AMAT Initiation

Overview

Applied Materials is the largest semiconductor capital equipment company in the world.  They are in the process of acquiring Tokyo Electron, the third largest semicap company in the world.  The deal is expected to close in Q4.  Once the deal is closed, and synergies are realized, I expect the combined entity to exhibit the following characteristics:

• Attractive valuation (8x EPS, 8x FCF, 6x EV/EBITDA on FY17 #’s)
• High and sustained ROE (> 35%) and ROIC (> 25%)
• Improving operating margins (from 19% to 25%)
• Large and growing addressable market with extremely high barriers to entry
• Massive economies of scale and #1 overall market share (2x larger than #2)
• Captive customers and pricing power
• Intelligent capital allocation (balance sheet and buybacks)
• High degree of deal complexity means the buy & sell side are not 100% up to speed on the potential earnings power
• Highly motivated, respected, and shareholder friendly management team

Barriers to Entry and Competitive Moats

The barriers to entry in the semiconductor capital equipment market are quite huge.  The amount of upfront R&D required in order to produce a piece of equipment is measured in the billions of dollars, and that’s before considering the costs involved with manufacturing and customer certification.  Intel considers its chief competitive advantage to be its manufacturing prowess…do you think they’d be willing to risk their reputation on an upstart equipment supplier?  No way.  In fact, the customer certification process represents perhaps the largest switching cost/competitive moat for the incumbent semicap companies.  Once a tool goes into a semiconductor fabrication facility it typically stays in there for a decade.  That means customers are hyper sensitive about testing their equipment ahead of time, and if you’re not on their pre-approved list already it’s unlikely that you’ll ever find yourself in a position to earn a spot. 

It’s a somewhat crude tool, but the Herfindahl-Hirschman Index (HHI) is a measure of the size of firms in relation to the industry they participate in, and an indicator of the amount of competition between them.  A rising HHI indicates lessening competition, while a lowering HHI indicates increasing competition. 

As you can see in these two charts, the semicap industry has consolidated over the last 10 years.  In fact, the HHI index has nearly doubled from .056 to .108.  While academic, it at least is a good representation of the fact that the semicap industry should be a beneficiary of a less cutthroat competitive environment going forward, especially once the AMAT/TEL acquisition closes.  The primary customers of the semicap industry have consolidated as well, even more dramatically than the semicap vendors themselves in fact.  One could make the argument that the semicap vendors were forced to consolidate in response to their customers’ own consolidation.  Either way, the net effect of these trends is that the big will get bigger, and the small with either have to be acquired or they will fade away. 

The state of the semicap industry post the AMAT/TEL deal will be one where there is a clear dominant end to end vendor with over 25% market share, followed by ASML which operates in the completely distinct lithography segment, followed by Lam in dep/etch and KLA in metrology.  After these big 4 vendors, the list drops off to niche and/or subscale vendors.  ASML will retain its lithography monopoly (and is still a great company), which means semiconductor manufacturers will only have to make once choice: do they want to go with the end to end approach at AMAT, or the specialty approach with LRCX and KLAC?  Ultimately, it does not seem unreasonable to think that each of these big 4 vendors will gain share at the expense of the sub-scale players in the market, and the competitive interactions amongst AMAT and LRCX/KLAC will be benign.   

Industry consolidation has had pronounced effects on various sub-groups within the technology industry over the last 10 years.  First the hard disc drive group consolidated down to just 2 duopolistic players, and their profitability exploded.  Then the analog semiconductor industry underwent a decade long consolidation that has resulted in expanding margins, cash flows, and distributions to shareholders.  More recently the effects of the DRAM industry consolidation have become apparent in MU’s incredible stock price action.  I believe the semicap group, led by AMAT/TEL, will be the latest instantiation of dramatically improved profitability as a result of industry consolidation in the technology industry.  It stands to reason that industries with high fixed costs and lower variable costs benefit from industry consolidation because if pricing in an industry like this stabilizes the incremental margins can be tremendous.  Perhaps the high fixed cost nature of the technology industry is why consolidation has had such a pronounced effect in many of its sub groups. 

Macro & Secular Drivers + Technology Factors

Two key points here: 1. Semiconductor equipment companies make most of their money at the leading edge lithography node.  2. The costs to keep up with Moore’s Law are accelerating.  The combination of these two factors has created the interesting dynamic where economies of scale are becoming more important than ever before for both the semicap vendors and customers.  IE, if you don’t have the capital to keep up with the leading edge node you will be left behind. 

