Description

Entegris (ENTG – NASDAQ): Compounding Semiconductor Complexity Compounds Value

1) Thesis Description

Entegris (ENTG) is a manufacturer and supplier of microcontamination control products (MC), specialty chemicals and engineered materials (SCEM) and advanced materials handling (AMH) products predominantly for semiconductor manufacturing processes. The semiconductor industry is undergoing change in two key aspects that are of importance to the company, breadth and complexity. Real and durable use case expansion for semiconductors driven by 5G, IoT, AI et al should become more apparent starting this year as the coronavirus pandemic abates, which has arguably accelerated secular digitization trends and its real-world secondary impacts. Further, increasing manufacturing complexity due to smaller transistor sizes (5nm and below) and performance limitations of traditional materials at those smaller scales are forcing semiconductor manufacturers to high-grade suppliers, while leaning on material specialists more intensely to solve their most pressing performance/manufacturing yield issues. Entegris is expected to benefit from these developments in a compounding fashion with improving revenue growth, margins and competitive dynamics after each successive transition to a new/smaller transistor node.

The thesis is as follows:

1)      At the current price of $94.37/share, the market assumes ~7% revenue per annum growth (~45% below its topline five-year average), and minimal value to margin improvements. Valuation ranges from $81/share to $268/share over five years, with scenarios ranging from no growth/no margin improvement to ~55% market share of the semiconductor materials supplier market.

2)      An opportunity is available due to several reasons:

a.      Entegris’s value proposition underestimated as the company has evolved from an ordinary to a one-of-a-kind comprehensive supplier, making direct comparisons to smaller competitors problematic. Since inception, the company has grown its portfolio from simple wafer carriers to a full suite of products/materials utilized in the majority of the numerous processes in semiconductor manufacturing. Competitors such as Pall (Danaher), Cabot Microelectronics, MKS Instrument, Ultra Clean Holdings, ACM Research and a handful of Japanese firms have products that address only individual steps in the overall manufacturing process, while Dow, Parker Hannifin and Merck KGaA have exposure to a few steps in the manufacturing process. Notably, most of these competitors have substantive business lines addressing other markets. Further, ENTG’s revenue is derived ~70% from consumables and ~30% from fab capital spending, unlike most competitors who are inversely proportioned. Primary focus on the semiconductor manufacturing market and a comprehensive, multi-touch point portfolio underpinned by recurring consumables makes Entegris a one-of-a-kind supplier, with only a few commoditized product lines. This uniqueness results in very different growth trajectories, customer/market dynamics, and therefore a variant value proposition compared to single-touch point competitors. As such, it would be more appropriate to compare Entegris to semiconductor suppliers who operate in an oligopolistic/duopolistic fashion (20x-25x EV/EBITDA) as opposed to direct competitors (10x-15x EV/EBITDA), who may lack semiconductor focus/comprehensive portfolios and, over time, should behave more analogous to secondary supply sources.

b.     The company’s expertise throughout the semi manufacturing process should become increasingly valuable as materials science drives semiconductor performance going forward. In the past, the improvement in semiconductor performance had to do with design variation/shrink and little to do with materials. Roughly 80% of incremental semiconductor performance improvement came down to transistor shrink and ~20% to materials. As transistors sizes continued to decrease in size the relationship flipped, and now ~80% of the incremental performance improvement in semiconductors is due to advances in materials. This is because traditional materials cease to work as expected at smaller transistor sizes given atomic-scale interactions between substrates and circuit lines, and even the smallest of impurities can destroy numerous transistors in these highly dense chips. Transistor shrink could reach an asymptotic level, either 1-2nm, in the coming years simply due to the fact that silicon atoms are ~2nm and smaller transistors in experiments have exhibited quantum mechanical interactions. Once that point is reached, materials science should drive almost the entirety of semiconductor performance improvements. Entegris’s unique ability to understand a customer’s entire manufacturing process and provide custom solutions, transports them to the forefront of the competitive landscape embedding them closer to their customers more so than in the past given the multiple intensifying industry trends that demand materials expertise.

c.     Transition to smaller node sizes results in non-linear increases in demand for Entegris’s filtration and specialty chemical products that amplify revenue growth beyond the linear secular expansion in semiconductor unit volumes. With each transition to a smaller node, more layers and transistors are packed onto 200mm or 300mm wafers. Accordingly, the number of manufacturing steps and amount of materials used increases exponentially. Further, increasing demands on materials purity means that at smaller node sizes manufacturers replace the expensive filters more often. This is evidenced by a 2x-3x increase in industry spend per wafer for logic materials (28nm to 5nm) coupled with 2x increase in filtration spending per wafer, and 4.5x increase in industry spend for 3D-NAND memory materials and filters (64 to 256 layers). Leading edge logic and NAND memory designs should become industry standard (trailing edge) in the coming years, essentially locking in higher spending levels with the benefit of expanding market share based on unit volume. This compounding dynamic in addition to the secular growth in semiconductors (10%-12% per annum) should result in higher and more consistent levels of revenue growth for Entegris’s MC and SCEM segments (~70% of revenue) than a simple historical analysis of other semiconductor firms/suppliers would imply, while AMH division (~30% of revenue) should continue to be linearly tied to wafer unit volume growth.

