Description

This morning I read an article in the Financial Times about Julian Robertson:  “He believed that an investment thesis should be able to be summed up in three bullet points on an index card.”  Here is the case for Lam:

• Leading participant in a growing oligopolistic industry

• Gaining market share

• Attractive valuation; especially for those with a long-term perspective

I will quickly flesh out each of these points below and am happy to respond to questions.  I am a generalist PM but have been following semiconductors and semi capital equipment stocks on and off for decades (see my VIC report on Varian Semiconductor from 2008).  Note that Lam has a June fiscal year.  Also, I will use GAAP earnings throughout though, from what I can tell, many on the sell-side use non-GAAP earnings; the difference is relatively minor.

A Growing Oligopolistic Industry

(For background I recommend two VIC reports from last year—one on Lam and the other on the iShares Semiconductor ETF, but really about the semi capital equipment industry.)

There are five critical suppliers of wafer fab equipment (WFE):  Applied Materials, ASML, KLA, Lam Research and Tokyo Electron. Furthermore, there are four key market segments.  ASML is dominant in lithography and KLA in process control and yield management.  AMAT, Tokyo Electron and Lam are the leading competitors in deposition and etching tools. Barriers to entry are extraordinarily high:  It is probable that these same five companies will be dominant a decade from now. Furthermore, as CEO Tim Archer noted at the Bernstein Conference in June, “Many of the companies in our industry, including Lam, have built incredible moats around our positions.”

In calendar 2022 the total WFE market should be between $90-100 billion. Supply chain issues and meaningful parts shortages (ironically including semiconductors) have impacted the industry, so some demand will spill over into 2023. Forecasts for the underlying semiconductor industry suggest L-T growth of 6-8%.  The equipment market should grow a little faster due to rising device complexity, larger die size, and a higher service component.  I project a doubling (9% CAGR) by 2030, although it is very likely that there will be one or possibly two down years over that time frame. Helping fuel this growth are a number of important governmental initiatives:  The recently passed CHIPS Act in the US, parallel legislation in Europe, and even a significant state initiative in India.

In sum, if one believes in underlying semiconductor growth and higher capital intensity over time, the WFE industry characteristics are very attractive.

Market Share Gains:

At its Investor Day in March, 2020, one of the first points made was that Lam had significantly outperformed the WFE in the years since the Novellus merger—with a CAGR of 16% from 2013-2019 versus 8% for the industry.  There, the company outlined a goal of gaining between 4 and 8 market share points by 2023 in both its key segments:  Etch and Deposition.  It appears to be on track, with clear market share gains in C21 (even though the memory segment, where Lam has long been strongest, grew more slowly last year than the logic and foundry segments). Dylan Patel, a close industry observer, calculated that Lam’s share of total spending grew from 9.3% in 2015 to 12.7% in 2021. 

A few thoughts on the sources of the market share growth:

—In recent years Lam has grown from a small position to becoming a very important supplier to Intel, which is in the midst of a surge in capital spending.  

—More broadly, Lam appears to be strongest in leading-edge fabs across the semiconductor market.  Consequently, it is a beneficiary of the arms race being fought by Taiwan Semiconductor, Samsung, and Intel in particular.  Historically, Lam has not been as strong in the foundry and logic areas and is clearly gaining ground here.

—FIner geometries require a shift in lithography to “High NA EUV”, or high numerical aperture extreme ultraviolet.  Working with ASML among others, Lam has developed a dry resist technology that opens up a new market for the company--one presently dominated by Tokyo Electron. On the most recent earnings call, Archer said that this represents $1.5 billion potential for Lam in 5 years (with, I believe, very significant growth beyond) as its technology will allow much greater utilization of extremely expensive ASM machines.  

Attractive Valuation:

I think that semi capital equipment companies have historically traded as if the industry is much more cyclical than it is in reality.  Yes, their customers are subject to the broad economic cycle (and increasingly so as semis are more ubiquitous) as well as industry cycles.  For example, Lam’s revenues declined 13% in FY19.  (The one down year since the GFC).  Undoubtedly, Lam will have another year—I estimate, with no precision at all, FY24—when revenues and earnings decline.  But I think there are a number of offsetting factors that long-term investors should consider and prize:

—Very high returns on invested capital.

—A highly variable cost structure.

—Stellar balance sheet.

—Extremely high barriers to entry.

—Some upward margin potential over time.

—Excellent capital management:  The vast majority of free cash is dedicated to share buybacks and a growing dividend.

My guess is that earnings rise in the current year, decline in FY24 to approximately current levels and then grow nicely from that point to the $55-60/share range by FY27.  On these assumptions, it is not hard to imagine the stock more than doubling over the next 4-5 years; a time frame I use as a taxable investor.  

Risks:

The cycle is obvious and I have no special insights.  I like the fact that some sell-side analysts (and I have only limited access to them) have already cut ratings and estimates based on an industry downturn.  My greater concern is China. The US government has expanded export restrictions to China in recent weeks, now blocking machines capable of producing advanced chips (14 nm or smaller).  In recent years over 30% of Lam’s business has been in China, although a significant percentage to firms like TSMC with fabs located there as opposed to native Chinese companies.  It’s unclear whether the new requirements include memory fabs.  

Catalyst

Not a catalyst driven investment.