about 10% on the price
of such panels.
Going Large
Despite the risks, most makers will feel the pressure to expand
capacity with Gen 5 fabs capable of producing larger size panels.
It's hard to argue about the competitiveness of the new fabs, although
some believe moving to Gen 5 is a matter of timing. Jeff Hsu, vice
president of CMO, says that his company is in no rush to expand
and that those who rush into design and construction -- especially
without plenty of outside technical assistance -- could risk creating
Gen 5 facilities that are less efficient than well-designed Gen
4 plants, the current standard.
Two Taiwan makers, AU Optronics (AUO) and QDI, will ramp up production
at their brand-new Gen 5 fabs in the second quarter of this year.
AUO took the lead in this area when it became Taiwan's fourth company
and the only pure-play TFT-LCD maker to list on the New York Stock
Exchange. AUO raised more than a half billion U.S. dollars to help
pay for its Gen 5 fab.
Building it might be risky, but the larger capacity will be needed
if makers here are to take advantage of the key area of current
and future sales growth: the increasing popularity of larger-sized
17-inch monitor screens and -- industry executives hope -- the onset
of the TFT-LCD television set industry in two years. Nearly 90%
of LCD consumer demand (measured by glass area) is currently attributed
to notebook and PC monitors. Gen 5 plants can cut twice the number
of 17-inch panels as their predecessors, so analysts predict nearly
100% growth in global 17-inch production capacity after the initial
Gen 5 plants -- in Korea and, this spring in Taiwan -- begin manufacturing
operations. This extra capacity will undoubtedly bring down the
end price of the wider screens, perhaps paving the way for 17-inch
LCD monitors to take the place of 15-inch screens as the standard
consumer choice. If so, this would only increase the early-mover
advantage, since companies with Gen 5 fabs would be able to garner
a larger customer base for their wider screens.
If that sounds too theoretical, it should be noted that the same
scenario has already been played out. In 1997, Korean makers began
stealing away market share from Japan by offering 14.1-inch notebook
panels, which quickly overtook sales of the Japanese-made 13.3-inch
models. According to a PIDA forecast, the same thing should happen
next year, with 17-inch monitors making inroads on the booming flat-screen
monitor market.
The larger-size panels that Gen 5 fabs can produce are also seen
as hastening the onset of the TFT-LCD television industry, which
currently accounts for very limited output. According to the MIC,
global demand for flat TVs reached only 120,000 units last year,
but that figure will hit 1.6 million in 2002. Impressive enough
growth, but it nevertheless pales in comparison to this year's forecasted
output of 30 million notebook screens and another 30 million PC
monitor displays.
Gen 5 fabs also can produce screens wider than 30 inches, which
industry executives say is roughly the maximum reasonable size for
a TFT-LCD TV screen, since plasma display panel (PDP) technology
is more appropriate for wider screens. (AUO says it is developing
PDP products that will be commercially available in 2003y.) Given
the 200 million to 250 million television sets sold around the world
annually, 99% of which still come equipped with a traditional cathode
ray tube (CRT), it is understandable why panel makers are rushing
to set up Gen 5 plants. In other words, to take advantage of what
could become the hottest market for displays in a few years, makers
have to upgrade to the next generation of manufacturing technology.
But at the same time, this rapid expansion of capacity is also
fostering concern that TFT-LCD panels are becoming a commodity-type
product. The theory is that the most popular large-sized FPDs are
following the same pattern as did DRAMs, the memory chips whose
prices can fluctuate wildly within just a few weeks. DRAM pricing
instability results from its product uniformity -- there is no differentiation
among different makers' chips, which leaves prices extremely susceptible
to swings in production capacity and market demand.
Likewise, some TFT-LCD industry executives admit that certain highly
standardized products, such as desktop PC monitors, can in fact
be considered commodities. That was made clear last year, when manufacturing
costs were 30% higher than today but the price of 15-inch monitor
panels fell through the floor. But other products, including small
screens for handheld IT devices and notebook panels, are differentiated
from one another by the widely varying design requirements of downstream
makers of finished IT products.
