What we are watching for in the upcoming earnings results …
Mr. Market can be very fickle. Microsoft has been a leader in AI for years, but now suddenly everyone is hot on the story. Same with Nvidia. But what about AMD – when will AMD be called the next big AI stock? On management’s end, that’s likely what this earnings call will be pointed toward.
AMD is releasing a chip that combines a CPU, GPU and memory into a single integrated design called the MI300. The company plans to launch the MI300 in the second half of this year although it’s not expected to contribute to revenue until 2024. The MI250 currently powers the world’s highest performing supercomputer.
Analyst notes on the data center are mixed with some expecting a slowdown for AMD and Intel (see below). For our purposes, AMD’s release of the 5nm Genoa Zen 4 in Q4 along with the strong performance of Milan should help to hedge a broader data center slowdown. Bergamo is also expected this year, which offers cloud optimized capabilities.
Management was quite clear on the last earnings call to expect a double digit decline sequentially for the data center but to also expect YoY growth from 2022 to 2023 with a stronger second half of the year. Embedded is expected to have a positive impact whereas PCs are expected to bottom in the March quarter.
Microsoft came in better-than-expected on PCs, which has some read through for AMD. Per our Microsoft write-up: “More Personal Computing declined by (9%) YoY to $13.3 billion and was better than the mid-point of the management guidance for a decline of (16.7%). The PC segment revenue witnessed better than expected results in all businesses. The guidance for the next quarter represents a YoY decline of (5.6%).”
Per Microsoft’s CFO: “We saw better-than-expected PC demand, as noted earlier, particularly in the commercial segment, which has higher revenue per license, although results continue to be negatively impacted by elevated channel inventory levels.”
Intel’s guidance in the most recent earnings report also reflects the PC downturn being near an end.
Financials:
Estimated Revenue and EPS:
AMD has seen downward revisions on revenue and EPS. At the same time, analysts are increasing their price targets. We’ve been talking a lot about a rebound for semis on our quarterly webinars both in Q1 and Q2. There’s still another weak quarter to go for AMD with Q2 being the anticipated low for the top line.
EPS is similar with growth returning in Q3 into the foreseeable future:
Margins:
Adjusted gross margin guide of 50% for adjusted gross profit of $2.65B, which is in line with previous quarters
Guided for $1.6B in operating expenses for an adjusted operating margin of 19.8%. This is lower than previous quarters due to lower revenue
Cash Flow:
Operating cash flow margin of 10% last quarter was lower than previous quarters when revenue was higher
Free cash flow margin of 7.9% last quarter was lower than previous quarters when revenue was higher
AMD has $5.9B in cash and $2.5B in debt
Noteworthy:
Last quarter, AMD missed with guidance for the March quarter at $5.3B compared to $5.52B expected. In addition to this, analysts were concerned about the lower adjusted operating margin of 23% versus 27% a year ago and lower adjusted net margin of 19.6% versus 23% a year ago.
Regarding margins, these are expected to return to normal once PCs rebound.
Management guided for data center to be down “double digits” sequentially per the last earnings call. Per AMD’s CFO Jean Hu: “Year-over-year Data Center and Embedded segment revenue are expected to grow, offset by lower Client and Gaming segment revenue.Sequentially, Embedded segment revenue is expected to increase. Client and Gaming segment revenue are expected to decline largely consistent with seasonality. Data Center segment revenue is expected to decline due to elevated levels of inventory with some cloud customers.”
Per the AMD’s CEO Lisa Su:
“Sure, Ross. So let's see. We said the Client and Gaming segments would be seasonal. So you would expect that the Data Center would be more than seasonal. So maybe to help you size that, think about the Data Center sequential drop as double digit, whereas the Client and the Gaming segments are more like single digit, if that helps.”
Also, per last quarter’s call, this quarter is expected to be the bottom for PCs.
Headlines:
AMD introduced the new Ryzen Z1 Series processors for handheld PC gaming consoles. AMD is partnering with Asus to launch the first Ryzen Z1 Series device with the Asus ROG Ally, a premium handheld PC console, featuring up to a Ryzen Z1 Extreme processor.
