Lam Research: Finding Opportunity Amidst Q4 FY 2024 Post Earnings Dip (NASDAQ:LRCX)

Lam Research (NASDAQ:LRCX) reported its fourth quarter FY 2024 earnings on July 31. The company exceeded its guidance and beat analysts’ estimates. The company’s first quarter of FY 2025 earnings guidance also beat analysts’ estimates. Seeking Alpha published an article post earnings that stated:

For the quarter in progress (first quarter FY 2025), Lam Research expects earnings per share ranging from $7.25 to $8.75 versus the estimate of $8.10. The semiconductor company anticipates revenue ranging from $3.75B to $4.35B versus the estimate of $4.04B.

Yet, the stock dropped 17% two days after the company reported earnings. The market was less than enthused with Lam’s fourth quarter FY 2024 report because of the potential rebound in Wafer Fabrication Equipment (“WFE”) spending that many expect will likely not take place in calendar year 2024. Additionally, virtually all hardware companies have declined over the last month due to fears that Artificial Intelligence may be in a bubble. The market also doesn’t like that the risk of a recession is still on the table. Last, Lam has company-specific risks that are large in investors’ minds and have caused some to steer clear of the stock.

However, experts still expect WFE spending to emerge from its downturn in 2025. Lam specializes in equipment that manufactures memory chips, and it already looks like the memory industry is potentially coming off a bottom. The company has invested heavily during this industry downturn in products that could help it gain market share during an upturn. I wrote about this stock in early July, and my thesis that the stock will rise due to a cyclical upturn over the next year remains intact.

This article will discuss how the company has prepared for a cyclical upturn and the company’s fourth quarter FY 2024 earnings report. It will also examine a few company risks, the valuation, and why the stock should remain on the growth investors’ buy list.

Preparing for a cyclical upturn

Semiconductor market forecasting company Yole Group projects the WFE market to stabilize in 2024 before meaningfully rebounding in 2025. After Lam’s fourth-quarter earnings, Seeking Alpha published an article that stated (emphasis added):

Analysts led by Joseph Moore noted that Lam reported a mixed June quarter supported by trailing edge spend from China, and September quarter guidance was in-line. They expect Lam to outperform in 2025 but see a tricky set-up in the second half of the year. The analysts said Lam will likely outperform the wafer fab equipment market in 2025 on the back of a recovery in NAND spending.

Analysts seem more worried about the company’s risks than the potential upside in 2025. Lam characterizes 2024 as an “investment year,” rather than a growth year. One of the initial comments that Chief Executive Officer (“CEO”) Timothy Archer made on the fourth quarter earnings call was (emphasis added):

As previously communicated, 2024 is a year of strategic investment for Lam, where we are prioritizing product development for key technology inflections, global R&D infrastructure close to our customers and digital transformation for operational efficiency at scale. We believe these investments will put Lam in a position to outperform as the industry moves into a period of multiyear WFE spending expansion.

Like the Yole Group assessment, Lam expects WFE spending to remain muted in calendar year 2024. The company forecasts WFE spending in the low to mid $90 billion range, boosted by Chinese customers spending more than initially projected and ramping demand for equipment to manufacture high-bandwidth memory (“HBM”). Lam also expects foundry logic, DRAM (Dynamic Random Access Memory), and NAND (non-volatile memory) to show growth in 2025 compared to 2024. The company also expects spending for older chip technology to flatline this calendar year.

Management projects AI to boost WFE spending in the calendar year 2025. CEO Archer said on the earnings call (Emphasis added):

Looking ahead to 2025, we see a positive environment for continued growth in WFE spending. The power of AI is a transformative business tool is still yet to be fully realized. Today, the focus on AI model training is driving strong demand for GPUs and HBM. However, as AI use cases expand, we believe inferencing at the etch will spur content growth of low-power DRAM and NAND storage in enterprise PCs and smartphones. Investments for AI-enabled etch devices play particularly well to Lam’s strengths. We anticipate that memory customers looking to scale capacity and lower bid cost will bias WFE spending toward technology upgrades of the installed base.