The trouble is that the commoditization of the PC monitor market
- and concomitant pressure on large-panel prices -- is precisely
what is currently driving TFT-LCD demand growth. Prices for 15-inch
panels, which climbed throughout the first half of 2002 to reach
US$250 in June after a serious dip in 2001, had fallen back to roughly
US$200 to $210 by late September. The drop was due partly to sluggish
demand from end consumers, but perhaps also to Korean makers' alleged
strategy of purposely driving down prices. In any case, given the
upcoming boost in capacity, prices are expected to continue declining,
with one industry executive suggesting they could fall as low as
US$150, significantly below production cost, by the end of this
year.
So there are great risks of overcapacity in the market -- but only
if all the proposed fabs in fact get built.
It seems clear, that all these makers have their eyes on the TV
market. "Interest in fifth generation fabrication plants is
mainly all about TV," says David Joe, vice president of global
sales and logistics at Hannstar. "And oversupply will kickstart
TV demand. I want to emphasize that price is the key issue -- as
long as the price is low, the TV market will fly."
Flat-panel manufacturers might be eagerly tuning in to the TV market,
but not everyone is as sanguine about the product's potential over
the short-term. "TFT-LCD TVs will become a big factor in the
future, but it's really not happening at this point," says
the MIC's Sean Wu, who believes that makers will have to wait until
the United States switches to digital TVs in 2006 before seeing
major sales growth. That would require surviving through two or
three years of production overcapacity, and it is unclear that all
the makers could weather the demand drought.
More interesting applications
Industry executives note that there are other potential FPD applications,
such as car audio-video systems. Some 50 million new cars are sold
annually, and makers say that auto companies are now considering
making CAV systems standard features in some models, including minivans.
And with 200 million to 300 million used cars rattling around the
globe, makers are also targeting the aftermarket potential. David
Joe of Hannstar emphasized the benefits of cheaper costs for this
largely untapped market, saying that consumer demand will kick in
when the price of an 8-inch panel car monitor reaches US$150 to
US$ 200. Of course, the technology must be developed to handle extremes
of temperature and other environmental factors, but these problems
are being handled now, he adds.
Another smaller panel application is the color PDA, sales of which
the MIC says will grow by 30% this year. An estimated 40% of the
products are bundled with TFT-LCDs. As for the promise of mobile
phones with color TFT-LCD screens, industrialists insist that it
does not depend on the success of 2.5G or 3G mobile data services.
They project a worldwide market of 500 million handsets next year,
with a color STN- or TFT-LCD bundle rate of 30%. "Once you've
used a mobile phone with a color screen, you?ll never go back to
black and white,? says Jeff Hsu of CMO. However, MIC analyst Sean
Wu again pours cold water on such enthusiasm, arguing that it will
take many years for the color cell phone market to develop.
If selling monitor panels is a commodity game, most industrialists
say they are looking elsewhere for growth. "Sometimes I wish
people would stop focusing on the pricing for monitor and notebook
panels", says one executive who asked to remain unidentified.
"These other applications have so much more potential."
Perhaps in the future, but LCD monitors remain the key driver of
industry growth now. And with prices trending downward, the Taiwan
makers are not the only ones who can see the potential for other
products.
The Korean Competition
The huge capacity and healthy finances of the world's largest
two makers will force more industry consolidation in Taiwan.
Can Taiwan makers compete on their own with the vast scale and
capital resources of the likes of Samsung and LG-Philips? The answer
most experts give is a resounding ?yes? -- but it may require some
industry consolidation. Indeed, while not as large as the two big
Koreans, Taiwan's AUO and CMO rank as the world's third- and fifth-largest
producers. according And AUO is itself the product of the September
2001 merger of Acer Display Technology with Unipac Optoelectronics,
a former subsidiary of semiconductor foundry United Microelectronics
Corp. "Even our customers are consolidating, so they're looking
for big suppliers to service them", AUO chairman Lee explains
his business strategy.
