On the surface, when I think about my investment strategy, I’m looking for companies with very strong competitive advantages, little competition, product differentiation, strong leadership, operate in a growing market, trade at reasonable valuations and those that provide a high amount of value to their customers.
Diving one level deeper, I’m looking for companies with the ability to continue increasing their value proposition over time. That there’s some underweighted characteristic like network effects that will drive returns over an extended period of time.
I think the large semiconductor capital equipment providers meet the first criteria, and likely don’t fit the second criteria. Meaning: I like the space’s investment potential a lot, but I think it’s unlikely to return unbelievable returns. Strong but not incredible.
Background on the Semicap Space
The value chain of semiconductors (starting from the end markets) begins with the purchasers of semiconductors. Major end markets include data centers, devices, automotive, industrials, and then a host of use cases for analog chips in toothbrushes, light switches, etc.
The major of advanced semiconductors come from fabless semi firms; meaning they outsource the manufacturing of their chips to TSMC, Samsung, or Intel in the future. Intel, Samsung, the analog semi companies, and memory semi companies mostly manufacture their own chips.
Finally, we get to the equipment that goes into the foundries. This is the semiconductor capital equipment space.
Since semiconductor manufacturing is the most complex process on the planet, the machines used in this process are nearly equally complex. The machines are so complex that most of the steps in the process have one or two companies that control 50%+ of the market, the most complex of which ASML has 100% market share in the most advanced machines.
This gives us a wonderful case study on the defining variables of competition in industries with high complexity. We can look here at the most complex industry on earth and see how technical differentiation is by far the most important variable in the space, and only a few companies are able to meet the technical standards to manufacture these components.
What makes the space so appealing?
If we go back to my original list of what makes a great company? It’s competitive advantages, low competition, growing market, strong leadership, high profitability.
This industry has 5 major players which each dominate in specific verticals. They are highly profitable, have gigantic moats, and return the majority of their FCF to shareholders through buybacks and dividends.
Additionally, for the extra long-term focused investors, these companies are heavily cyclical which means they provide wonderful buying opportunities when the cycles turn for the worst. Now this assumes the long-term success of these companies’ and management’s ability to navigate these cycles; and we have multiple decades of the proof of ability to manage these cycles.
Risks?
Of course, these companies come with risks. The most common risk to a company is competition; we live in a world where margins get eaten by competition. These companies face much less risk to competition as evidenced by their consistent market share in their verticals and the fact that few companies exist for competition. Additionally, building a company from scratch is near impossible without a new, disruptive way to analyze these companies.
The risks as I see them are as follows:
Ignorance Risk - the space is unbelievably complex and the number of investors who can truly understand the nuances of the technology and industry is probably under 100 in the world. Being long-term focused helps remove some of this complexity predicting cyclicality but some level of diversification across semicap is wise in my eyes.
China Risk - this is the largest risk outside of ignorance. China makes up 30% of sales for these companies on average. China is also investing heavily to build out their semi industry. They will copy the semicap company’s IP wherever they can, give this to domestic companies, flood them will capital, commoditize the tooling, and that could remove a majority of the semicap China revenue over a decade. The top-end tools likely won’t be replicated but we don’t know that. Additionally, any end markets like analog semis that are replicated by China remove revenue from Western countries which then removes revenue flow to semicap companies (assuming China builds out their semicap industry).
Risk of Semiconductor Architectural Changes - The flat innovation risk is unlikely in my eyes. We’ve seen consistent innovation from the major companies and very very few new semicap companies in the last two decades. The greater risk goes back to not fully understanding how the importance of each piece of technology rises and falls based on semiconductor architecture trends. As semi architecture changes, so does the importance of different tools in the stack.
Risk of Infrastructure Buildout - We’re seeing a massive reshoring of semiconductor manufacturing across the world. This brings the risk that we’re pulling forward large amounts of revenue over the next 2-3 years that will result in 8-10 years of flat or negative growth.
Risk of Semicap Capital Intensity Decreasing - Right now, a large percentage of the gains in leading edge logic semiconductors are coming from tooling improvements, specifically lithography. However, if this changes, and less of the improvement gains come from these tools, their pricing power and margins will naturally decrease.
Summary & Key Questions
This was a very quick breakdown of my investment thoughts on the space. I think they are wonderful companies with huge competitive advantages. Understanding the businesses and how they are affected by trends is challenging. The key questions I’m left with in the space are:
What does a valuation process look like when earnings and profits are so cyclical? What data can we look at historically to reasonably analyze this?
In worst case China scenarios, what does this industry look like?
If a fab infrastructure buildout pulls revenue forward, when do we start seeing revenue rise again?
What IRR should justify the risks in this space around China?
Should you diversify by owning shares in the 4 major western semicap companies? This decreases some risk around which tooling gains momentum.
What does the long-term growth look like of the semicap industry?
What do I need to model for financial modeling to understand the business? best-worst case scenarios on: semi industry size, semicap intensity, revenue of 0-30% from china, what cycles may look like.
Back to studying.