Analyst Notes:
“March notebook shipments were up 41% month-over-month, 18% above Citi's expectation, driven by new product launches and pull-in demand at quarter end, Citi analyst Christopher Danely the analyst tells investors in a research note. The firm says overall Q1 shipments were down 9% quarter-over-quarter, well above seasonality and Citi's forecast of down 15%. It reiterates Neutral ratings on Intel (INTC) and AMD (AMD), expecting the data center downturn to negatively impact both companies.”
“Morgan Stanley analyst Joseph Moore raised the firm's price target on AMD to $102 from $87 and keeps an Overweight rating on the shares ahead of the company's report scheduled for Tuesday, May 2. The setup for the quarter is "interesting," as the stock has appreciated 40% year-to-date into "an obvious negative event at least as far as data center is concerned" and 90% of the firm's conversations are "appearing to lean negative," the analyst tells investors. However, AMD is pointing to relevant exposure to AI "for really the first time in the company's history around MI300," and "anyone with AI exposure has rallied materially," the analyst added.”
“Cleveland Research is lowering its estimates for AMD below consensus for Q1 and Q2 on further weakness in near-term Server demand and notes that its 2023 forecasts are "moving slightly lower" given its expectations for total server units to be down year-over-year.”
Buy/Sell Plan
Please check the forum tomorrow for the Buy/Sell plan and chart update from Knox.
Taiwan Semiconductor Manufacturing Company’s Q1 2023 revenue declined by (4.8%) YoY to $16.72 billion. The revenue came at the lower end of the management guidance of $16.7 billion and $17.5 billion. It missed analyst revenue estimates by 1.6%.
EPADR (Earnings per American Depository Receipt) came at $1.31 and beat estimates by $0.12 (9.7% beat). The Q2 revenue guidance was lower. However, there is scope for a better bottom line than expected in H2 due to the better utilization rate and cost controls.
TSM will be able to better withstand the macro challenges and the company’s long-term growth opportunities are still strong due to the leadership position in the advanced nodes. The recent generative AI trend is another tailwind for the company.
Financials:
Revenue declined by (4.8%) YoY to $16.72 billion and missed estimates by 1.6%. The company released its monthly revenue figures earlier this month. The softness in the revenue was expected and we have covered in our pre-ER here.
The guidance for the next quarter is $15.2 billion to $16 billion based on the exchange rate of $1= NT$30.4. This represents a YoY decline (14.1%) at the mid-point of the guidance. It missed estimates by 4% as the inventory adjustments are expected to continue due to the challenging macro environment and slowing end-market demand.
Wendell Huang, VP and CFO of TSMC, said, “Our first quarter business was impacted by weakening macroeconomic conditions and softening end market demand, which led customers to adjust their demand accordingly.” He further added, “Moving into second quarter 2023, we expect our business to continue to be impacted by customers’ further inventory adjustment.”
Gross profit declined by (3.6%) YoY to $9.42 billion. The gross margin was 56.3% compared to 55.6% in the same period last year. The gross margin was expected to be lower due to lower capacity utilization. The gross margin was higher than the management guidance of 53.5% to 55.5%. The management guidance for Q2 is 52% to 54%. The next quarter’s gross margin will be negatively impacted by lower capacity utilization and higher electricity costs in Taiwan. The higher electricity costs will negatively impact the gross margins by 0.60% in Q2 and by about 0.50% for the full year 2023.
The operating income declined by (5%) YoY to $7.6 billion. The operating margin was 45.5% compared to 45.6% in the same period last year. The operating margin was also higher than the management guidance of 41.5% to 43.5%. The company’s cost control efforts led to a reduction in operating expenses and improvement in the margins. Wendell Huang, CFO of the company, said in the earnings call, “Total operating expenses accounted for 10.8% of net revenue, which is lower than the 12% implied in our first quarter guidance mainly due to stringent expense control and lower employee profit sharing.mainly due to stringent expense control and lower employee profit sharing. The management guidance for Q2 is 39.5% to 41.5%.
