by James Montgomery, news editor
May 27, 2011 - Building "cumulative headwinds" -- in inflation, consumer skepticism, energy costs, and Japan recovery -- are causing semiconductors and display manufacturers to push out near-term orders, according to Applied Materials execs laying out their 2Q11 books and future projections. As a result they see a softening 3Q11 and the need for a 4Q11 turnaround to meet their as-yet-unchanged annual projections.
Rising inflation in emerging markets, reduced consumer confidence, uncertainty in energy/fuel costs due to Middle East unrest, and Japan's tragic events are causing customers in semiconductors and displays to push out near-term orders, noted CEO/president Michael Splinter, in the conference call to investors and industry watchers. Nevertheless "the fundamental drivers within the electronics sector remains strong," and he's confident that AMAT's customers "have a firm foundation for a multiyear investment cycle." He pointed to record orders in 2Q11 for the company's process diagnostics/control systems, and repeat business for mask inspection at DRAM and NAND customers, with "improved traction" at foundries.
|A snapshot of AMAT's 2Q11 numbers:|
|Overall: Net sales $2.86B, up 6% from 1Q11 (slightly better than projected), reversing 1Q11's -7% decline. Gross margins were down nearly a full point to 41.5%. Net income $489M, vs. $506M in 1Q11; EPS $0.37 within the expected $0.34-$0.38 range.|
|Silicon (SSG): Sales down -3% to $1.45B (vs. expectations of flatness); operating income decreased -10% to $491M, attributed to "weaker mix." Orders up 7% to $1.71B, better than the 5%-6% projected. Customer mix showed less foundry business than in 1Q11 (notably lower at 65nm) but more for flash and DRAM: 47% foundry (54% in 1Q), 25% logic (23%), 16% flash (10%), 12% DRAM (13%).|
|Solar (EES): Record sales of $637M, a 34% spike reversing 1Q11's -22% slide and well above 10% expectations. Orders declined but only -8% to $612M instead of the expected -25%/$500M sink.|
|Services (AGS): Record sales of $614M, up 8% (vs. 3% expectations), thanks to continued demand for 200mm equipment. Orders back up again to $603M, 9% growth in line with estimates. Note that AGS is selling its parts cleaning business (to Quantum Global Technologies), resulting in $24M of asset charges.|
|Display: Sales of $158M/7% growth better than outlook, and orders blew out 40% Q/Q expectations with 80% growth to $255M, thanks to needs from producers of touch panels and mobile device displays.|
|And outlook for 3Q11 and FY11:|
|Overall: Net sales down -3% to -10% to $2.6B-$2.8B; non-GAAP EPS $0.31-$0.37, including anticipated $0.01 for Varian acquisition. But prior full-year prediction of $11B in sales and $1.50 EPS (non-GAAP) are at risk, now requiring a big bounce in 4Q11 (specifically SSG), admitted CFO George Davis in the teleconference, "and will depend on our customers seeing evidence of strengthening consumer demand over the next few months." He thinks this could happen -- "we haven't walked away" from the full-year projections -- and Splinter added that 4Q spending (like 1Q) is typically a big month. "We have to have a pretty solid October to reach those numbers." Splinter kept to his earlier prediction of 2012 WFE of $35B "plus or minus a couple of billion."|
|Silicon (SSG): Sales flat to down -7%, with foundry pulling back. 2H11 sales about flat with 1H11.|
|Solar (EES): Net sales down -15% to -25%, due to digestion of capacity additions.|
|Services (AGS): Sales up "by more than 25%."|
|Display: Sales also up at least 25%. Total LCD capex (including low-temperature poly-Si and touch panels) will be down -10% to -15%, but those two areas will see "very strong" growth in 2012. Big question: how quickly Chinese companies will make investments, if they go through in 1H12 it'll be a very strong year for AMAT's display group, he said.|
Who's spending, who's not
DRAM prices remain soft even after bottoming out in 4Q10, and DRAM capex is "low and limited to technology conversions." Supply constraints for leading tablet product and "flow adoption" for new ones impacted tablet sales, which meant weaker-than-expected orders from DRAM and foundry customers, Splinter noted.