d.    Competition and pricing pressures should decrease at the leading and trailing edge with each transition to smaller transistor nodes. Historically, as products matured and rolled off the leading edge, smaller competitors would then offer similar products at lesser prices resulting in 1%-2% pricing decreases per annum for market leaders. Transition to smaller transistor sizes and the purity demands at those levels should allow the leading materials suppliers to gain share and lock in a higher quality product standard on the back of higher R&D intensity, which then amplifies R&D scale advantages for the next node transition. The net effect results in smaller players not only getting pushed out of new higher margin designs but also lacking the R&D scale to produce substitutes of high enough quality when the semiconductor market moves that particular design to the trailing edge. In addition, changing customer behavior due to purity concerns (the increasingly detrimental impact on yield) makes them less inclined to purchase lower priced substitutes. As node transitions occur (10nm to 5nm and 1nm eventually) and smaller nodes mature comprising more of semiconductor capacity, Entegris should observe greater revenue and less competition/pricing pressure in its market than it’s historically experienced.  

e.      Concerns of obsolesce due to new advances in semiconductor design fail to recognize the company’s increasing visibility into customers development pipeline. In time, the semiconductor industry should reach transistor nodes in the 1-2nm size range.  With 2nm being the limit of silicon substrates, a new material would probably be used to get to 1nm and future improvements in semiconductors would be based on materials variation/innovation. As evidenced through multiple node transitions, Entegris is being involved earlier in semiconductor design programs and is able to not only provide customer solutions that span almost the entire manufacturing process but also have visibility into its customers future developments, increasing customer retention. From there, the company is able to either provide for that future design or purchase a smaller firm with the missing expertise/product before it reaches an inflection point in industry adoption. This visibility compounds with each transition to a smaller node and should become more pronounced as materials drive the semiconductor industry forward.

f.      Fears over the bifurcation of the semiconductor market due to geopolitical rifts likely misplaced for Entegris given close geographic proximity to its largest customers and the growth opportunity from the redundancy now being built into the industry’s manufacturing base. Despite trade disputes between the U.S and China over the last few years, Entegris has experienced minimal disruptions due to the placement of facilities near their largest customers in the U.S., Taiwan, South Korea and Japan. Notably, only 10%-15% of ENTG’s revenue is derived from China. Further, the company’s U.S. produced materials are not the target of trade tariffs/sanctions; the end product is, leading edge semiconductors. Accordingly, semi manufacturers (Samsung, TSMC) are planning to build U.S. domiciled facilities to minimize impacts from future geopolitical conflicts. Whereas in the past, if these tensions did not exist, the manufacturers would have simply expanded capacity in their current locations, the most efficient use of capital. Instead, their actions, while more costly for them, present an opportunity for suppliers going forward as they sell into duplicate facilities of the same manufacturer, the only difference being the political sphere of influence in which they operate.

3)      The largest micro risks in the name is execution issues with manufacturing expansion/utilization, M&A execution issues and potential customer switching. The largest macro risks are semiconductor unit volumes, growth capital spending and drastic changes in semiconductor design that don’t require advanced materials.

2) Business Analysis

A Brief History – Expansion Throughout Semiconductor Manufacturing Process Elevates Company’s Industry Standing and Value Proposition

Entegris, headquartered in Billerica, MA, was founded as Fluoroware in ’66 then renamed Entegris in ’99 following the merger with EMPAK and went public in ’00. At that time, the company was primarily focused on wafer handling products. The entire semiconductor manufacturing process will not be discussed in its entirety; there are hundreds of steps during which a variety of materials are applied or removed and many steps are repeated multiple times depending on the complexity of the design. However, the main steps that concern Entegris are at the front end of the manufacturing process and include: raw wafer and chemical transportation, implantation, lithography, deposition, dry etch, wet clean and chemical-mechanical polishing (CMP).

The merger with Mykrolis, in ’05 resulted in a $1B+ valuation, expansion into filtration products and a Delaware incorporation. Then in ’14, Entegris acquired ATMI for ~$1.1B which expanded the company into critical materials and handling solutions. Since then the company has added capabilities throughout the semiconductor manufacturing process, buying smaller and product-specific firms.

In ’18, the company bought SAES Pure Gas for ~$340M at ~3.25x EV/Revenue or ~9x EV/EBITDA and Particle Sizing Systems for ~$37M at ~2x EV/Revenue. SAES expanded ENTG’s gas purification portfolio and PSS enabled online real-time particle size analysis in fluids. Notably, the SAES acquisition was purchased ahead of a node transition and before industry-wide adoption of its specific products. Entegris believes that the acquisition can achieve ~20% per annum revenue growth for the next 3-5 years.

In ’19, Entegris purchased MPD Chemicals for ~$160M at ~4x EV/Revenue, Anow for $73M at ~2.5x EV/Revenue and Digital Specialty Chemicals for ~$64M at ~2.5x EV/Revenue. MPD and DSC focus on specialty chemicals and Anow produces unique membrane and filtration solutions in Asia, all of which deepen ENTG’s portfolio.

In ’20, the company bought Sinmat for $75M at ~2x EV/Revenue and GMTI for ~$35M at ~3.5x EV/Revenue. Sinmat provided new slurry technology for CMP processes and GMTI brought high precision analytical instruments for CMP of new substrates such as SiC (silicon carbide) and GaN (gallium nitride).

Through this methodical portfolio expansion, Entegris is one of the only companies that interacts with wafer growers, OEM semiconductor tool manufacturers, device makers, the semiconductor manufacturers R&D/operations teams as well as its’ chemical/gas suppliers. The company has materials solutions that span the upper fab air system, fab floor, sub fab and outer fab facilities where chemicals are shipped/stored.

Entegris has facilities in U.S, Taiwan, South Korea, Malaysia and Japan as well as partnerships in China.

Entegris has three main market segments: Specialty Chemicals and Engineered Materials (SCEM), Microcontamination Control (MC) and Advanced Material Handling (AMH).