With five Taiwanese firms engaged in the manufacture of large-sized
TFT-LCD panels, Lee believes there are "two or three"
too many. However, he offers no forecasts for further mergers and
acquisitions within the industry. "When the opportunity comes,
it will be quite clear," he says. "But this is probably
not the best time, since most companies are profitable and there
is little pressure to consolidate."
While 2001 was the worst year for the industry and everybody lost
money right across the board, the Koreans invested anyway, garnering
almost half of the worldwide large-sized FPD market share that year
and, at least for LG-Philips and Samsung, putting themselves in
the leadership position of Gene 5 fabs. This raises fears that Taiwan,
particularly the smaller-scale manufacturers, will be squeezed out
by their larger, faster-moving competitors.
Nonsense, say local makers. For one thing, they argue that unlike
the Korean conglomerates, their generally pure-play TFT-LCD business
model allows them to outsource their materials and components, which
is more cost efficient than the Korean model of producing nearly
everything in-house and manufacturing under their own brand names.
Also, they dismiss the notion that the South Koreans will enjoy
an advantage in scale for very long. By the end of 2003, Taiwan
TFT-LCD production capacity (based on area) is forecast to grow
by 61%, compared to just 28.6% for Korea. "Afterthe next four
Gen 5 fabs are built, Taiwan's capacity will be double that of Korea's,"
says David Joe of Hannstar, referring to plants scheduled to be
built by CMO, CPT, and his own company over the next two years.
"That will be fun!"
The Japanese Allies
The world's second largest economy learns the benefits of sharing
-- or at least selling -- its manufacturing technology.
If the depression year 2001's heated competition was tough for
Taiwan makers, it was misery for the Japanese. They saw their share
of the worldwide market fall from almost half in 2000 to 33.5% in
2001, a figure that further fell to 22% this year, according to
Taiwan's Electronics Research Service Organization (ERSO), a division
of the Industrial Technology Research Organization (ITRI). Many
Japanese vendors reacted by moving away from commodity products
into more profitable, smaller-scale lines, such as screens for cellular
phones and personal digital assistants, and into low-temperature
polysilicon (LTPS) TFT-LCD technologies for small-panel applications.
In fact, Taiwan should be grateful to Korean FPD makers, who have
been squeezing their Japanese counterparts for years. After all,
it was this intense pressure, coupled by Japan's decreasing manufacturing
competitiveness and Taiwan's increased technical capabilities, that
forced the Japanese to transfer their industrial know-how to Taiwan
starting in the late 1990s. It is seen as something of a unique
case, and interviewees agree that it is unlikely to be repeated
with any other industry.
The Japanese TFT-LCD makers did not come to Taiwan for charity.
They had no real alternative, other than quitting the business entirely.
With Japan?s economy in a slump for more than a decade and South
Korea on the rise, Japan's world-class consumer electronics companies
decided it made more sense to license their technologies to Taiwan
firms than to invest in costly fabs at home. This way they have
been able to stay in the business while transferring most of the
capital risks to Taiwan.
"I think they are going to make a lot of money out of Taiwan,
since of course we have to pay them for their technology,"
says Liu Yi-hwa of the CEPD.
Rachel Huang, director of ERSO's International Program and Public
Relations Divisions, says her agency was instrumental in convincing
the Japanese to work with Taiwanese makers, something they were
initially reluctant to do. "We built up a base for negotiations
with the Japanese by developing our own technologies," she
says. "When they saw that we had the ability to do it ourselves,
they realized that they had to cooperate with us or else watch us
become their competitors anyway."
The Japanese had another reason for transferring thei technologies
to Taiwan: the nation's massive output of IT products that require
FPDs. Taiwanese manufacturers control 60% of global LCD monitor
production, plus half of the notebook and digital camera markets.
In its first year of operations, CMO, which originally partnered
with Fujitsu, focused all of its capacity on 14.1-inch monitors,
with 70% of its output dedicated to the Japanese market. "Even
as a small player, we made our name, made a good profit, and penetrated
the Japanese market - all within the same year," says Jeff
Hsu.
