The core of the AI semiconductor bubble | How far will Tokyo Electron (8035) grow in 2027?

[Summary]

Tokyo Electron (8035) is a company at the core of the AI semiconductor market.

However, the essence of the company is that it is not just a "semiconductor manufacturing equipment stock."

What's really important is

“A company that provides manufacturing capabilities for mass production of AI semiconductors”

It is to be.

In the financial results for the fiscal year ending March 2026 announced on April 30, 2026, sales were 2,443.5 billion yen, operating income was 624.9 billion yen, and net income attributable to owners of parent was 574.4 billion yen.

Sales and net income are at record high levels.

Furthermore, the forecast for the first half of the fiscal year ending March 2027 is as follows:

We are expecting.

Company materials say that in the first half of FY2027, demand for AI servers will be the driving force, and the company expects record high sales, gross profit, and operating profit.

Importantly, Tokyo Electron's strength lies in more than just temporary equipment sales.

The company has

• World's largest installed base of approximately 96,000 units
• Field solution sales expanded to 626 billion yen in FY2026
• 100% share among EUV compatible coaters/developers *A group of equipment that crosses etching, film formation, cleaning, coating and development *Lock-in structure jointly developed with customers

There is.

For this reason, Tokyo Electron is not a company riding on the AI semiconductor boom.

“A company that holds the manufacturing bottleneck in the AI semiconductor hegemony structure”

should be viewed as such.

First, the conclusion

When looking at Tokyo Electron in 2027, the most important thing is not the short-term ups and downs of the stock price.

What you should see is

“Will the profit structure shift from relying on mature nodes in China to being driven by AI, HBM, and advanced logic?”

It is.

One of the major factors that supported the company's performance from 2024 to 2025 was its investment in mature nodes in China.

However, from late 2026 to 2027, the company expects shipments to increase mainly in DRAM and advanced logic.

Furthermore, in company materials,

• Significant increase in investment in DRAM and advanced logic
• Etching sales YoY25% or more in FY2027 *FY2027 sales YoY 60% or more with advanced packaging
• First-half results expanded due to demand for AI servers

is shown.

In other words, Tokyo Electron in 2027 will be

“Semiconductor market recovery stocks”

rather than

“A company that monetizes the increasing difficulty of manufacturing AI semiconductors”

It is.

GPU is not the only thing that matters for AI infrastructure

In the current stock market, NVIDIA's GPUs are attracting attention as the core of AI infrastructure.

This is natural.

This is because the performance of generated AI is greatly influenced by the GPU and AI accelerator.

But what investors should look at next:

“The very ability to create GPUs”

It is.

A true supply chain for AI infrastructure can be organized as follows:

AIdemand
↓
GPUKey point
↓
HBM
↓
Key point
↓
Key point
↓
semiconductorsKey point
↓
Key point

As AI demand increases, not only GPUs but also HBM, DRAM, logic, and advanced packaging will be required.

As a result, the manufacturing process becomes more complex.

What is needed here is

• Film formation
• Coating and developing
• Etching *Washing
• Bonding *Test

It is.

Tokyo Electron covers a wide range of these important processes.

In other words, the company

“A company that solves the supply constraints of AI semiconductors”

It is.

What the HBM and CoWoS shortages mean

In the AI semiconductor market, the lack of HBM and lack of advanced packaging capabilities are major points of contention.

HBM is high-bandwidth memory essential for high-performance GPUs.

However, HBM is more difficult to manufacture than regular DRAM, and stacking, bonding, inspection, and yield management are important.

Advanced packaging such as CoWoS is not just a post-process.

This is the core process that affects the performance of AI semiconductors.

Tokyo Electron says in its company materials that it expects sales of Advanced Packaging to be approximately 200 billion yen in FY2026, and sales YoY of more than 60% in FY2027.

This is important.

The market tends to view semiconductor manufacturing equipment as "front-end process stocks," but in the AI ​​era, the line between front-end and back-end processes will become thinner.

The more value-added advanced packaging becomes, the more profit opportunities will expand for Tokyo Electron.

Business Model 1｜Recurring structure that does not end with selling

Tokyo Electron's greatest strength is that we don't just sell the equipment.

The company's corporate profile lists approximately 96,000 units as one of the world's largest installed bases.

Additionally, it is estimated that 4,000 to 6,000 new devices are shipped to customer factories each year.