In ‘20, revenue is comprised of ~40% from MC and ~30% from SCEM and AMH, respectively. Approximately 70% of revenue are consumables and ~30% is tied to capital spending at fabs. Geographically, revenue is comprised of ~25% from the U.S., ~20% from Taiwan, ~15% from Korea, 10%-15% from Japan, 10%-15% from China and less than 10% from Europe and SE Asia, respectively.  Over 90% of revenue is derived from the semiconductor market (~70% to fabs) and ~10% from industrials/medical/other markets. Notably, Taiwan Semiconductor amounts to ~11% of revenue.

Specialty Chemical and Engineered Materials (SCEM)

The company’s SCEM segment provides highly customized and highly pure chemistries, liquids, gases and other materials for chemical vapor/atomic layer deposition, cleaning, CMP and integration of complex materials in semiconductor manufacturing. Many of these products are created in conjunction with a semiconductors R&D team for logic and memory applications, while precursors are designed closely with OEM tool manufacturers. Importantly, it is through this segment that Entegris is able have increased visibility of its customer’s development pipeline. That information is then utilized in coordination with the MC and AMH segments to create other custom products that interact in lock-step with the special chemicals developed at SCEM.  Revenue is ~95% unit driven.

Products in this segment include: SDS gas cylinders, specialized graphite, silicon carbide, conductive foam, precursors for deposition of tungsten, titanium, cobalt and aluminum films as well as high quality coatings for dry and plasma etch, chemical vapor deposition and ion implantation. Other chemicals are provided for medical and other industrial processes.

Microcontamination Control (MC)

Entergris’s MC segment offers solutions to purify liquid chemistries and gases used in the photolithography, deposition, planarization and surface etching/cleaning processes. These products are designed to handle specific elements/molecules and require tight integration with the SCEM segment and the semiconductor manufacturer. Additionally, the filters are sold to chemical suppliers and device makers. Revenue is 70%/30% unit and capital spending driven.

Products include: air/water/nanometer membrane filters as well as facility-wide gas purification systems.

Advanced Materials Handling (AMH)

The company’s AMH segment develops solutions to monitor, protect, transport and deliver liquid chemistries, wafers and substrates to and from wafer fabs and semiconductor fabs. Raw chemical suppliers also utilize these products to deliver to semiconductor fabs. This segment is the least technical of ENTG’s business units, and does not exhibit the same compounding dynamics, visibility, stickiness and margins as the other business units, though does complete an end-to-end product portfolio. Revenue is 50%/50% unit and capex driven.

Products in this segment include: high-volume wafer carriers, reticle protection products for photolithography, fluid transfer valves, fittings, tubing, pipes and customer fabricated systems as well as digital flow control software and particle sizing instrumentation.

Management History – Minimal Insider Ownership (<2.5%), Long Tenured Senior Management

Bertrand Loy, CEO, joined Entegris in ’05 and was EVP of supply chain, COO and then CEO since ’12. Prior to that, he was CFO of Mykrolis from ’01-’05.  Loy’s compensation comprises of a ~$870K base salary, a cash bonus of ~$1M and a stock bonus of ~$3.5M or ~65% of total compensation (part long-term compensation plan). He is required to own 6x his base salary in shares, owning ~0.9% of shares outstanding.

Gregory Graves became Entegris’s CFO in ’08 though joined the company in ’02 as VP.  Prior to that, he was a VP of M&A at Piper Jaffray, MD of Corporate Finance at RBC and in other IB/Corporate Development roles. Mr. Graves’s compensation comprises of a ~$465K base salary, a cash bonus of ~$375K and a stock bonus of ~$1.25M or ~60% of total compensation (part long-term compensation plan). He is required to own 4x his base salary in shares, holding ~0.05% of shares outstanding.

Compensation performance metrics are based on EBITA margins (~75%) and key business objectives (~25%).

The majority of the senior management have tenures at Entegris or its previous entities spanning 10-25 years.

The executive team and insiders own ~1.5% of ENTG.

Customer Dynamics – Intensifying Demands for Specialty Materials and Purity to Drive Chip Performance Improvements Make Customers Less Price Sensitive and More Dependent on Entegris

The company sells its products to semiconductor companies, device makers, semi equipment manufacturers and chemical suppliers to these companies. With greater material complexity and purity demands, Entegris has observed tighter coordination earlier on in a customer’s design process. Chemical suppliers to semi companies are under similar purity scrutiny and utilize ENTG to ensure their products meet acceptable tolerances.

Semiconductor manufacturers typically exhibit long development lead times and are increasingly outsourcing materials expertise to providers who understand the upstream and downstream effects of a material or process change. Most vendor changes occur at node transitions, given the high switching costs during full-run production. Node transitions occur roughly every 3-5 years, and design adjustments ensue in 6-9 month intervals afterward.

Importantly, ~80% of the incremental performance improvement in semiconductors is due to advances in materials. And with transistor shrink reaching an asymptotic level in the coming decade (1-2nm, current leading edge node is 5nm), materials should be of the upmost importance for semiconductor performance. Importantly, Entegris’s customers compete fiercely on chip performance. In so many words, Moore’s Law, the doubling of transistors on a chip every two years, is breaking down and materials innovation is more dominant in computing performance improvements.

Additionally, the complexity of the semi manufacturing process and greater amounts of materials used are forcing customers to focus more intensely on yields. This results in consuming more specialty chemicals and swapping out filters more often, decreasing their price sensitivity to ENTG’s leading products as even the smallest impurity in both materials and filters can cost much more than the amount saved by switching to less expensive substitutes. In comparison to semiconductor equipment, Entegris’s products (~70% consumables) are barely a rounding error in total fab capital spending.

The top ten customer’s amount to ~45% of revenue and Taiwan Semiconductor equates to ~12% of revenue. Consolidation is ongoing but may benefit Entegris as they have facilities purposely adjacent the largest customers in the industry.