Also, and in contrast to the South Koreans, most original equipment
manufacturers in Taiwan have traditionally eschewed brand names,
which means that they do not compete with their OEM customers. In
fact, the one notable Taiwanese exception, Acer, spun off its display
subsidiary (which later became AUO) with the declared intention
of eliminating the inherent conflict between simultaneously making
its own brand and those of other makers. "We're a pure TFT-LCD
panel maker, so we have no conflict," says H.B. Chen of AUO.
"The Koreans are major brand name monitor makers, and that's
a clear conflict."
The combination of Taiwan"s OEM competitiveness with Japan's
manufacturing technology and world-class brand names should allow
both sides to take full advantage of the predicted boom in flat
panel TV sales. Sharp is currently the world's largest maker of
TFT-LCD TVs, but the flat television is still a niche product with
just 1% of overall global TV sales this year. "The Japanese
are here for the TV market," says the MIC's Sean Wu. "TV
makers like Sony don't want to continue losing out to the Korean
brands."
But did the Japanese have any viable alternative to Taiwan? Probably
not when the Korean pressure began to bite in the late 1990s. But
that may be changing, as shown by NEC's recent announcement of intentions
to build a Gen 5 plant in the Shanghai area by 2004.
Is TFT-LCD a "China-Proof" Industry
" ?
The mainland is considered far from ready to become a major
player in the capital-intensive and technically complicated FPD
industry.
As noted above, Taiwan's FPD industry suffers from the nation's
all-too-common malady of fragmentation and over-competition. But
almost everyone believes that it is at least partially immune to
that other common bug -- mainland fever! "This may be the one
major industry that is China-proof," says a foreign analyst
who closely follows Taiwan's economic and technological developments.
Industrialists and observers agree that China is years behind in
terms of technical development, logistical support, and ability
to weather the business cycle in such a capital-intensive industry
as FPDs. While the mainland?s IC manufacturing capability has come
a long way, TFT-LCD production entails far more complex upstream
supplier relationships than China can support. With more components
to be sourced, efficient first- and second-stage production of FPDs
would require a major upgrade in the nation's logistical capabilities.
While NEC is indeed planning a Gen 5 plant with a Chinese partner,
they are talking about ramping up production in 2004, relatively
late in the game.
Nevertheless, China is still benefiting from the rising worldwide
market for flat-panel displays. Its huge labor market has attracted
interest from both Korean and Taiwanese makers, who are setting
up their labor-intensive TFT-LCD panel assembly operations (the
third and final stage of production) in China. Not only is this
more cost competitive, but it also provides geographic proximity
to the assembly operations of the FPD makers' downstream customers,
whose IT product assembly plants are also in China. The theoretical
result should perhaps be a major boost to Taiwan's cross-Strait
exports, while the highest value-added parts of the manufacturing
process -- the array and cell processes - remain here for the next
ten years at least.
But there may be flaws in that theory. Intriguingly, Japanese FPD
makers Sharp and Epson have set up their cell processing facilities
in China. Whether or not they can succeed remains to be seen, and
while there is no shortage of skeptics on this side of the Taiwan
Strait, the Taiwanese FPD makers are certainly paying attention.
In any case, industrialists here say they do not see the need to
move their highly technical operations to China or anywhere else
for many years to come, because of the intense requirements of supporting
infrastructure, technology, and engineering. "Whenever any
of our machinery goes down, we can call our equipment suppliers
here and get the failed module replaced very quickly," says
K.Y. Lee of AUO. "It's a big advantage for Taiwan." So
while China is becoming the assembler of FPD screens for both South
Korea and Taiwan, the real horse race only has two entrants. The
industrial chain that is growing up around the Taiwanese TFT-LC
industry should make it difficult for China to "hollow out"
this manufacturing sector for a long time.
Coming next: OLEDs, Organic and Cool !
New technologies have arrived in the (liquid) crystal ball.
TFT-LCD technology might be the Big Kahuna on Flat Panel Beach
right now, but other contenders are shining up their boards. If
new and upcoming FPD applications are more than hype and are actually
going to succeed, panel makers must first develop screens that are
physically thinner and more versatile, offer a wider viewing angle
and still brighter, clearer picture, and have faster reaction times
than current amorphous silicon TFT-LCDs can provide. With help from
foreign technology developers and government agencies like ERSO,
two small Taiwanese panel manufacturers already offer commercially
available screens built with the latest FPD technologies, and the
larger panel makers have noticed.