If CEO Archer’s statement about low-power DRAM and NAND memory customers being biased toward upgrades rather than new equipment purchases is authentic, that’s beneficial to Lam, as it means existing customers will purchase upgrades from Lam instead of potentially opening up bids for new equipment from competitors.

During the industry downturn, Lam has been busy. It has made multiple product investments to manufacture smaller, more powerful chips. DIRECTDRIVE is a new tool for plasma etching (dry etch) wafers that the company believes is the best in the industry. This new tool is excellent at removing thin slices of material from the surface of a semiconductor wafer, which is necessary for producing the newest advanced memory chip, 4F2 DRAM, a more compact and efficient version of DRAM compared to previous generations. On the earnings call, the CEO highlighted the importance of precise tools designed for 4F2 DRAM: “In 4F2 devices, the nature of the bit line placement requires precise etching of ultra-small high aspect ratio silicon structures to avoid device shorts or leakage.”

For generative AI’s Large Language Models (LLMs) to learn quickly and give speedy responses when queried, they require a memory that can retrieve large data sets rapidly. DRAM chips are a computer’s short-term memory, and the chip industry developed 4F2 DRAM to improve the performance of generative AI and other computing needs that require top-of-the-line short-term memory performance.

Another product that the company worked heavily on is a conductor etch (“CE”) for EUV (Extreme Ultraviolet) patterning for gate-all-around (“GAA”) and DRAM chips. These chips require precise material removal (etch placement) in the etching process. CEO Archer said about the CE tool (Emphasis added):

Conductor etch is becoming a critical enabler for EUV patterning for gate-all-around and DRAM due to the need to reduce etch placement error. For nodes below 2 nanometers, the requirement is for roughly 40% tighter control than at 5 nanometers. Our new conductor etch tool delivers a 30% reduction in feature roughness, which is one of the main contributors to etch placement error. In addition, we can achieve one to two orders of magnitude improvement in defectivity for a given EUV dose, further helping customers reduce the overall cost and improve the capability of the EUV patterning process.

Computer manufacturers use GAA as logic chips in applications requiring energy-efficient and speedy processors. One of the worries with LLMs is the massive power consumption in training and inferencing. GAA chips help lower the power consumption for generative AI uses while speeding up the data processing times.

On the date of the fourth quarter earnings call, the company announced several products for 3D NAND chips. Memory manufacturers construct next-gen 3D NAND by stacking tiny memory chips horizontally and vertically. Lam designed a product to help manufacturers stack memory chips wider and higher more logically. The name of this product is Cryo 3.0, the company’s third generation of cryogenic etch technology. Lam’s website states (emphasis added), “Since 2019, Lam has been the only company with cryo tools running in production for NAND manufacturing worldwide, boasting an install base of nearly 1,000 chambers.”

The second product that the company designed for NAND technology is a deposition product for gap-fill, a process that fills in the space between the memory chips stacked on top of each other. The industry has traditionally used polysilicon and tungsten to fill the gap. CEO Archer discusses the problem with polysilicon and tungsten gap filler and Lam’s solution in the earnings call:

Polysilicon and tungsten gap fill materials have typically been used to enable tier stacking in high-layer count NAND. Integration of these materials, however, has resulted in poor control of critical dimensions and overlay, negatively impacting yield and performance. Lam’s innovative PECVD-based pure carbon and gap-fill process provides an attractive alternative material. With the unique combination of high etch selectivity, superior mechanical properties, and simplified dry post-process removability, it also reduces the number of process steps required in some cases by approximately 50% compared to traditional approaches.

Computer manufacturers often use NAND memory chips as solid-state devices (SSDs) for long-term memory. Like DRAM, generative AI requires faster read/write times of information stored in memory, which requires 3D NAND. Another benefit of 3D NAND is it can store more data than regular NAND chips and has the endurance to withstand repeated reads and writes to the long-term storage device. Solid-state memory can wear down over time.