Notably, semiconductor manufacturers have historically put pressure on suppliers and many have cited 50% gross margins as a ceiling. ASML’s competitive position and customer dependency on EUV lithography machines results in them being one of the only semiconductor suppliers with gross margins durably above 50%.

Deep understanding of a customer’s processes and their increasing risk aversion towards materials, allows Entegris to become locked in as a supplier at each design change or node transition with decreasing price sensitivity. Lastly, with materials driving closer to 100% of improvements in chip performance, customer should depend on ENTG to a greater degree. 

Supplier Dynamics – Most Raw Materials Widely Available, Ability to Moderately Pass on Costs

Entegris’s inputs comprise of many materials which are then developed into specialty chemicals and other products. The products with the most raw material inputs are typically highly customized for clients and are small batch runs, making them less susceptible to inflationary pressures. Further, there is wide availability on many of these supplies, though single sources or limited groups of suppliers for membranes, petroleum coke and polymer resins have necessitated multi-year supply agreements.

Input pricing pressure hasn’t been observed thus far at the company. Price increases for competitive reasons has almost never occurred for mature products, but the company has been able to pass along base materials inflation on newer products given minor design tweaks that usually occur every 6-9 months.

Competitor Dynamics – Intensifying Industry Complexity Pressuring Those Who Do Not Focus on Semiconductors and Shutting Out Price-Oriented Subscale Competition

In the past, material purity demands and complexity were not high on a semiconductors list of priorities concerning chip performance and manufacturing yield. Competition therefore, came from many suppliers who focused on a particular process or material. Consolidation of both the customer base and supplier base has decreased the number of players involved and as a whole the industry has become more price rational and less cyclical due to strong demand drivers. The landscape now comprises of these single process/material suppliers that are now under the umbrella of larger parent companies (Pall under Danaher, Versum under Merck KGaA, several under Parker Hannifin), independent suppliers (Cabot Microelectronics, Brooks, MKS, ACM) and subscale local price-oriented suppliers. In reality, Entegris competes with the larger parent companies whose subsidiaries span a few steps in the semi manufacturing process, and to a lesser degree the independent single step providers. Lastly, the company is more focused on recurring consumables (~70% of revenue) than equipment sales (~30%), whereas most of its competitors are positioned inversely.

While ENTG’s competitors that reside under a parent company are equally capable in several sub-segments, increasing R&D intensity at the leading edge is forcing those parents to reassess investments to the semiconductor market. This was most acutely observed when the industry transitioned to 7nm and 5nm designs, and Entegris was able to capture more share than the previously anticipated. Additionally, ENTG’s semiconductor focus should be a benefit as customer’s desire certainty in supplies, where competitors with other priorities may waver.

Smaller single step/independent providers are at increasing disadvantage over time as well, since materials and their increasingly costly upstream/downstream knock-on effects in the fabs benefit players whose products/expertise span most of manufacturing process.

Lastly, local price-oriented suppliers are at the most disadvantage as the industry evolves, since they may lack capabilities of even the single step providers coupled with customers less likely to be price focused.

The competitive dynamics cited above are most intense at the leading edge (20nm and below) which comprises ~25% of total semi capacity at the moment. Over time, the leading edge and trailing edge capacity percentages should remain the same. However, 5nm-10nm nodes should expand to a greater capacity percentage at the trailing edge (30%-35%), while 3nm nodes and below should become the leading edge (~25% share).

All that said, the industry is highly unlikely to 100% single-source products (Intel is an exception), though given the dynamics above this norm should moderately change and competitors beyond the few industry leaders may become more of a second source only if necessary. At present, Entegris has 50%-60% market share in MC and as a whole the company has ~35% market share. Market shares range by product from ~30% to as high as ~90%. Considering MC is at the leading edge of the industry it is reasonable to assume SCEM and to a lesser degree AMH can achieve similar share levels over time.

Market Trends – Secular Growth in Semiconductors Accelerating, Material Intensity Should Continue Increasing

Semiconductors

From ’15 to ’20, semiconductor growth amounted to ~6.5% per annum. In ’20, the semiconductor market grew ~5% to ~$435B despite the coronavirus pandemic. In ’21, the market is expected to grow between 8.5%-12% to ~$475B, and grow 10%-12% per annum thereafter through ’25 according to World Semiconductor Trade Statistics and others. This is driven by 5G, IoT, AI, EVs et al, which is essentially large scale use case expansion. Further, memory makers are expected to grow even faster in ’21 at ~12% minimum.

Notably, Taiwan Semiconductor is expected to increase capital spending over 50% y/y to ~$28B, not only to build out for the next leading edge nodes but to get ahead of future expected unit demand. Additionally, South Korean fabs are expected to increase capital spending ~20% this year to ~$20B, an all-time high.

Lastly, ASML, who essentially has a monopoly in Extreme UV lithography equipment, sold 31 EUV units last year, more than the 5 sold in ’19. This indicates the growth at the leading edge nodes is accelerating as equipment is purchased before a ramp in unit volumes.  

Transistor Designs and Materials

For logic, leading edge designs (20nm and below) should comprise over 25% of total logic capacity in a few years, and even encroach into mainstream logic capacity (~35% of capacity) but not likely analog/discrete capacity (~30%).

At present, the semiconductor industry’s leading edge for logic is gaining momentum at the 3/5nm transistor size level. Silicon (standard wafer material) has a ~2nm atomic size and transistor shrink is increasingly more difficult or simply not physically possible as quantum mechanics starts to encroach at such small scales. Wafer designs utilizing silicon germanium, silicon carbide and gallium nitride are possible replacements and are being used in high power applications today. New gate designs at 2/3nm may be necessary, such as the Gate-All-Around Field Effect Transistor (GAA FET). The knock-on effects from these materials-based constraints impact all other aspects of the semiconductor industry.