This huge group of operating equipment forms the basis of our field solution business.

Semiconductor factories operate 24 hours a day, 365 days a year.

Equipment outage is a big loss for customers.

Therefore, even after installing the equipment,

• Maintenance *Parts replacement
• Modification *Upgrade *Remote support
• Improved utilization rate

Demand continues.

Field solution sales in FY2026 were 626 billion yen.

This is approximately 26% of the company's total sales of 2,443.5 billion yen.

6,260hundred million yen ÷ 24,435≒ 25.6%

This is an important pillar supporting Tokyo Electron's earnings stability.

There is a cycle in new investment in semiconductor equipment.

However, once installed, the demand for maintenance, modification, and parts remains.

For this reason, field solutions provide a cushion that supports profits even during downturns in the semiconductor cycle.

Business Model 2｜Sales leverage for high value-added equipment

Tokyo Electron is not simply a mass-producing manufacturer.

The company's strengths lie in equipment development, process proposals, and mass production start-up support that meet customers' technical requirements.

The official company profile also explains that the company's strengths include incorporating customers' technological needs from the development stage and providing high-quality products in a timely manner.

In other words, the company

“Company that makes equipment”

At the same time,

“A company that creates semiconductor processes for customers”

It is.

With this business model, profits tend to increase significantly when sales increase.

The reason is that when the ratio of high value-added equipment increases, both unit price and profit margin tend to increase.

Operating profit margin for FY2026 was 25.6%.

The forecast for the first half of FY2027 is an operating profit margin of 27.5%.

This indicates not only an increase in sales, but also the possibility of improving the product mix and creating higher added value.

Business Model 3 | Joint development platform that locks in customers

Cutting-edge semiconductors cannot be created by purchasing general-purpose equipment.

For 2nm, GAA, HBM, and advanced packaging, equipment requirements vary greatly depending on the customer's process.

For this reason, equipment manufacturers collaborate with customers on development from an early stage.

Tokyo Electron's corporate brochure also explains that the company understands customers' technological needs and issues from the development stage and optimizes related processes.

Once customers enter this joint development line, they cannot easily switch equipment manufacturers.

The reason is

• Process conditions
• Yield data
• Equipment recipe
• Engineering know-how
• Verification of mass production launch

This is because it is deeply connected to the device.

This is Tokyo Electron's lock-in structure.

We don't just sell products,

“Integrated into the customer's mass production process”

That is our strength.

AI power industrialization raises technological barriers

The essence of the AI semiconductor market is not just competition for GPU sales.

As AI becomes a huge power industry, semiconductors will

• Low power consumption
• High density
• Faster
• High reliability
• Measures against fever
• Yield improvement

is required.

This change is a tailwind for Tokyo Electron.

Because the higher the manufacturing difficulty,

• High precision film formation
• Fine etching
• Extreme cleaning *GAA compatible
• Joining for HBM *Advanced packaging

because it becomes more important.

Company materials indicate the HARC process, wiring process, and GAA process as priority areas for etching.

The company also plans to capture significant investment opportunities in DRAM and advanced logic.

In other words,

“The more AI evolves, the more difficult the technology required for Tokyo Electron's equipment will become.”

This is the structure.

AI semiconductors cannot be mass-produced using ASML alone

When many investors hear about cutting-edge semiconductors, they think of ASML.

EUV exposure equipment is certainly essential.

However, semiconductors cannot be made by exposure alone.

There are many steps before and after exposure, including coating, development, film formation, etching, and cleaning.

Tokyo Electron explains in its company information as follows:

• Covers four important continuous processes: film formation, coating and development, etching, and cleaning. *Many products rank first or second in each market *The world's most advanced semiconductor chips are processed by our company's equipment at some point during manufacturing.
• Coater/developer that can be integrated in-line with EUV exposure equipment has 100% market share

This point is important.

Mass production of AI semiconductors cannot be achieved by ASML alone, TSMC alone, or NVIDIA alone.

Mass production will only be possible if equipment manufacturers like Tokyo Electron solve bottlenecks in each process.

Tokyo Electron's strengths by equipment

Tokyo Electron's strength lies not in a single device but in its device portfolio.

What is especially important is that they not only exist separately, but also combine in the customer's process development.

As semiconductors become more complex, optimization of multiple processes becomes more important than just the performance of a single device.

This "cross-processing ability" is Tokyo Electron's competitive strength.