Lam is one of the most innovative companies in WFE. It heavily invests in research and development (R&D) to develop cutting-edge products to improve NAND, DRAM, and logic chip manufacturing. These investments should position it to potentially gain market share in WFE in the medium to long term.

Company fundamentals

The company referenced Lam Malaysia in the above image because it’s a new manufacturing facility built near its most significant Asian customers. It’s likely that the better this plant performs, the better the company’s long-term fundamentals. The company references that the Malaysian plant shipped 5000 chambers. A chamber is a sealed environment where Lam’s etching, deposition, or cleaning occurs. Sometimes, the company’s communication with investors will reference the number of chambers sold to customers. This number is important for the initial sales figures and CSBG (Customer Support Business Group) sales.

Lam manufactures highly specialized equipment that chipmakers integrate deeply into its manufacturing processes. Chipmakers don’t often switch suppliers because it would be expensive, time-consuming, and disrupt the chip manufacturing processes. So, the company enjoys a high switching cost moat. One of the benefits of that moat is that when the parts for its equipment wear down, customers only have one source for replacement parts and upgrades. CSBG provides Lam’s customers various support services, including spare parts, upgrades, technical training, software upgrades, and service calls. This business segment includes Reliant Systems, Lam’s specialty and legacy products provider. Specialty products are products outside of logic and memory chips. The company describes the types of products that manufacturers would use Reliance tools for on its website:

For Specialty Technologies, we are enabling customer roadmaps by developing new solutions, including critical steps. Specialty Technologies is a growing market, covering devices such as power, CMOS image sensors (CIS), radio frequency (RF), optoelectronic, analog and mixed-signal, and micro-electromechanical systems (MEMS).

The following table shows that CSBG generated $1.7 billion in the June quarter, 44% of its total revenue. The segment grew around 14% year over year. Although CSBG grows slower than systems revenue in an upcycle, this revenue source provides a stable revenue source in a downturn.

Lam’s fourth-quarter FY 2024 revenue increased by 20.71% to $3.87 billion, returning to quarterly year-over-year growth after multiple quarters of declining growth. This revenue was above the company’s fourth-quarter guidance mid-point and beat analysts’ estimates by $39.92 million.

Lam’s fourth-quarter earnings call and the accompanying investor presentation emphasize non-GAAP (Generally Accepted Accounting Principles) metrics. I’m not fond of some companies emphasizing non-GAAP numbers because non-GAAP metrics appear to be there primarily to exclude stock-based compensation (“SBC”), which makes GAAP profitability look worse. However, that’s not the case with Lam. Lam’s GAAP to non-GAAP reconciliation table shows that the company has multiple one-time or non-cash expenses that could make comparing profitability over different periods challenging. The company also lacks SBC.

EDC likely refers to Employee-Deferred Compensation plans. These plans are not the same as SBC or a pension plan. An EDC is a benefit that allows employees to defer a portion of their salary into a tax-deferred savings account. The company holds these funds until a specific date, usually retirement. Multiple factors can change an EDC valuation, but the likely culprit in this case was the increase in interest rates across 2022.

Transformation and restructuring are similar concepts with only a slight distinction. A transformation may change a company’s processes and technology, while a restructuring focuses more on changing the company’s organizational structure and the people running those new divisions. Amortization is a non-cash accounting technique for gradually writing off the cost of acquiring intangible items, such as trademarks, patents, copyrights, and goodwill, over time.

This article will primarily look at non-GAAP metrics to better understand how its core profitability is trending. June 2024 gross margin showed a 280 basis point (“bps”) year-over-year increase to 48.5%, at the top end of the company’s guidance. Chief Financial Officer (“CFO”) Douglas Bettinger said on the earnings call, “June quarter gross margin benefited from continued improvement in factory efficiencies, which largely offset the headwind we saw in customer mix.”

