ASML Holding NV (ASML) is the world's largest manufacturer of semiconductor equipment. It makes lithography equipment, which is the most critical / costly machinery a chip-maker buys. In 2011, ASML had sales of EUR 4.9 billion out of a total lithography market of EUR 6 billion and a total fab equipment market of EUR 25 billion. Over the last 10 years, ASML has grown revenue at about an 11% CAGR. This growth has largely come from lithography consistently taking share of semiconductor capex budgets and from ASML taking market share from other lithography players.
I like it when investments remind me of patterns that worked well with other securities in the past:
1) ASML is becoming a monopoly in a winner-take-all industry. In some sub-sectors of technology, the leading company in the niche ends up with virtually all the market share. This is partially because the leading company is the only one in the industry that can afford to invest in the research and development necessary to drive technology forward. ASML's market share went from about 30% to 80% in the last 10 years. The company spends about 5 times more in R&D than Nikon, its closest competitor. This has resulted in 90%+ share for the most advanced lithography tools. Intel (INTC) is currently Nikon's 2nd biggest customer and recently decided to make a large investment into ASML. The street seems to be underestimating the long term implications of this move. It pretty much guarantees that ASML will be a monopoly in the not-too-distant future. I also think that it is interesting that ASML's gross margins are currently significantly lower than the gross margins of Intel and many other chip companies. ASML will earn much higher gross margins long term as it benefits from its monopoly position.
2) Opportunities usually arise when investors get overly concerned about short term cyclical weakness and forget about secular growth. Intel recently lowered its 2012 capex budget. The memory companies (DRAM and NAND) have also basically stopped spending money on anything other than technology improvements. They are not adding any wafer capacity. Recently, there were rumors that Samsung would significantly reduce its capital expenditures for 2013.
Several sell-side analysts have lowered growth expectations for semiconductor capex for both 2012 and 2013. I expect that semiconductor capex will very likely be weak through the rest of the year. However, for cyclical stocks, the best time to buy is when business is weak in the short term. Next year should be a better year and the worse that 2012 is, the better 2013 growth will be. The memory segment cannot spend much less than what they are spending now. Furthermore, SSD adoption and tablet demand will likely drive an upturn in memory-related capex at some point next year. In addition to the rumored iPad mini, several Android based tablets are expected to be launched by the end of the year.
Microsoft (MSFT) will also start selling its Windows based tablets. Intel usually grows its capex in the first year of a line-width shrink, which bodes well for 2013 spending. PC demand may also benefit somewhat after new version of Windows is released. While Intel lowered capex in the short term, one should remember that Intel is now facing a much more competitive environment. Intel dominated the PC processor market, but the convergence of tablets and laptops and Intel's ambitions for new end markets like mobile phones put it on a collision course with many other Arm-based chip companies like Qualcomm (QCOM), Broadcom (BRCM) and Apple's (AAPL) own internal semiconductor design division. This new competition will likely make Intel's business more capital intensive over the next several years.
Rumors regarding Samsung's capex are also misunderstood. Like other memory companies, Samsung's memory related capex can't go do down much. If Samsung were to reduce capex significantly, it would have to come from its logic-related capex. Samsung's logic spending would only go down significantly if Apple moved to a different supplier. If this were to happen, the spending would just shift to someone else like TSMC (TSM). Lithography's share of semiconductor capex is also set to accelerate over the next few years. In his initiation report of ASML on June 21, 2012, Bernstein analyst Pierre Ferragu stated that he expects lithography's share of logic semi capex will rise from about 11% to at least 17% from 2012 to 2015. This is because the lithography step of semiconductor manufacturing gets more important with every successive generation of chip technology. According to Ferragu, "Modernizing a logic fab from 45nm to 32nm represents in our estimate a minimal EUR 23m net incremental spending on lithography, assuming no wafer capacity increase. A modernization from 32nm to 22nm would cost EUR 400m."
While the above two patterns remind me of other technology stocks that turned out to be good investments, ASML is still very unique. It is the only company I have ever seen where the industry's top 3 customers actually became shareholders. Intel, TSMC and Samsung will own a combined 23% of ASML. These customers have also agreed to fund a significant portion of ASML's R&D. They realize that ASML is the only viable long-term player for lithography equipment. Rather than trying to fight the trend by continuing to keep Nikon alive, they have decided to give in and participate in ASML's upside by being investors in the company. They agreed to fund close to EUR 1.4 billion of ASML's R&D over the next 5 years because they desperately need ASML to deliver newer generation equipment. Despite this momentous news, ASML's stock is trading only slightly higher than the price these chipmakers will pay. The stock trades at 12.5x 2013 EPS and at a mid-teens multiple of 2012, which is reasonable given that 2012 is likely going to be the cyclical trough for the company. For those that are concerned about the 22% revenue growth rate expected by the Street in 2013, note that the company doesn't need much, if any, industry capex growth to make those numbers. Virtually all of the incremental revenue modeled by the Street can be explained by new products that the company already has orders for. Furthermore, since these new products are "development tools" (equipment customers are buying initially for R&D purposes), they will not be cannibalizing other equipment ASML makes. Margins will actually be depressed in 2013. The street correctly is modeling gross margins to come down because these new products will initially have low profitability. Gross margins should rise again in 2014 as the new products achieve higher scale.
Putting possible short term news aside, a long term investment in ASML at a reasonable valuation and at almost the same price as its three largest customers should work out well. The company has a very large and sustainable competitive advantage, which should drive share gains and margins long term. The coming war between Intel and Arm and the explosion of NAND based devices will make the semiconductor industry more capital intensive. Lithography has steadily taken a larger share of overall semiconductor capex and there are some reasons to believe that this trend will accelerate over the next few years. Finally, the company offers a 1% dividend and has consistently bought back stock. ASML's management has stated that the company will resume buying back shares after the 3 chip manufacturers become investors.
The main risk is timing-related. I do not know when the upturn will start and it is certainly possible news gets worse before it gets better. Technically, ASML's chart is bullish above all moving averages. I suggest taking an initial position now and leaving room to add on weakness.
Additional disclosure: I may trade in and out of shares of securities mentioned in this article without notification.