We had also highlighted in our pre-ER about the margins, “So, we understand that the margins will be lower due to higher R&D expenses along with lower utilization due to inventory adjustments, ramp-up, overseas fab expansion, and inflationary pressures.”
The net profit declined by (6.3%) YoY to $6.8 billion. The net profit margin was 40.7% compared to 41.3% in the same period last year. The GAAP EPADR (Earnings per American Depository Receipt) came at $1.31 and beat estimates by $0.12. Return on Equity was 27.5% compared to 36.2% in the same period last year.
The free cash flow was $2.72 billion compared to $3.94 billion in the same period last year. The free cash flow margin was 16% compared to 22% in the same period last year. The company has a stable balance sheet. The company had cash and marketable securities of $52.24 billion and debt of $28.16 billion at the end of the March quarter.
Capex increased by 6.1% YoY to $9.94 billion. The management has reiterated the full year Capex guidance of $32 billion to $36 billion.
Smartphone declined (27%) QoQ and accounted for 34% of revenue. HPC declined (14%) QoQ and accounted for 44% of revenue. IoT declined (19%) QoQ and accounted for 9% of revenue. Automotive increased 5% QoQ and accounted for 7% of revenue. Digital Consumer Electronics decreased (5%) QoQ and accounted for 2%. Others decreased by (18%) QoQ and accounted for 4% of revenue.
Other important earnings call updates:
The management had mentioned in the previous earnings call that 2023 will be a slight growth year for the company. However, the recovery is taking longer than expected due to inventory adjustments expected to continue due to the challenging macro environment and slowing end-market demand. So, the company expects 2023 revenue to decline low to mid-single digit in U.S. dollar terms. Revenue in the 1H is expected to decline by 10% YoY and the H2 revenue is expected to be better than the 1H. The key takeaway is that the company will perform better than the industry.
C.C. Wei, CEO of the company, said in the earnings call. “3 months ago, we said we expect fabless semiconductor inventory to start gradually reducing 4Q 2022 and we forecast a sharper reduction throughout the first half of 2023. However, due to weakening macroeconomic conditions and softening end market demand fabless semiconductor inventory continued to increase in the fourth quarter and exited 2022 at a much higher level than we expected. In addition, the recovery in end market demand from channels reopening is also lower than our expectation. Therefore, the fabless semiconductor inventory adjustment in first half '23 is taking longer than our prior expectation. It may extend into third quarter this year before rebalancing to a healthier level.”fabless semiconductor inventory adjustment in first half '23 is taking longer than our prior expectation. It may extend into third quarter this year before rebalancing to a healthier level.”
“For the full year of 2023, we do our forecast for the semiconductor market, excluding memory, to decline mid-single-digit percent while foundry industry is forecast to decline high single-digit percent. We now expect our full year 2023 revenue to decline low to mid-single-digit percent in U.S. dollar terms and our business to do better than both semiconductor ex memory and foundry industries, supported by our strong technology leadership and differentiation.”We now expect our full year 2023 revenue to decline low to mid-single-digit percent in U.S. dollar terms and our business to do better than both semiconductor ex memory and foundry industries, supported by our strong technology leadership and differentiation.”
Management comments on N7 recovery. “It will be recovered but slowly. As I said, most of the N6 and N7's technology loading still in HPC and smartphone. However, looking into the future, some of the specialties such as RF, connectivity, WiFi, all those kind of things will start to build up the loading their demand. And we expect in the long term, 7-nanometers loading will become more healthier.”
The demand for N3 chips are strong. C.C. Wei said in the earnings call, “Our 3-nanometer technology is the first in the semiconductor industry to high-volume production with good yield. As our customers' demand for N3 exceeds our ability to supply, we expect N3 to be fully utilized in 2023 supported by both HPC and smartphone applications. Sizable N3 revenue contribution is expected to start in third quarter and N3 will contribute mid-single-digit percentage of our total wafer revenue in 2023.”Our 3-nanometer technology is the first in the semiconductor industry to high-volume production with good yield. As our customers' demand for N3 exceeds our ability to supply, we expect N3 to be fully utilized in 2023 supported by both HPC and smartphone applications. Sizable N3 revenue contribution is expected to start in third quarter and N3 will contribute mid-single-digit percentage of our total wafer revenue in 2023.”