Viewers are demanding wider screens, but nobody wants them any
thicker. Because TFT-LCD panels require as components color filters,
backlights, and liquid crystals, the thinnest panel is still about
5 millimeters thick. One promising technology can reduce panel thickness
to as little as 1.5 millimeters: Organic light emitting diode (OLED)
panels have a simpler structure that does not require these three
components, instead substituting a thin layer of light-emitting
organic material. OLED panels also offer a brighter and higher contrast
image, a faster response time, and a viewing angle of 160 degrees
- all vital for the flat TV application.
Additionally, getting rid of the liquid crystals eliminates the
costly and time-consuming process of applying them, making OLED
especially suitable for large-sized screens. According to DisplaySearch
estimates, worldwide OLED revenues will grow by 110% annually over
the next four years to reach US$2.2 billion. If the technology works,
it could displace conventional TFT-LCDs in the large-panel market.
"OLED has the most potential of any FPD technology over the
next decade," says PIDA's Cheng Te-kuei.
Truthfully, OLED has been around for years, but the high cost of
the key organic material and the short lifespan of OLED products
have limited its commercial success. There are currently just two
OLED panel suppliers in the world - Japan's mighty Pioneer Electronics
and Hsinchu-based RiTdisplay Corporation. Established in 2000, RiTdisplay's
two fabs are now running at full capacity to fill orders from mobile
phone suppliers, says president and CEO D.C. Wang. "We're confident
that we'll break even and become the number one OLED panel supplier
next year," he says. The company plans to issue a Taiwan IPO
in the third quarter of 2003 as preparation for a GDR or ADR issuance
in the hopes of raising US$100 million the following year. Another
recently commercialized technology, low-temperature poly-silicon
(LTPS) TFT-LCDs, allows for system-on-a-panel screens, in which
the ICs that drive and control the system are mounted directly on
the glass. Because amorphous silicon panels, the current norm, must
be processed at temperatures of up to 500 degrees Celsius and higher,
the ICs cannot easily be mounted directly on the glass. This is
why most FPDs available now have a frame surrounding the screen
(the controllers and drivers are inside the frame), and also a reason
why the speed of the image is insufficient for many multimedia applications.
Put simply, the distance between the ICs mounted in the frame and
the TFT circuits on the panel slows down the system's reaction time.
Because of the atomic structure of poly-silicon, the IC drivers
and controllers can be safely mounted directly on the screen at
lower temperatures, eliminating the need for a surrounding frame,
increasing speed and resolution, and lowering the production costs
of small-sized displays.
Japanese panel makers have invested heavily in LTPS TFT-LCDs, and
Taiwan's large panel suppliers are likewise trying to develop the
technologies. But in Taiwan, only Toppoly Optoelectronics Corp.
has ramped up operations so far, and it claims to have pulled in
25% of the world LTPS market after less than a year of production.
The company is perhaps unique among Taiwanese TFT-LCD makers in
that its technology is all being developed in Taiwan, either in-house
or in consultation with the government-backed Electronic Research
and Service Organization (ERSO),
AUO, Taiwan's largest panel maker, is hedging its bets by developing
both OLED and LTPS technologies. "At this moment, we have not
made any decision on ramping up commercial production of any of
the new technologies," says company president H.B. Chen. Fortunately
for companies that have already invested heavily in amorphous TFT-LCD
plants, the basic principles behind producing the conventional panels
are similar to both LTPS and OLED manufacturing. MIC analyst Sean
Wu says that it won't be extremely hard for makers to spend "a
couple hundred million" to convert their fabs: "They won't
enjoy it very much, but when the time comes they're not going to
have any choice."
- ends -
The above article, courtesy of the American Chamber of Commerce
in Taipei, (www.AmCham.com.tw) is part of a bigger story written
for TOPICS, the chamber's monthly print and online magazine. It
was edited for contents and brevity to suit our international readership.
Ed.
|