Operating expenses for the June quarter increased $99 million from the prior year’s June quarter to $689 million. Most of the operating expenses came from R&D. CFO Bettinger said the following on the earnings call (Emphasis added):

We continue to prioritize spending in research and development to extend our technology differentiation as well as expand our product portfolio. I’d just point out that more than 70% of our total operating expenses were concentrated in research and development.

June 2024 operating margin improved by 40 bps year over year to 30.7%, above the company’s guidance range. Non-GAAP diluted earnings-per-share (“EPS”) of $8.14, up 36.3% over the previous year’s June quarter. GAAP diluted EPS was $7.78, above the midpoint of guidance. The company reduced diluted shares from 134.39 million in the June quarter of 2023 to 131.11 million in the June 2024 quarter, increasing the EPS for existing shareholders.

The following chart shows that Lam has cash from operations (“CFO”) to sales of 31.21%, the second-highest WFE company I track. The higher the CFO-to-sales, the greater the potential for increasing free cash flow (“FCF”).

Lam generated $4.256 billion in trailing 12-month FCF in the June quarter.

The company ended the quarter with $5.84 billion in cash and cash equivalents against $4.47 billion in long-term debt and finance lease obligations. The company had $4.983 billion in total debt and $4.685 EBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization). Its debt-to-EBITDA ratio is 1.06, meaning it has enough EBITDA to repay its debt in a reasonable amount of time. The larger this number, the longer it will take for a company to repay its debt. When this number goes over 4.0, the company may be on the verge of financial distress.

The following table shows the company has a June 2024 inventory turn of 1.9, which means it sells its entire inventory 1.9 times per year. This number was 1.5 in June and September of 2023, around the bottom of the cycle. These low numbers indicate sales were likely poor in June and September of 2023. During the company’s last upcycle, in the March 2021 quarter, Lam had an inventory turn of 3.2. A higher inventory turnover often means the company has a better sales performance. Investors should monitor this number in future earnings reports. A rising inventory turnover potentially indicates the company is in an upcycle.

During the June 2024 quarter, the company returned capital to shareholders through $382 million in share repurchases and $261 million in dividends. In FY 2024, the company returned $3.7 billion, or 88% of FCF, to shareholders. The company’s long-term plan is to return 75% to 100% of FCF to shareholders.

Risks

Some consider Lam Research an AI infrastructure stock. Over the last month or two, sentiment toward AI infrastructure companies has declined. Sequoia published an article in 2023 titled AI’s $200B Question, later updated in June 2024 to AI’s $600B Question. These articles discussed a growing gap between companies investing in AI infrastructure and the lack of enough AI applications and services to provide value to customers. This growing gap has created fear among some that the AI infrastructure market may be in a bubble. The following chart shows that many of the most significant AI hardware infrastructure stocks have declined over the last month. A potential perception that the AI infrastructure market may be in a bubble could hurt the stock’s near- to medium-term performance.

Another risk that investors should consider is the concentration of revenue from China. The chart below shows China made up 39% of the company’s June quarter total revenue, which is $1.5 billion.

In 2022 and 2023, the U.S. instituted export controls on sending advanced semiconductor manufacturing equipment to China. Today, China is buying as much legacy equipment as it can before the U.S. government potentially decides to create even more restrictive rules on semiconductor manufacturing equipment sales to China, which the U.S. government has considered. Additionally, Lam’s customers in China might eventually switch away from the company’s products the longer these export controls remain in place. The U.S. has spent much political capital getting foreign countries like the Netherlands on board with restricting companies like ASML Holdings (ASML) from sending equipment to China. However, how long can the U.S. government continue getting foreign governments to follow this policy? For instance, Tokyo Electron (OTCPK:TOELY)(OTCPK:TOELF), a direct Lam competitor from Japan, may be exempt from potential new upcoming export restrictions.

Valuation

Lam Research’s price-to-sales (P/S) ratio is 7.155 at the end of the June quarter, well above its five- and ten-year median, suggesting overvaluation. When I wrote my last article on Lam Research, I ignored how much the stock traded above its P/S median in favor of what its forward price-to-earnings implied about what the stock should trade at. This time, I wanted to examine why the stock trades so much above its median P/S ratio.