A 100mm² chip today now has over 1b transistors on it.  Leading edge films are now 2000x smaller than a human hair and require the selective removal of atoms that are smaller than human DNA.  This stuff is becoming incredibly complicated.  Semiconductors have now in fact reached the physical limit with respect to the possible transistor shrink on a 2D surface.  As a result, the industry is rolling out chips with 3 dimensional surfaces.  These chips are known as 3d finfet in the logic space and 3d NAND in the flash memory space.  For our purposes, the most important aspect of this transition from 2d chips to 3d chips is the associated increase in capital intensity.  Foundries will face a 25-35% increase in capex as they transition to finfet architectures while memory suppliers will face a 35-50% increase in capex as they transition to 3d NAND (for an equivalent number of wafers).  As a result of these trends, wafer fab equipment (WFE) spend is forecast to increase at a steady rate in the coming years. 

Interestingly, previous node shrinks have been driven largely by lithography advancements, hence ASML’s massive share gains over the last 10 years.  However, the transition from 2d to 3d architectures will actually result in a much greater degree of incremental capex going to the materials engineering firms (AMAT/TEL and LRCX to the greatest extent).  As a result, I expect ASML’s share gains to take a pause until 2018 or so when EUV is rolled out en masse, and I expect AMAT/TEL and LRCX to see increasing share gains for at least the next 5 years. 

Margins and Cash Flows

Margin expansion and improved free cash flow generation are key components to the AMAT thesis.  The boom/bust historical nature of the semicap industry is now moderating.  The amplitude of future cycles will be dramatically smaller vs. previous cycles due to consolidation on the part of the semicap vendors as well as their customers, leading to a greater degree of predictability and higher margins.  As a result, the semicap industry is making a transition from an extremely competitive, technologically intensive industry to an industry that more closely resembles an industrials group.  The analog semiconductor group has gone through a similar transition over the last 10 years, and profitability has expanded dramatically over that period of time.  For example, from 2004-2013, TXN’s operating margins expanded from 17.7% to 28.7%, and its distributions to shareholders increased from $715m to $3.3b annually.  ADI’s operating margins expanded from 18.7% to 31% while its cash distributions increased from zero to $600m.  I believe AMAT will follow a similar pattern. 

AMAT’s operating margin was 13.7% in FY13.  Since Dickerson and co took the reins they made margin expansion a key priority; operating margins are headed toward 19%+ this year.  I believe with further cost efficiencies, top line growth, and synergies from the TEL deal AMAT will be able to do 25% operating margins by FY17, with room for further expansion from there.  Although management has provided a target model with 25% operating margins to the street I do not think the street is fully discounting this outcome (skepticism is pervasive, many feel the target model is far too aggressive).  I believe they can hit and even exceed this target model over time, and companies that are expanding operating margins at a clip of 200bps per year are often great investments, in my experience. 

In addition to this aggressive margin expansion, Applied is poised to benefit from a lower tax rate.  Its current rate of 24% is poised to drop to 17% soon after the TEL acquisition closes.  Additionally, since the combined company will reincorporate in the Netherlands, the long term tax rate will end up near 8% (starting in 2020).  As a result, free cash flow conversion will increase, and the company will have an easier time moving its cash between tax jurisdictions – making buybacks, dividends, and acquisitions easier to execute.  Ultimately, I see FCF of $1.7b in FY14 expanding to nearly nearly $4b in FY17 and continuing to expand thereafter, with another jump up in 2020 when the tax rate drops to 8% (reflected in my DCF). 

Management Team & Incentives

Gary Dickerson took the helm at Applied in September of 2013.  He has been an absolutely transformational figure at the company.  His background includes the President & COO role at KLAC from 1986-2004, and CEO of Varian from 2004-2011.  He sold Varian to Applied in 2011, and was subsequently named President at Applied before taking over as CEO in ’13.  When he was named President in June 2012 the street interpreted this as a signal that he would eventually take over as CEO and the stock began the rally that has lasted to this day.  The street absolutely loves Mr. Dickerson because of his focus on creating shareholder value, his ability to innovate and execute, and his history of achieving high levels of customer satisfaction.  Mr. Dickerson gets loads of credit for turning KLA into the metrology powerhouse it is today, and is equally credited for getting a fantastic price when he sold Varian.  He is the architect behind the Tokyo Electron acquisition and his long term relationships with the TEL management team really made this deal a unique possibility.  Under Mr. Dickerson’s leadership, Applied has invested significantly in field technical support in an effort to increase customer captivity, reallocated resources from G&A to R&D in an effort to accelerate new product development, and increased its aggregate market share by 1.4% in the last year alone.  He has also dramatically pulled back on solar investments and pulled that group up to the breakeven level vs. -250% op margins last year.  One last anecdote: the difference in energy between Mr. Dickerson and his predecessor Mr. Splinter is incredible.  Dickerson brings a palpable dynamism with him wherever he is, and that is exactly what Applied needs right now. 

