Applied Materials Still A Buy After 59% YTD Gain

| About: Applied Materials, (AMAT)

Summary

New product areas have little or no competition.

Capital intensity is growing in the semiconductor industry.

AI chips will boost semiconductor fabrication equipment spend.

Applied Materials (AMAT) is a leading manufacturer of capital equipment used to make semiconductor chips and display screens. From the open on Sept. 27, before the Analyst Day presentation, to the close on Sept. 28, the stock price jumped $3.89 per share or 8.3% to close at $50.62. The stock is also up (at $51.60 close of Wednesday) 59% since the beginning of 2017. Given those gains, I would not blame anyone for taking some profits, given the inherent uncertainty of the world and its markets.

Chart AMAT data by YCharts

However, there is good reason to believe that Applied is still undervalued, given its likely future path.

The big easy: FY 2020 guidance and buyback plan

Towards the end of the Analyst Day on Sept. 27, Applied Materials supplied guidance for its fiscal year 2020 (ending October 2020) and added $3 billion to is current share repurchase program.

The non-GAAP EPS target for FY 2020 is $5.08. If Applied is still projecting revenue and profit growth at that point, (and interest rates on bonds are still relatively low) a P/E ratio of 20 would be reasonable. That would project to over $100 per share. Not a bad investment, if it come true. Even at a P/E of 15, it would project to over $75 per share. Note this is a non-GAAP P/E, not a GAAP P/E.

How good is Applied's guidance?

I consulted my old notes and Seeking Alpha articles to see how past projections of Applied Materials' CEOs compared to later results. For instance, in "Applied Materials' Process Advantages Could Build Market Share," from Aug. 21, 2013, I relied on a July 8, 2013, press release targeting non-GAAP adjusted earnings per share of $1.50 to $2.15 by fiscal 2016.

What was non-GAAP adjusted EPS in fiscal 2016? FY 2016 non-GAAP EPS came in at $1.75. Just below the mid-range of the 2013 projection, which is not a bad estimate for three years out. My summary for my notes on the fiscal Q4 2016 analyst conference states: "Continues to show strong growth, but the stock price reflects little of that, yet."

It was after reporting Q4 2016 that the stock finally began to climb, gradually at first, then accelerating after quarterly revenue and profits continued to show strong growth. So, $5.08 per share for 2020? I would say that is as good of an estimate as any, until we have more data closer to that date.

Understanding the market share argument

Listening to the recent Analyst Day, some of the dynamics of Applied Materials position within the industry became clearer to me. Several years ago, in 2011, an important part of the semiconductor creation process, lithography, was forecast for increase as a percentage of WFE (wafer fabrication equipment) total spend. Applied believed this forecast was wrong, and so invested R&D dollars in other areas of WFE. It turned out the lithography actually declined during the period. That is how Applied gained 6 percentage points of WFE market share in the period from 2011 to 2016.

It also shows that some of the critics who predicted the decline of AMAT were correct when they pointed to competition taking market share within lithography. AMAT gave up some territory in lithography in order to win the overall WFE war.

Endura PVD machine. Source: Applied Materials.

Many times I have heard Applied execs, led by CEO Gary Dickerson, talk about "materials technology" rather than equipment. It is important to understand that the end products that need to be created, now and in the future, have some fundamental differences from legacy semiconductors. The legacy devices were mainly planar, constructed on one flat surface. Newer devices, like 3D NAND, are multi-story. So memory has been transitioning from being lithography enabled to being materials enabled. How high you can build is materials enabled, just as one has to change from wood or brick for a single story building to structural steel for a skyscraper. One of the new materials being used, for instance, is cobalt for interconnects.

The resolution of lithography is no longer the primary limiting factor in making ever-more powerful chips. The limitation is in putting the features on the chip where you want them. This requires both precision placement and precision removal of materials. Applied has created equipment to do this. Selective removal technology, for instance, is dominated by Applied and will generate $200 million in revenue in 2017. By 2017 that revenue is projected to grow to $450 million.