Before I go into this analysis, it would be helpful for you to understand that I am making a rather large and straightforward assumption that this stock will return to trading at its median five-year P/S ratio. This assumption is simplistic because not all stocks immediately return to their median, and the five-year median is not static. It has risen over time and is well above the ten-year median. It is possible that when Lam matches its five-year median again, it may be at a higher P/S ratio than 4.994 because the market may award the company higher P/S valuations over time. After all, the value of the equipment and services it sells may increase in the market’s eyes. Let’s look at what Lam’s five-year median P/S ratio implies about the company’s sales.

For Lam Research to reach a five-year median PS ratio of 4.994, the price needs to drop to $534.50, or the company’s sales will need to grow into that valuation. For instance, its latest trailing 12-month sales per share is $107.03. The company would need to achieve trailing 12-month sales per share of $161.77 to match the median of 4.994 without the stock price declining from its closing price on August 9, 2024 of $807.90. Multiplying $161.77 times 131,112,000 diluted shares the company recorded at the end of the June quarter equals $21.21 billion in TTM revenue.

The chart below shows that at the end of its fourth quarter FY 2024, it had an annual revenue of 14.91 billion. The highest TTM revenue it had ever achieved was in the December quarter of 2023, where the number reached $19.05 billion. So, Lam has a long way to go to reach $21 billion in TTM sales. Although the stock beat revenue estimates in the fourth quarter, some who value the stock using metrics like the P/S ratio see that sales must increase 42% to reach the five-year P/S median at the current stock price — a potential reason the stock slid after earnings.

On a TTM basis, Lam’s revenue appears to have already begun to rebound. There are several things that investors must consider here. Lam is a cyclical company. Cyclical companies will often look the most highly valued on a P/S basis at the bottom of a downturn because sales are depressed from the downturn. At the same time, investors will drive up the price in anticipation of a rebound in sales. A cyclical stock like Lam can sometimes have sales rebound faster than the market expects, which is why some investors buy in anticipation of a rebound instead of waiting to see proof that a sales rebound has taken place.

Now, let’s look at what analysts expect. The following table shows analysts’ revenue estimates for Lam over the next three fiscal years. If you believe analysts’ revenue estimates, the company could reach $21 billion in sales within two to three calendar years.

Right now, the market is erring on the side of caution because it’s not outside the realm of possibility that the U.S. market will have a recession. According to the Estrella and Mishkin indicator, the recession probability is 56.29%.

However, because analysts are cautious, they may be understating Lam’s revenue estimates if the economy avoids recession and achieves a soft landing. In that case, if the Federal Reserve starts lowering interest rates in September 2024, by FY 2025, FY 2026, and FY 2027, actual revenue may be far higher than current analysts’ estimates for those years. The company could achieve $21 billion in TTM revenue, far faster than the market expects. Let’s look at the current P/E ratio.

Lam’s price-to-earnings (P/E) is 27.83, above its five- and ten-year median of 20.71 and 18.89, respectively.

Lam’s five-year median P/E implies a stock price of $563.11. However, the price doesn’t necessarily need to decline to that level to reach its five-year median EPS. The company can also get to its five-year median by growing into its current valuation. By raising its TTM diluted net income per share (same as EPS) 34.51% from the $29.00 it recorded in its fourth quarter FY 2024 report to $39.01, the stock would reach its five-year median P/E of 20.71 at $807.90. This EPS growth is feasible. So, although its P/E is below its median, the market may not necessarily overvalue the stock.

The memory chip industry was last in a down market from 2018 to 2019. The following chart shows that once the business cycle turned upward, EPS quarterly growth peaked at nearly 100% year-over-year in about a year. The lesson is that earnings growth can ramp rapidly when the business cycle turns. If it is true that the memory chip industry has already bottomed, Lam may not look overvalued at these current prices a year from now. Today’s prices may eventually look like a good entry point using hindsight a year or two from now.