Bob Halliday was brought in as CFO at Applied in 2013 largely on the recommendation of Mr. Dickerson.  The two had worked together very successfully at Varian since 2004.  The first words out of Mr. Halliday’s mouth at Applied’s recent analyst day were on the topic of taking care of shareholders: “The stock price going up is how we define winning.  At Varian we debated replacing the conference room chairs even though they were falling apart.  We asked ourselves, would buying new chairs drive shareholder value?  We are very shareholder oriented with respect to each of the investments we choose to make.”  I like management teams with that type of orientation.  Dickerson and Halliday make a great team. 

Insider ownership and activity: From June 2011 through October 2012 multiple directors made several open market purchases of the stock at levels between $10-12.  Insider sales have been relatively modest over the last 2 years, even as the stock has nearly doubled.  Aggregate insider ownership is still < 1%, but importantly the current management’s team’s compensation is > 90% stock oriented, which aligns their interests with ours.  I have incorporated future stock option grants into my share count assumptions. 

Capital Allocation

AMAT management considers TXN to be the “gold standard” in returning cash to shareholders in the technology industry.  TXN management has committed to return 100% of its FCF to shareholders, and it has worked very hard to improve FCF conversion to impressive levels.  I believe AMAT will undergo a similar transition going forward. 

Management has already committed to doing a $3b buyback in the 12mo following the closure of the TEL acquisition.  This is a nice gesture, but I think it’s just the beginning.  In fact, AMAT will acquire $2.72b in cash from TEL’s balance sheet once the deal closes, and after suspending buybacks for FY14 their cash balance would end up at $6.5b if they didn’t immediately do a sizable buyback.  Add in a $3b annual FCF run rate (on its way to $4b by FY17) and you’ll find a company swimming in cash pretty quickly.  Additionally, take into account the fact that AMAT/TEL will be doing $5b in EBITDA by FY17 and you’ll find that the company will only be levered up to .6x Gross Debt/EBITDA (and it will still have a net cash position of at least $3b even after an initial round of buybacks).  My perspective here is that AMAT will be able to take on at least an additional $3b in debt just to get to 1x gross debt/EBITDA (.2x net debt/EBITDA), and allocate that cash toward increased buybacks.  With $2.72b cash from TEL, $3b in additional debt, and $9b in FCF between FY15 and FY17 AMAT will be able to fund $14.5b in buybacks over the 3 years following deal closure.  At the current price of $21.10 that would represent 687m shares, or 38% of total shares outstanding post deal closure…in just 3 years.  Wow. 

I have also included annual dividend raises which would get to a 3.3% yield by FY17 in this calculation.  Should this management team decide they could actually handle some real leverage you can add $5b to the aggregate buybacks for every 1 turn on net debt to EBITDA (e.g. if they go to 1x net debt/EBITDA add $5b, 2x net debt add $10b etc).  Every additional $5b would represent an additional 235m shares, or 13% of the float.  So at 1x net debt/EBITDA AMAT could retire 51% of its float!  Obviously I am expecting stock price appreciation to mitigate the efficacy of these buybacks, but even a 20-30% reduction in float will still be extremely compelling. 

Tokyo Electron Deal

AMAT remains convinced that the TEL merger is on track to close 2h14.  I get the sense from the company that they are aiming for a November/December close, but regulators are sometimes unpredictable in their timing.  MOFCOM, China’s Ministry of Commerce, is the surprising regulator who may have the most problems with the deal.  They typically finish their deal review process within 180 days, but just last week they passed the 180 day review limit and AMAT had to re-file.  This is unusual and could be an indication that MOFCOM may ask for additional deal concessions.  AMAT has already offered to divest $600m in revenues post deal, so that is already built into my model, but if this number goes up it could be dilutive to my expectations.  If MOFCOM’s demands are too great there is a chance that AMAT could decide to walk away from the deal, which would obviously be a bad thing for the stock.  There is some apprehension in the market around this deal closing, but my best guess is that it closes by year end with $600m in divestitures (although we will have to watch developments here closely). 

Management has already produced a target model post deal closure:

My estimates are largely inline with the company’s upside case model, although I am using a much lower share count (1.38b vs. 1.6b) to arrive at my $2.59 EPS vs. management’s $2.40. 