After 1nm, if we get there, quantum computing may be the next transition, which would be even more highly materials intensive since transistors may not be gates but nanomagnets, single particle magnets where +/- would be equivalent to a gates electrically generated 1/0. Much of this is very far off and speculative, but still highly dependent on materials.

For memory, 3D-NAND designs with over 96 layers should comprise ~90% of total memory production capacity in a few years. From there, the verticalization of NAND memory should continue to grow with designs over 500 layers.

Similar to logic chip development, future memory wafer designs are likely to utilize smaller structures to fit more data on a similar sized chip, such as DNA (~2nm in size) or even crystal structures. Designs that are likely still highly dependent on materials.

3) Why now?

Entegris compounds many elements of their business with each advancement at the leading edge and the current node transition to 5nm should also be coupled with secular growth driven by worldwide digitization accelerated by the pandemic. At present, the name trades at ~18.75x EV/’22 EBITDA based on consensus revenue growth of ~12% in ’21 scaling down to ~8% growth  by ’23, ~35% below the company’s ~12.5% historical topline growth rate. However, the company’s biggest customers are expected to increase capex 20%-50% this year, and unit volumes are expected to grow 10%-12% per annum above the five-year average of ~6.5% per annum.  If the company grows at a more probable ~15% per annum (M&A ~250 bps growth per annum) through ’26, ENTG trades at ~17x EV/’22 EBITDA, ~15% below the low end of the valuation range for a secular compounder in the semiconductor space. We advocate entering into a position as the company’s advantage at the latest node begins to compound and the market lags in catching up to the paradigm shift.  

A few key points below illustrate the company’s value potential at this point in time:

1)     Strong Organic Growth from Ongoing Node Transition and Secular Semiconductor Demand: As evidenced by the growth plans from TSMC, Samsung and ASML, industry-wide capacity expansion at the 5nm node is accelerating in ’21. This particular transition is significant given the difficulties some manufacturers (INTC) have encountered due to materials complexity and the increased market share that has already occurred in Entegris’s favor. Further, the semiconductor industry as a whole is expected to grow low double digits this year. Historically, ENTG has grown organically 8.5%+ per annum equating to 200 bps of outperformance to industry growth. More recently, with compounding materials demand at smaller nodes, this outperformance has increased to 300-400 bps per annum bringing organic revenue growth into the low double digits. We anticipate this outperformance to continue as the materials complexity/demand and market share gains only intensify, permitting Entegris to lock in their products as design standard which creates a long tail of recurring revenue. Demand for semiconductors driven by the digitization of the physical world (5G, IoT et al) underpin stronger expected long-term industry growth of 10%-12% per annum, above levels seen historically (5%-10% per annum). With most fabs at/near capacity, semi manufacturers are debuting large increases in capital spending for leading edge designs to get ahead of future unit volume growth, reinforcing our growth expectations. Going forward we anticipate Entegris grows organically ~12.5% per annum through ’26, above an expected and conservative ~8.5% per annum industry growth rate.

2)    Company’s Unique Strategy and Industry Trajectory Should Drive Increasing Customer Visibility and Retention; Tactfully Bolt On Acquisitions Ahead of Inflection Points: Entegris is one of the only materials science suppliers that focuses predominantly on the semiconductor industry, and has a comprehensive portfolio/customer base spanning almost the entire front end manufacturing process. Intensifying semiconductor complexity coupled with materials performance and yield concerns are forcing customers to work more closely with Entegris and reveal their future development pipeline. The company is able to not only increase customer retention by providing custom solutions for their customers but also gain insight into the direction of the industry. As such, Entegris can get ahead of industry trends by developing new products internally or purchasing a firm with said capabilities before industry-wide adoption, ensuring their competitive positioning as the industry evolves. This M&A strategy was most apparent with the SAES acquisition, where ENTG could accelerate the target’s revenue growth to ~20% per annum for several years following a node transition in ’18. In fact, sometimes a customer pushes Entegris to acquire a smaller company, so as to improve their operations and de-risk that particular supplier for the manufacturer. M&A typically amounts to ~250 bps of incremental growth per annum, with substantive revenue synergies and modest cost synergies acquired for valuations below 10x EV/EBITDA. As Entegris gains further visibility into their customers design plans, M&A should continue to be highly accretive and should increase the company’s top line growth to ~15% per annum through ’26.

3)    Higher Margins at the Leading Edge Should Become More Durable Long-Term as Trailing Edge Becomes More Complex as Well: In the past, node transitions typically allowed a supplier to charge higher prices to compensate for higher R&D, while over time, smaller copycat competitors could compete on price. This allowed semi manufacturers to demand price concessions (1%-2% per annum) as technology nodes matured. Despite these conflicting forces, Entegris’s margin profile has improved substantially over the years, from ~21.5% EBITDA margins in ’15 to ~30% in ’20, ~170 bps per annum, partly due to manufacturing scale. However, transition to the 5nm node and the greater purity requirements at the near atomic level put onerous demands on suppliers, while making customers more risk adverse/less price sensitive than before. As the 5nm node reaches the trailing edge, this margin improving customer behavior should endure. Importantly, this has already been observed at Entegris with pricing pressure in recent years decreasing to ~1% per annum. This should result in gross margins increasing from ~45.5% in ‘20 to ~50% in ’26. While we anticipate ENTG’s gross margins top out at ~50%, should the firm evolve in its market similarly to ASML it may achieve comparable 50%+ gross margins. Notably, the company’s incremental EBITDA margin which has historically averaged ~40%, inflected upward to ~46% in ’20, a function of the node transition and improved pricing environment. Going forward, we anticipate Entegris can conservatively expand its EBITDA margins from ~30.5% in ’21 to ~39%, equivalent to its historical ~170 bps per annum due to gross margin expansion and manufacturing scale.  