From China's mature node to cutting-edge AI investment

From 2024 to 2025, one of the factors that supported Tokyo Electron's performance was sales to China.

Investment in mature nodes is strong in China, and there was a time when China accounted for a high proportion of the company's sales.

However, what will be important as we move toward 2027 is that the quality of our revenue sources will change.

According to company data, sales of new SPE equipment in the first half of FY2027 are expected to increase by 41% compared to the same period last year.

Furthermore, shipments are expected to increase mainly in DRAM and advanced logic until the second half of 2026.

This is Tokyo Electron's growth engine.

“Special demand for mature nodes in China”

From,

“DRAM, HBM, and advanced logic investment for AI servers”

This indicates the possibility of moving to

The key point in 2027 is whether this transformation will actually progress.

How to look at valuation

When looking at Tokyo Electron's valuation, a simple PER is not enough.

What is important is the sustainability of EPS.

Semiconductor manufacturing equipment stocks are cycle stocks.

If capital investment increases, profits will increase significantly, but if investment stops, profits will decline.

However, in the case of Tokyo Electron,

*Field solution sales

• Huge installed base
• High market share in advanced processes *Long-term investment in AI
• Joint development with customers

supports the sustainability of EPS.

In other words, what the market should be looking at is

“Will 2027 profits be a one-off or the beginning of the next profit stage?”

It is.

If the latter is determined to be the case, Tokyo Electron will be more likely to be reevaluated as an AI infrastructure stock rather than just a cycle stock.

Growth market in 2027

There are three growth markets that Tokyo Electron has the potential to capture in 2027:

AI GPUmarket
↓
Key point
Key point

HBMmarket
↓
DRAMinvestment
Key point

Key point
↓
Key point
Key point
Key point

All three are directly connected to AI data center investment.

Unlike the semiconductor cycle of the past, which was dominated by PCs and smartphones, the AI ​​cycle is connected to data centers, the cloud, national policy, and power infrastructure.

Therefore, even though there may be peaks and troughs in the cycle, the depth of the investment theme may be greater than in the past.

Maximum risk

Tokyo Electron also has risks.

Of particular note is the anticipated advance in AI investment.

If the stock price rises too much first, even a slight failure to reach guidance will result in a large sell-off.

Therefore, investors should not only focus on themes;

• FY2027 first half results
• H2 guidance *Field solution sales
• DRAM/advanced logic ratio
• Advanced packaging sales

you need to check.

Checkpoints that investors should look at

If you look at Tokyo Electron as a theme for 2027, the following five points are important.

Operating profit margin is especially important.

The forecast for the first half of FY2027 is an operating profit margin of 27.5%.

If this level can be maintained, the market may reconsider Tokyo Electron as a "high-profit AI infrastructure stock" rather than a "cycle stock."

Final conclusion

Tokyo Electron's 2027 is not just a recovery in the semiconductor cycle.

The essence is

“An era where the ability to mass-produce AI semiconductors is insufficient”

The company is a company that solves these bottlenecks.

After GPU shortages come HBM shortages, advanced packaging shortages, and manufacturing capacity shortages.

Tokyo Electron is on the side to solve this manufacturing capacity shortage.

Moreover, the company has

• Installed base of approximately 96,000 units *Field solution sales of 626 billion yen
• 100% share of EUV compatible coater/developer
• Cross-process capabilities including etching, deposition, and cleaning *Lock-in through joint development with customers

There is.

Therefore, Tokyo Electron in 2027 will be

“Behind the scenes of AI semiconductor hegemony, a company that controls manufacturing capabilities”

may be reevaluated as

What investors should be looking at is not the immediate rise and fall of stock prices.

What you should see is

“How will AI demand translate into the company's sales, profit margins, and field solution revenue?”

It is.

The more this transformation is confirmed in numbers, the more Tokyo Electron will strengthen its presence as a favorite in AI infrastructure.

source

• Tokyo Electron "Fiscal Year Ending March 2026 Financial Results Presentation Materials" April 30, 2026
• Tokyo Electron "Financial Results/Financial Results Briefing Materials"
• Tokyo Electron "Investor Information"
• Tokyo Electron Corporate Profile 2025
• Tokyo Electron "Medium-term Management Plan"
• Tokyo Electron "Deposition TELINDY Series"
• Tokyo Electron "Etching Episode UL Series"