There is a surprisingly small degree of product overlap between Applied and Tokyo Electron, although there is obviously 100% overlap in customers.  Even in the areas where the two companies do overlap - such as dry etch, dielectric etch, silicon etch, and CVD – there is a clear dominant player between the two companies, making it obvious who will be driving the bus with regard to each of those particular product lines.  As a result, the costs that can be taken out range widely from overhead, sales and marketing, cost of goods, as well as some efficiencies to be gained in research and development.  The combined company will be the only end to end supplier of semiconductor capital equipment in the world…period.  Their breadth of product offerings will make them a one stop shop, and their involvement throughout the entire lifecycle of semiconductor manufacturing will enable them to see every aspect of the fabrication process which will in turn give them a leg up on new product development (anticipating their customers’ needs more quickly than the competition). 

The mechanics of the deal are straightforward.  This is a 100% stock deal.  Each AMAT shareholder will receive 1 share in the new entity for every AMAT share held, while TEL shareholders will receive 3.25 shares for every TEL share held.  According to my math, the current deal spread is 4.4%.  If the deal closes by the end of the year you can make an additional annualized return of 10.6% by buying TEL shares instead of AMAT. 

Valuation

My approach to valuing AMAT/TEL is to determine the earnings power of the combined entity once it is fully integrated in FY17 and then discount those earnings back to the present day.  When I do this (see attached model) I arrive at $2.59 in EPS in FY17.  Then I discount this earnings stream back 2 years at 11% per year, and apply a 14.5x forward PE multiple.  This analysis produces a 12mo price target of $30.52, which is a 45% premium to the current share price.  An alternative approach is to apply 14.5x to 2.59, yielding a $38 price target for those that are willing to hang onto AMAT shares for 2-3 years.  I believe a 14.5x earnings multiple is actually conservative, and should AMAT execute on the TEL acquisition according to its plan it would re-rate higher.  TXN is the ideal that AMAT strives to achieve, and TXN trades at 17.2x forward earnings – a multiple that is certainly more aggressive, but not out of reach for AMAT over time.  If we were to apply a 16x multiple to our earnings power figure it would imply AMAT shares will double by 2016. 

I always like to do a gut check on my valuation work with a long term DCF.  AMAT’s tax rate declines to 8% in 2020 and the DCF captures that nicely.  My current best guess DCF indicates a current equity value of $62b for AMAT, which equates to $38.55 per share, which makes sense when compared to my earnings power analysis. 

Recommendation

Buy AMAT shares for All-Cap.  If by some crazy miracle AMAT dips to below a $20b market cap in the near term start buying it for mid-cap, too.  We started with a 100bp position and we have been adding over the last week.  My inclination is to continue adding at a steady pace between now and this fall.  There may be a period of increased angst around fears of a delayed or cancelled closing of the TEL acquisition over the next 3-4 months, but once the deal closes we’ll be off to the races and I’d like to make this a big position.  I would take it up to 200bps over the next month and look to make it 300bps as we get into the fall. 

Risks & Mitigating Factors

• Valuation: Standalone AMAT is only forecast to earn $1.05 this year.  At $21 the stock is already trading at 20x this year’s EPS, which is not cheap. 
• Competition: The combined AMAT/TEL organization will be a behemoth.  It is not uncommon for a merger of two giants to lose share to smaller specialists, LRCX and KLAC in this case.  Should AMAT/TEL lose share it will fall short of its target operating model in FY17. 
• Japan: Tokyo Electron is a Japanese company with a culture that is very different from the Silicon Valley culture at Applied.  Should there be a culture clash this combination could fail.  Additionally, it is extremely tough to fire employees in Japan which could present a risk in achieving the combined entity’s target model.
• Should the TEL acquisition fail to close, AMAT’s stock would likely decline to the $15-19 range immediately. 
• MOFCOM is China’s Ministry of Commerce and they have held up the AMAT/TEL acquisition for longer than the typical 180 day review period.  They are likely asking for concessions of some sort.  AMAT has preemptively offered to divest $600m of revenues from the combined company, so that is already contemplated in their long term model, but there is a decent chance that regulators in China, Korea, Japan, Europe, or the US come back with more onerous terms. 
• EUV: if EUV rolls out and is extremely successful by 2018 it could lead to a decline in the number of deposition and etch steps required in a typical fab.  That would hurt AMAT, TEL, and LRCX. 
• Moore’s Law: if it becomes no longer economically viable to produce leading edge semiconductors in mass quantities due to a diminution of Moore’s Law semicap vendors will suffer as their previous generation tools can be used for longer periods time, hurting new tool sales.