4)    With Each Node Transition, Advantages/Influence Accrue to Market Leaders and Competitive Environment Should Moderate: Growing semiconductor complexity raises the R&D costs for both manufacturer and their suppliers. This R&D intensity at the smallest nodes is partly the reason semiconductor manufacturers are consolidating, with benefits accruing to the largest players. Suppliers are undergoing similar shifts though in a lagging downstream fashion. Notably, this has already occurred amongst the semi equip suppliers with ASML taking inordinate share at the latest node, given their unique solution for EUV lithography. The materials supplier market is more fractured than the semi equipment market and represents an opportunity for Entegris. The company has continued to spend 7%-8% of sales on R&D, expanding their R&D scale advantages with each successive node transition, while consolidating through M&A. Over time we expect ENTG to extend their leadership position to a greater degree with competitive dynamics and customer influence improving in their favor similar to the top-tier semi equipment firms at the leading edge.  

5)    Option Value for Large M&A and Other Industries Growth: In the past, the company has attempted large transformational deals, ie. Versum which eventually was acquired by Merck KGaA. ENTG currently has plenty of financial firepower to effect a transformational deal going forward. Entegris’s current net-debt-to-EBITDA is ~0.9x, with the availability and willingness to reach ~3.75x. This equates to debt issuance capacity of at least ~$1.5B for future M&A that is far beyond our expectations of ~250 bps of inorganic revenue growth per annum (which at most amounts to $200M-$250M per annum in later years). The only gaps in the company’s portfolio, where they would be likely to engage in large scale M&A, is in CMP slurries (Cabot’s dominant semiconductor product) and in some photoresist/photolithography products. Lastly, ~10% of revenue comes from industrial, medical and other industries. While not a core focus, expanding use of Entegris’s chemical/biologic packaging for the medical supply industry, water filter membranes and advanced graphite materials for industrial applications could present solid option value as those industries evolve. With an added effect of damping out volatility in the company’s core semiconductor market.  

4) Concerns/Thesis Pressure Points

Semiconductor Growth

The growth in semiconductor use cases underpins the thesis of secular growth and therefore valuations in excess of what has typically been observed in the industry. Should 5G, IoT and all the other acronyms fail to generate substantive demand for semiconductors, the value of the industry would be expected to moderate from these levels.

Offsetting this risk is confirmation of expanding use cases from automotive/EV growth, strong consumer 5G phone demand, industrial automation demand and growing IoT devices, all of which is causing a near-term chip shortage and fab capital spending increases in excess of historical norms.

Node Transition Execution Risk

At each node transition, suppliers compete to have their products locked in at the leading edge, which then leads to a long tail of recurring revenue. Should Entegris fail to win a contract with a manufacturer shifting to the latest node, they may be locked out until the next design iteration while also missing out on long-term recurring revenue.

The company’s market share gains at the latest node and increased visibility of their customer’s development plans mitigate this concern.

Competition Risk

The company’s competitive environment is quite fractured at the moment, with subscale local competitors able to compete on price but only a few large global competitors that pose a threat. While we anticipate diminishing competition to an oligopolistic/duopolistic level, if the remaining global competitors are determined to increase their R&D spending to keep pace with their customer’s higher standards, the supplier market may consolidate more slowly than anticipated.  

Evidence of increased market share, public statements by large competitors concerning the level of semiconductor R&D investment and analogous consolidation evolution in different parts of the semi market underpin the contention that competition should moderate.

Concentrated Customer Risk

Semiconductor manufacturer consolidation has accelerated in recent years and Entegris’s top ten customer’s amount to ~45% of sales, with Taiwan Semiconductor equaling ~11% in ’20. With increased influence, these company’s may push back to a greater degree on their suppliers, offsetting advantageous supplier market share gains and increased dependency on materials specialists.

Notably, concentrated customers have not been an issue in the semi equipment market for critical leading edge products and we anticipate the industry’s demand for secondary supply sources generally continues.

Customer Country Risk

Over 75% of semiconductors are made in Asia, the majority of which is in Taiwan (20%-25%), Japan, South Korea and Malaysia. Taiwan in particular is a flash point in the geopolitical tug of war between China and the U.S. Conflict surrounding Taiwan or trade issues with China may cause unforeseeable uncertainty to the semiconductor industry.

Despite being U.S. domiciled, Entegris has not had many issues with trade conflicts as they have local facilities adjacent to their largest customers. Further, the leading edge semi manufacturers plan to build facilities in the U.S. de-risking both themselves and the industry as a whole.  

Semiconductor Design Risk

Although the current development path for semiconductors involves considerable materials expertise, new designs that utilize completely different materials and require less raw materials/supplies are a risk. Developments outside of this paradigm that are completely divergent from the current ecosystem are highly unlikely.

Notably if this shift was occurring it would first be investigated/developed within a semiconductors R&D team, where Entegris would already be and would therefore have visibility into.

5) Business Valuation

Entegris should exhibit double digit organic growth over a number of years as the semiconductor industry evolves with more complexity and dependency on materials science and associated products. Additional top line revenue growth should come from acquisitions given the company’s ability to see a customer’s future design requirements and then purchase a critical subscale suppliers ahead of increases in demand for the target’s products. The company generates revenue through the manufacture and sale of specialized materials, filters and materials handling products for semiconductors, their suppliers and others in the ecosystem.  