Some pushouts in NAND flash spending resulted from soft tablet demand in 1Q11, Splinter said, but he thinks flash makers will ramp investments again in 2H11. What's holding back DRAM makers, meanwhile, is bit growth "in the 40% to 50% range, which really isn't enough to drive significant capacity," he explained. "Certainly, the big 3 DRAM companies have cash available and could invest in capacity if they need it," he said -- but "they have to have demand to be able to invest."
He expects more DRAM spending in 2H11, but less than in NAND flash and "still very small on relative terms, comparing back to any of the recent peaks of wafer fab equipment spending" (it's about a third of those levels now). Splinter recast AMAT's FY11 wafer-fab equipment investment outlook to $31B-$34B, down from the $34B-$36B raised expectations in 1Q11. Wafer starts are relatively flat as capacity added over the past half-year is being absorbed.
Note that the difference in AMAT's WFE outlook isn't due to any change in spending guidance from chip companies, but rather that equipment forecasts were running ahead of capex forecasts and are coming back down in-line, points out Satya Kumar from Credit Suisse. Also, some of that WFE was for equipment shipped in 2010. He thinks 2012 could be up 10% Y/Y.
Three keys for 450mm
Asked by an analyst for an update on 450mm tool progress, Splinter said "there's a lot of work to do" in areas such as automation and chamber design, and AMAT will "have to start investing more in 2012." While he's not ready to offer a timeline for alpha tools, "it will be when customers need them," i.e. production in the 2015-2017 timeframe.
Splinter also has a unique perspective, one analyst pointed out, having been at Intel's 300mm transition, and now sits on the other side of the wafer-size-transition fence at a supplier. For AMAT's discussion with key early-adopters of 450mm, he says he's stressing the need for both early adopters and suppliers to establish some middle ground where there's "a win-win" for both sides, "a space where they can get their cost reduction and we can still have a reasonable market that's going to grow over a period of time," he said. Another 450mm development necessity: pick and stick to a technology node for the switch -- the 300mm transition had an initial fab at 250nm but real conversion didn't "really start in earnest until 130nm," he pointed out. Another key factor is keeping to a specific timeframe -- the 300mm transition "got delayed several times because of the industry cycles," he noted.
Solar equipment concerns?
Regarding the solar business, questions about solar policies in Italy and associated delays in projects mean AMAT sees significantly less PV installations in 2011, 18GW-22GW (not the 21GW-25GW it expected in 1Q11). Nevertheless the company's wafering and Baccini systems saw another record quarter in sales and profits, shipping >6GW of nameplate screen print capacity, primarily to China. "We expect to see lower shipments in Q3" and are "monitoring capacity digestion," Splinter noted, though he predicted "another strong quarter for EES."
Asked by an analyst about a possible pause in solar equipment spending: "we're seeing the leading-edge guys continue to invest in capacity because they want to be ahead on capacity, especially in a market where material costs are coming down, so that they don't have to have cells in inventory which are depreciating in value at a relatively fast pace," Splinter said. There's always a big surge just ahead of policy and incentive changes (see Italy), which he sees continuing through the end of this year. Look for effective capacity going into 2012 to be around 40GW in the customer base, ahead of installations (as it is this year), he noted.
AMAT summary: Bulls vs. bears
Barron's blogger Tiernan Ray lines up the Wall Street bullish/bearish reactions to AMAT's numbers and comments. First, the "Bulls" -- Stifel Nicolaus' Patrick Ho thinks the weak outlook is "timing related." Gus Richard from Piper Jaffray believes some foundries have overbuilt capacity, and logic makers are stretching too far too ("Intel is adding too much capacity and is likely to pull back next year," and Global Foundries' 140% capex/sales ratio "is unsustainably high"), but excess capacity should be soaked up by demand for gadgets including tablets and phones. Needham's Edwin Mok sees business picking up in 2H11: increasing fab utilization, incremental capex, and a cycle extending into 2012. Now, the "Bears": Weston Twigg from Pacific Crest thinks AMAT's weak outlook is shared by other tool suppliers. Tim Arcuri from Citigroup thinks AMAT will miss its targets in 4Q, and thus its 2011 targets will come in noticeably lower than predicted: $2.3B sales (not $2.6B-$2.8B) and $0.28 EPS (not $0.31-$0.37). Caris & Co.'s Ben Pang agrees the company has an "uphill scenario" to achieve full-year 2011 targets, and has slightly lowered his AMAT outlook for both 2011 and 2012.