Analyst Notes:
Susquehanna upgraded TSMC to Positive from Neutral with a $126 price target. The analyst says a "worst-case earnings scenario" is now reflected in investor expectations. The ramp of new products is helping with modest revenue improvement for TSMC in the second half of 2023 following a severe wafer shipment decline in the first half of the year, the analyst tells investors in a research note. With the utilization rate rebounding in fiscal Q3, the company's earnings should rebound at a faster pace than revenue, a trend that should gain momentum in 2024 as new product ramp.
Conclusion
TSM’s results are good taking into consideration the tough macro environment. The revenue guidance is lower. However, the bottom-line beat was the main highlight of the report. The company’s cost control efforts helped to improve the margins. Even though the recovery has been delayed, the company’s long-term growth outlook is still positive due to its leadership position in producing advanced chips that will be used in HPC and smartphones.
When we provide coverage to our readers, our goal is to discuss what is of highest value to an investor group. Although there were many exciting things announced during Nvidia’s GTC conference, one of the topics that AI semiconductor investors should take the time to discuss is the launch of cuLitho as it pertains to extreme ultraviolet (EUV) lithography.
At GTC, Nvidia announced a new cuLitho library to extend support for GPUs into its lithography software products. The future goal of this partnership is to move beyond the advanced 3nm node and to achieve a 2nm or smaller by “pushing the limits of physics” by combining EUV lithography process GPUs and software to employ algorithms that simulate the production process.
Years ago, we discussed Huang’s Law on the I/O Fund Forum. This was around 2020 when the Wall Street Journal published a paywall article entitled: “Huang’s Law is the New Moore’s Law.” The article was also discussed on Mother Jones where it pointed toward Nvidia’s pace of innovation on GPUs exceeding Moore’s Law. In 2018, Jensen Huang stated at GTC that GPUs were “25 times faster than five years ago” whereas Moore’s Law would have resulted in only a ten-fold increase. The more important part of Huang’s Law as it relates to ASML is that Huang stated: “The innovation isn't just about chips — it's about the entire stack."
Particularly, as microchips become smaller, it becomes harder for chip advancement to meet the speed of Moore’s Law. As investors, we felt that Moore’s Law running out of room, so to speak, would narrow the road for the winners. The thoughts we’ve been publishing for many years point toward entering an era of excellence where the most talented teams/competitive designs would forge forward by combining advances in architecture, interconnects, memory and algorithms. However, we have to give a solid nod toward the process in how these chips are made, as well – especially the chips that are surpassing the limitations of Moore’s Law.
To say the semi war is heating up is true, but it’s equally important to understood who the victors will be and why. This is why we’ve dedicated an enormous amount of time to writing about Nvidia’s A100 and H100 GPUs (27 articles in 4+ years!), and AMD’s Zen architecture. For ASML, we want to dig deeper into the tools and processes that power the market-leading foundries that are producing advanced semiconductors.
What is EUV and DUV Lithography:
Note: I’ve bolded the main points for easier reading.
Taiwan Semiconductor is the global leader in manufacturing advanced nodes, which are defined as those less than 10nm. Currently, TSM is producing a 3 nm chip compared to competitors who are delayed at the 7nm size. Powering TSM’s ability to shrink node size are ASML’s EUV lithography machines. We use words such as “A100 Ampere” or “H100 Hopper” or “Genoa,” “Milan,” “Bergamo” for AMD – but helping to drive the robust demand for these designs is the advanced node, or nanometer size.
Extreme ultraviolet lithography builds on ASML’s success with DUV lithography, or deep ultraviolet lithography. The word lithography means to “write on stones.” In the semiconductor industry, lithography uses a light-sensitive polymer or “photoresist” to write patterns on silicon wafers. Light is projected through the blueprint of the pattern that is printed. When the pattern is encoded in light, the system’s optics shrink and focus the pattern onto a photosensitive silicon wafer. After the pattern is printed, the system moves the wafer and makes another copy. Lithography creates the formation of 3D images on the substrate for the transfer of the pattern to the substrate (or the base of the silicon wafer). For each circuit, the lithography and pattern transfer are repeated anywhere from 10 to 30 times.
