The DisplaySearch Monitor
Summary of the Q4’05 TFT LCD Supply/Demand and Capital Spending Report
By Charles Annis At the end of December the Q4’05 Quarterly TFT LCD Supply/Demand and Capital Spending Report (QSD), which presents detailed insights on actual and forecast AMLCD fab activity, panel supply/demand, and the capital equipment market, was released. Furthermore, the report provides a broader prospective on the FPD industry through discussion of PDP, OLED and color filter fab activity, total announced capital spending, DisplaySearch’s forecasting methodology and a variety of other relevant topics. The user-friendly PowerPoint based report is now covering a wide 13 quarter rolling forecast range, with this edition evaluating Q3’05 – Q3’08. Annual forecasts cover 1999 – 2008, with long range capacity forecasts reaching to 2009. On continued growing optimism for the opportunity in LCDs, particularly in televisions, panel manufacturers keep on increasing the size of capacity expansions plans and pull in next fab investment schedules. In Q4’05 this trend drove increases to the forecasts for the amount of capacity to be installed, ramped capacity available for production, and in the capital equipment forecast. Table 1: Metric Q405 m2/mnth Q305 m2/mnth Difference Q4-Q3 % Difference / Q4 Q3 - Q4’05 increased capacity forecast 2000 4,556 4,556 0% 2001 7,083 7,083 0% 2002 10,261 10,261 0% 2003 15,116 15,116 0% 2004 23,790 23,790 0% 2005 38,845 38,543 301 1% 2006 53,741 53,264 477 1% 2007 73,542 70,692 2,850 4% 2008 88,153 81,450 6,703 8%
CF capacity is growing at a rate similar to array, with 2000 – 2009 CAGR forecast to be 39%. CF capacity is consistently 3-8% higher than array capacity. However, because of some yield loss, color fitler capacity needs to be higher than array capacity. If array utilization is maintained at a high level, color filter capacity will be tight for some panel manufacturers at some substrate sizes. For PDPs, 2006 capacity growth will be only about half of 2005, which is expected to create tight supply and slower price declines. This will likely drive upside potential for PDP investments, some of which are already coming to light, like Matsushita’s announcement in January that it will proceed to build a 2nd new factory in Amagasaki, Japan. The new line will be a state-of-the-art factory, using large 8-Up 42” equivalent substrates and will start production in 2007. FHP also announced that it will pull in the ramp schedule at its new Miyazaki 4 line to help fill the supply gap. The outlook for OLEDs continues to be mixed. PMOLEDs have attained success in MP3 players, particularly in China, but capacity is being added at a high rate, so supply will like exceed demand well into the future. AMOLED mass production is expected to expand in 2006. However, manufacturing and business model difficulties remain as emphasized by early technology leader Pioneer’s recently announced exit from the market. In terms of capacity share, in Q3’05 the top five manufacturers AUO, CMO, LPL, Samsung and Sharp accounted for 72.6% of total capacity. These top tier manufacturers will maintain their leading positions, and will have 72.6% share of
January 2006
Page 1
The DisplaySearch Monitor
ramped capacity in Q3’07. LPL and Samsung have essentially an equal amount of capacity accounting for around 20% each of the total, are followed by AUO and CMO with around 13-14% share each, and then Sharp follows at about 7%. Panel demand in 2006 turned out to be higher than just about anybody expected a year ago. Component supply was unable to keep up with demand; particularly CCFLs for backlights were in shortage during the 2nd half of the year. CCFL capacity will ease somewhat moving into 2006, but glass supply will be limited. So despite rapidly growing array based capacity, supply will be limited to a relatively tight 5 – 6% surplus. Regarding spending for TFT LCD array, cell, module and color filter capital equipment, after reaching an all time high in 2004, equipment spending was somewhat less in 2005, even so, at $9.58 billion it was the 2nd strongest year in LCD history. Now 2006 spending is expected to exceed 2005 by at least 10%, or $10.52 billion. There is still some upside potential in 2006 as some of the investments in the later part of the year are not yet firm. 2007–2008 are now forecast to be in the $7-8B range, lower than the 2004 and 2006, but still relatively healthy for equipment vendors. If higher than expected demand continues to push new investments there is significant upside potential to the long range forecast as shown below. Figure 1: 1999 – 2008 AMLCD Equipment Spending Forecast and Upside Potential
$14.00
$12.00
$10.00
$US Billions
$8.00
$6.00
$4.00
$2.00
$Revenues Upside Potential
1999 $3.63 $3.63
2000 $6.70 $6.70
2001 $3.82 $3.82
2002 $5.21 $5.21
2003 $6.35 $6.35
2004 $12.65 $12.65
2005 $9.51 $9.51
2006 $10.52 $11.26
2007 $8.01 $11.29
2008 $6.79 $9.04
The net change to our forecast of new TFT LCD array, cell, module & color filter capital equipment was an upward revision in cumulative 2005 – 2008 spending by 11%, from $ 31.4 billion to $34.9 billion From 2002 through 2005 fifth generation fabs accounted for the large majority of total dollars invested into new factories. Although there are still significant gen 5 investments in 2006, the shift to larger glass sizes accelerates to gen 7 and larger substrates.
Page 2
October 2005
The DisplaySearch Monitor
Figure 2: Equipment spending by generation
100%
80%
60% Share
40%
20%
0%
8 7.5 7 6 5 4 3.5 3.25 3 <3
1999 0% 0% 0% 0% 0% 0% 4% 56% 56% 14%
2000 2001 0% 0% 0% 0% 0% 11% 41% 26% 26% 4% 0% 0% 0% 0% 11% 10% 32% 18% 18% 21%
2002 0% 0% 0% 0% 45% 16% 12% 25% 25% 1%
2003 0% 0% 0% 11% 59% 27% 0% 1% 1% 0%
2004 2005 0% 0% 13% 24% 40% 20% 1% 3% 3% 0% 0% 0% 22% 27% 38% 8% 1% 1% 1% 0%
2006 0% 10% 43% 11% 26% 9% 0% 1% 1% 0%
2007 2008 0% 34% 12% 20% 16% 15% 1% 1% 1% 0% 19% 19% 22% 21% 6% 13% 0% 0% 0% 0%
Spending on gen 4 and smaller fabs will continue into the future, but except for a few exceptions in China, small glass investments will be for LTPS and AMOLED manufacturing. A key, yet to be clearly answered question, is what will be the next glass size greater than Sharp’s 2160 × 2460 mm substrates. A variety of proposals are on the table that include, highly efficient “Super G5.5” fabs, glass sizes in the range of 2200 × 2600 mm all the way up to 2600 × 3000 mm. Details on the benefits of various glass sizes and the cost of such fabs, as well as a lot of other interesting and relevant information can be found in the Q4’05 Quarterly TFT LCD Supply/Demand and Capital Spending Report.
January 2006
Page 3