Applied Materials is not so much taking share from other WFE manufactures in legacy technologies, as creating new categories of product where there is (so far) little or no competition. This is also true in the improvements it has made in the last five years within existing equipment categories.

Capital intensity argument

Capital intensity refers to the value of the equipment needed to create a semiconductor chip. It is going up, which benefits all WFE manufacturers, but particularly Applied Materials. I already discussed the use of multiple layers in a single chip. If there are ten layers in a chip, then it may take ten times as many machines to generate the same number of chips as when the chips had one layer.

Chips are made on silicon wafers. The bigger the wafer, the more chips per wafer. For some time the industry assumed that the 300 mm wafers of yesteryear would give way to 450 mm wafers. Such wafers exist, but the transition to them has been much slower than expected, and does not seem to be taking off. A 450 mm wafer should go through a process as fast as a 300 mm wafer, so that would mean higher speeds of production (for the same number of same-sized chips). So, again, earlier projections for a lesser need for WFE equipment have proven wrong.

Every few years the size of the features on chips get smaller, measured in nm (nanometers). So otherwise identical devices would shrink at each smaller node. But going against that trend is the consumer demand for more computational power and memory. That makes devices, at a given node, larger. When average chip sizes increase despite node shrinks, more WFE spend is needed to keep up with demand.

The needs of AI (artificial intelligence) devices is pushing for even larger, more complicated chips. Larger chips create another factor that is bad in itself, but good for Applied. A single defect can ruin a device. The more defects per wafer, the more chips on the wafer are ruined. Large chips mean fewer chips per wafer to begin with, and a higher chance that chips will be defective. Applied Materials' technology is crucial to reducing the number of defects per wafer.

One thing I like about investing in AMAT is that I don't have to predict who will win the battle for AI devices to run self-driving cars. Applied will make many of the machines that make the semiconductors for those devices.

Caveats

What could derail Applied's 2020 scenario?

The global economy could stall or erode or collapse. Whatever probability you assign to those scenarios, you should factor into your price target. I am sure the guidance does not include the possibility of global growth slowing to zero, much less going to the negative.

It is conceivable that the semiconductor manufacturers could delay their plans for equipment purchases for their fabs. This could be a belt-tightening measure, or could stem from an actual lack of end consumer demand or projections of a lack of demand. Many scenarios are possible, like a lack of further smartphone expansion into developing countries, or consumers in advanced countries keeping their current phones longer before replacing them. That seems unlikely, but demand slowdowns have happened in the past. Usually they followed a period of overproduction, and so far, have always just been a pause before demand went back to ramping.

Investors could continue undervalue the stock even if it meets its targets. In many ways what Applied Materials does is much more amazing than what Tesla (TSLA) does, but human takes on the future are hard to predict. People can see a Tesla car; they never see an Applied Materials machine produce silicon with 7 nm transistor widths.

Semiconductor equipment manufacturing is competitive. It is conceivable that a competitor or set of competitors could eat Applied's lunch. So far, however, Applied has thrived on competition.

Q4 results back up the argument

Fiscal Q4 results support Applied Materials arguments and projections. Management called it "the best quarter in its 50-year history." Of course, that is the kind of quarter that can make some people overconfident about the future.

Revenue was $3.74 billion, up 33% from $2.82 billion year earlier. GAAP EPS was $0.85, up 85% from $0.46 year-earlier. Non-GAAP EPS was $0.86, up 72% from $0.50 year-earlier. Cash ended at $8.34 billion, debt at $5.30 billion. Cash flow from operating activities was $1.37 billion, so there is plenty of cash to support the dividend and ongoing stock buybacks.

Conclusion

Applied Materials remains substantially undervalued, and a good investment at this price point. By 2020 the stock is likely to reach a range of $75 to $100 per share, depending on the outlook at that point and how investors are valuing earnings. It also should continue to rise gradually, short term, from its current price as earnings continue to rise y/y.

Disclosure: I am/we are long AMAT.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: I have owned AMAT stock since March 2008. At the time this article was written it represented 6.2% of my stock portfolio.

About this article:

Expand
Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500.
Tagged: , , , Semiconductor Equipment & Materials
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here