A summary of the Base Case assumptions for the company is below:

1)     Short-Term Guidance 1Q/21 Revenue Growth: ~25% y/y, ’21 Organic Revenue Growth: 11%-13% (100 bps bump from COVID vaccine transport products), EBITDA margin: ~30%, Capex: ~10% of revenue.

2)      Long-Term Guidance: Organic Revenue Growth: 9%-10% per annum, EBITDA Margins: 30%+, R&D: 7.5% of sales, Opex (incl. R&D):  ~15% of revenue. Capex: ~7% of revenue.

3)      Total revenue should grow ~16% to ~$2,155M in ’21 and average ~15% growth per annum to ~$4,270M by ’26. M&A should amount to 250 bps in revenue growth per annum.

4)      M&A metrics assume ~27.5% EBITDA margins and acquisition multiples of ~8.5x EBITDA (assumes 17.5% cost synergies, and revenue synergies of ~10% realized in one-to-two years).

5)      EBITDA margin of ~30% in ‘21 should improve to ~39% by ’26, a ~170 bps expansion per annum, in line with historical expansions despite improved pricing environment going forward. Effective tax rate increases from ~19% in ’21 to 21% in ’22 and beyond.

6)      Capex: ~10% of revenue in ’21 scaling down to ~7% thereafter.

7)      Programmatic returns to shareholders: Dividend: ~$45M per annum. Repurchases: ~$60M per annum

8)      ~22.5x normalized EV/EBITDA multiple for Entegris. Reasoning behind the multiple is in the Peer Analysis section below.

9)      Discount rate at ~10% (large-cap).

Five-Year Operating Model

A simple five-year operating model is utilized to determine value.

Base Case:

Base Case assumes the company grows ~15% per annum (12.5% organic growth) from ’21 through ’26 and expands EBITDA margins to ~39%.

-           Base Case Valuation: $183.50/share.

Upside Case:

Upside Case assumes the company grows ~17.5% per annum (~15% organic growth) through ‘26 and captures ~55% of a ~$20B end market.

-           Upside Case Valuation: $268/share.

Downside Case:

Downside Case assumes the company grows ~7.5% per annum (5% organic growth) from ’21 through ’26 and does not expand EBITDA margins, and is at a 20% discount to our expected 22.5x EV/EBITDA normalized valuation.

-           Downside Case Valuation: $81/share.

Peer Analysis – Trading Comps

Entegris has no direct public comparables, as it’s one of the only pure play semiconductor materials suppliers and essentially a hub with many spokes throughout the industry and customer’s manufacturing processes. The list of comparables below consider firms which exhibit similar characteristics as to end markets, products, competitive positioning and customer dynamics. Analysis below utilizes a normalized EV/EBITDA metric to account for the margin implications of scale and market consolidation.

Public Comparables – EV/EBITDA

1)      ACM Research (ACMR) – ~25x normalized EV/EBITDA; currently at ~26x FTM.

2)      ASML Holdings (ASML) – ~27.5x normalized EV/EBITDA; currently at ~30x FTM.

3)      Brooks Automation (BRKS) – ~20x normalized EV/EBITDA; currently at ~22x FTM

4)      FormFactor (FORM) – ~15x normalized EV/EBITDA; currently at ~17x FTM.

ACM Research is a developer of semiconductor wet processing technology. The firm provides a range of applications for semi cleaning and wafer level packaging. Their technology at the latest node has exhibited minimal competition thus far and the firm as a whole is expected to grow rapidly (~40% per annum) for several years as leading edge semiconductor volumes ramp. Notably, this is the only direct competitor utilized for our comparison analysis, due to their singular focus on the semiconductor market and their rapid growth/minimal competing alternatives at the leading edge.

ASML develops and produces semiconductor manufacturing equipment specifically for photolithography. The company’s solution for 5nm nodes, EUV lithography equipment, is highly unique and has resulted in a near monopolistic position at the leading edge. Growth should average 10%-12% per annum over the next three-to-five years.

Brooks is a manufacturer of automation solutions for the semiconductor and life sciences industries. Approximately 55% of revenue is derived from semiconductors with capital equipment products in wafer handling and contamination control. Notably, this is one of the competitors most cited by as a comparable by analysts due to their similar focus on microcontamination in semiconductor fabs, though Brooks is more focused on equipment outside of the process chamber. Growth should average ~10% per annum over the next three-to-five years.

FormFactor develops and manufactures advanced semiconductor wafer probe cards, just downstream of the front end manufacturing process. The company is experiencing similar customer and competitive dynamics as other front end suppliers and is a net beneficiary of the increasing complexity of memory manufacturers. Growth should average ~10% per annum over the next three-to-five years.

The group’s normalized EV/EBITDA multiple equates to ~22.5x and has ranged between 10x-40x from ’15 to ’20 and currently trades at ~24x FTM assuming ~11% per annum revenue growth. Entegris currently trades at ~18.75x ’22 EV/EBITDA (consensus 12% and 9% revenue growth in ’21 and ’22, respectively).

Notably, our valuation multiple aligns with our previous work on Vicor (20x-25x EV/EBITDA) though Entegris likely has lower risk and better long-term growth possibilities. Detailed above, we observe increasing valuations as we ascend the front end manufacturing process from FORM (~15x), to BRKS (~20x), AMCR (~25x) and finally ASML (27.5x), which corresponds with the increased level of consolidation, customer dependency/expertise the closer the company is to the manufacturer. Consequently, as a market leader in a consolidating sector, ENTG could exhibit competitive characteristics and a valuation closer to ASML over time than our analysis currently gives credit.

Peer Analysis – EBITDA Margins

EBITDA Margins

1)      ACM Research (ACMR) – ~20% normalized EBITDA margin; currently at ~16% FTM.