The following table shows Lam’s forward P/E ratio and EPS growth estimates. One simple rule that investors can use to determine whether the market overvalues or undervalues a stock is observing a company’s forward P/E and EPS growth estimates for a given year. When the two numbers match, the company would have a forward Price-to-Earnings-to-Growth (“PEG”) ratio of one, which many consider fairly valued. The market may undervalue the company’s potential EPS growth rate when the consensus EPS estimated growth rates exceed the forward P/E. Alternatively, if the consensus EPS estimate growth rate is lower than the forward P/E, the market may overvalue the stock. In FY 2026, when Lam should be well within a rebounding memory market, the stock looks undervalued with a one-year forward PEG ratio of 0.62. If it traded at a forward P/E equal to its EPS growth estimate in FY 2026, the stock price would be $1298.68, up 60.15% over its August 12 closing price. Be aware that analysts can underestimate a cyclical stock’s EPS growth estimates coming out of a downturn, and the actual EPS growth rate may be much higher.

Let’s examine the company’s reverse discount cash flow analysis (“DCF”). I set the terminal growth rate, which is the rate at which I expect the company’s FCF to grow in perpetuity, at 3%. This number shows an expectation of the company to increase revenue and FCF growth rate above the U.S. gross domestic product growth rate over the long haul. I use a discount rate of 10% because I consider Lam’s risk normal. A higher discount rate would mean I assume the risk is higher than the average stock. A lower discount rate would mean that I believe the risk was lower than average. I also use the company’s levered FCF in the following reverse DCF.

Lam Research Reverse DCF

FCF (Trailing 12 months) margin	28.55%
The second quarter of FY 2024 reported Free Cash Flow TTM (Trailing 12 months in millions)	$4,256
Terminal growth rate	3%
Discount Rate	10%
Years 1 – 10 growth rate	10%
Current Stock Price (August 12, 2024)	$810.89
Terminal FCF value	$11.474 billion
Discounted Terminal Value	$63.196 billion

Click to enlarge

At an average FCF margin of 28.55% over the next ten years, the company would need to achieve an annual revenue growth rate of 10% over the next ten years to justify the August 12 closing price. However, the company’s median FCF margin over the last ten years was 21.83%. Assuming an average FCF margin of 22% over the next ten years, the implied revenue growth rate is 13.5% to justify the current stock price. Lam has achieved a 17% annual revenue growth rate over the last ten years, from $3.6 billion in FY 2013 to $17.43 billion in FY 2023. The big question is how long the company can maintain a 17% revenue growth rate with competition increasing in its primary deposition and etch markets. The company must navigate potential market share loss to competitors like Applied Materials (AMAT) and Tokyo Electron. The company may also face pressure to lower its prices to compete. Alternatively, Lam has invested in unique technology enabling its customers to manufacture advanced logic chips like Gate-All-Around transistors and DRAM chips. The company has potentially positioned itself to gain market share within the markets it competes in. Suppose I assume a few percentage points of growth loss and use a 15% revenue growth rate at an average FCF margin of 22% over the next ten years; the intrinsic value would be $901.26, up 11% over the August 12 stock price.

Although Lam may look overvalued based on several valuation metrics and Seeking Alpha Quant rates the valuation a D-, investors should be careful about labeling a cyclical stock overvalued at the potential bottom of a cycle because as the cycle turns up, the company could go from looking overvalued to looking undervalued in a short period as fundamentals improve.

Lam remains a buy

Although this company faces risks and challenges, Lam has invested significant amounts in R&D to put itself in a position to benefit from the increasing adoption of 3D NAND chips, faster DRAM chips, and high-bandwidth memory, essential for high-performance computing and generative AI. The company could gain share throughout a potential upcycle with some of its newest products. If you are a growth investor interested in investing in AI infrastructure companies, consider an investment in Lam. The stock remains a buy.