2)      ASML Holdings (ASML) – ~37.5% normalized EBITDA margin; currently at ~32.5% FTM.

3)      Brooks Automation (BRKS) – ~22.5% normalized EBITDA margin; currently at ~17.5% FTM

4)      FormFactor (FORM) – ~20% normalized EBITDA margin; currently at ~20% FTM.

ACM’s operating margins are expected to increase ~250 bps per annum over the next three years as the company achieves scale with its core technology. Capital expenditures should average ~6.75% of revenue.

ASML’s operating margins have increased ~55 bps per annum since ’15 though are expected to expand ~160 bps per annum for the next three-to-five years given their position in EUV lithography. Notably, the company’s gross margin is expected to exceed 50% during this period, one of the only semiconductor suppliers above 50%. Capital expenditures have averaged ~5% of revenue.

Brooks’s operating margins have increased ~175 bps per annum since ’15 and should scale ~135 bps per annum going forward.  Capital expenditures have averaged ~8.5% of revenue.

FormFactor’s operating margins have increased ~280 bps per annum since ’15 and are expected to expand ~100 bps per annum for the next five years as gross margin fractionally improves and manufacturing scale takes hold. Capital expenditures have averaged ~5% of revenue.

Entegris’s comparable group generates ~25% normalized EBITDA margins, underpinned by operating margin expansion of ~160 bps per annum. Entegris generates ~30% EBITDA margins and should expand margins ~170 bps per annum.

Peer Analysis – Conclusion

For Entegris, a ~22.5x EV/EBITDA normalized valuation appears reasonable based on the representative comparable group of semiconductor suppliers whose segment of the market is further along in consolidation (20x-25x). Further, this valuation multiple equates to a ~20% discount to the company (ASML) that operates in a competitive environment several years ahead of where ENTG is in the semiconductor materials supplier market today. Notably, we did not include a strategic acquirer analysis as we anticipate the company being more valuable as an independent company and a consolidator as opposed to an acquisition target.

6) Market Expectations/Perceptions

Entegris’s is covered by 10 analysts with an average price target of ~$112.75/share. Seven analysts have a Buy rating and three have a Hold. Belief in the company’s growth trajectory and the ability of management is widely acknowledged by market participants. Management is known for being highly focused on operations/returns and conservative on forecasts, typically beating expectations. Notably at their November ’20 Analyst Day, the company raised their three-year organic revenue growth target to 9%-10% per annum, up from 5%-8% previously. For reference, ENTG’s prior three-year organic revenue growth rate (8.5%+ per annum) was above this prior guidance.  

Consensus forecasts ~12% revenue growth for ’21 slowing to ~8% by ’23, roughly 35% below its historical five-year average for total revenue growth.  Sell side consensus for EBITDA expansion of 80 bps per annum to ~33% margins in ’23 is ~50% below the company’s historical average. The market’s growth forecast appears to anticipate slowing organic growth, while not including inorganic growth opportunities. Underlying the growth disconnect is likely the lagging recognition of semiconductor use case expansion due to the acceleration of digitization by the pandemic.

Further, concerns of customer consolidation fail to recognize the importance of materials suppliers/expertise going forward when transistor shrink is no longer the performance driver. Lastly, there is a lack of recognition around Entegris’s compounding dynamics and its’ ability to see further into their customer’s development pipeline.

7) Downside Protection – Where’s the Margin of Safety?

The company’s downside is well protected as Entegris provides crucial materials with few substitutes to the leading semiconductor manufacturers and their suppliers. If the company were to go away overnight, the industry would be in turmoil and would set the leading semiconductor manufacturers back a few years.

At ~$100/share, the market assumes revenue grows ~55% below the company’s historical norm, with little improvement in margins at a ~22.5x EV/EBITDA normalized valuation multiple. At ~$75.25/share, the market would be assuming no growth/margin improvement at a ~22.5x EV/EBITDA normalized multiple.

Adjusting growth and margins to consensus estimates implies a normalized ~15.5x EV/EBITDA valuation for the business based on the current market price of $94.37/share. Which highlights the markets lagging recognition of the changes occurring at both the company and industry level.  

Notably, while our normalized valuation multiple appears elevated compared to the firms’ historical average, it may be justified given the industry/company’s strong secular growth and rational/consolidated marketplace in existence today. A highly relied upon company, ASML, has a normalized ~27.5x EV/EBITDA valuation, which is predicated on one crucial product. Whereas, ENTG has a wider breadth of products in a part of the industry that is evolving towards the dynamics in the equipment market. Which ultimately makes our normalized valuation possibly conservative.

8) Conclusion       

Entegris is a well-run semiconductor materials supplier company with compounding revenue growth, competitive advantages and value proposition with each transition to a new/smaller transistor node. The company’s strategy and market positioning enable it to create a unique platform unlike any in the semiconductor industry, which generally makes it more difficult for financial participants to ascertain its intrinsic value. This is observed in the discrepancy in the solid growth outlooks for semiconductors, the valuation of firms in the space further along in their market consolidation and Entegris’s underestimated growth outlook/value proposition by the investing public.

The company’s valuation should inflect upward from strong growth and expanding competitive/financial advantages as the industry evolves at the increasingly complex leading edge.

As of February 5, 2021, the name is trading at $94.37/share. With a Base Case valuation of $183.50/share, we believe there is ~95% upside, equating to ~14.5% annualized IRR over five years.  

Catalyst

·         Revenue growth in excess of historical norms/expectations

·         Industry node transitions

·         Accretive/Transformational M&A transactions

·         Design wins for new nodes

·         Growth and capital spending plans by leading edge semiconductors