In simple terms, each silicon wafer has dozens of thin layers that make up billions of transistors. These layers are printed with lithography – which are extremely precise rays of light projected through a mask of the chip design. When light hits the surface, it prints the miniscule designs. Keep in mind, Apple has 10 billion transistors in each iPhone.
EUV is a new process that is needed for more advanced nodes, whereas DUV has been powering semiconductor fabs for decades. Deep ultraviolet lithography uses a 254 to 193 nm light whereas EUV uses a 13.5 nm light. This is 14X shorter than DUV lithography. As the wavelength of the light source becomes very narrow, fabs can produce chips with smaller features. Hence, the smaller nodes that TSM produces (3nm, 5nm).
EUV lithography machines are only produced by ASML as the process of creating EUV light is very complexvery complex. Outside of ASML’s machines, EUV only exists in outer space.
Per the company, “in our laser-produced plasma (LPP) source, molten tin droplets of around 25 microns in diameter are ejected from a generator at 70 meters per second. As they fall, the droplets are hit first by a low-intensity laser pulse that flattens them into a pancake shape. Then a more powerful laser pulse vaporizes the flattened droplet to create a plasma that emits EUV light. To produce enough light to manufacture microchips, this process is repeated 50,000 times every second.”
According to TSMC, “the aim is so precise that it’s like shining a laser from the moon to hit a coin from the earththe aim is so precise that it’s like shining a laser from the moon to hit a coin from the earth.”
Mirrors are an important component to how ASML achieves precision “down to the atom.”
ASML is partnered with a company called ZEISS for high-precision lenses and mirrors. The numerical aperture (NA) measures how much light the lens system can collect and focus. This is important because ASML is creating a machine called High-NA that will increase the numerical aperture from 0.33 to 0.55. High-NA is especially important at the advanced node level of 3nm or smaller. TSM is expected to have High-NA machines by 2024.
EUV lithography uses mirrors instead of lenses. According to ASML, these mirrors are the flattest surface on earth with smoothness that is less than one atom thick.
Production:
EUV machines produce 3,000 wafers per day, and there can be hundreds of chips on a 300-millimeter wafter, and up to 10 billion transistors per chip. The machines cost $200 million and are expected to reach an average sales price of $300 million in the future. Due to this cost, ASML has very few customers albeit large customers. TSMC makes up 40% of ASML’s revenue. Intel and Samsung are the second and third largest customers.
ASML’s Customers
ASML supplies fabrication plants, which produce semiconductor chips for design companies. Due to the cost of the machines and complexity of the process, ASML has exactly five customers. A few of the biggest fabs in the world are ASML customers: TSMC, Intel and Samsung. The other two are memory suppliers, Micron and SK Hynix.
ASML is capable of producing around 50 units of EUV equipment per year with a lead time of one year to six months. This is leading to a strong pipeline for both EUV and the next-gen High-NA lithography equipment. The company is expected to grow production capacity to 60 billion units in 2023. Management has stated (quite clearly) that average sales price (ASPs) will increase over time, for example, what was a $200 million machine will eventually cost $300 million.
Here is the production capacity ASML is expecting:
“[We have] plans of 600 DUV (deep ultraviolet), 90 EUV systems by 2025-2026 and 20 EUV High-NA systems by 2027-2028.”
TSMC:
We own TSMC stock but we’ve also been clear there is geopolitical risk to weigh. Taiwan operates in grey area where it has its own constitution and democratically-elected leaders, yet Beijing considers Taiwan as its territory. The island hosts the world’s most important foundry for advanced node semiconductors, and China is seeking a “reunification” with Taiwan. Meanwhile the United States is at odds with China by supporting TSMC in building two chip plants worth $40 billion in Arizona.
Taiwan Semiconductor, or TSMC, is ASML’s largest customer making up 40% of sales. In terms of advanced nodes, 90% are produced by TSMC. Customers include Apple, Qualcomm, Nvidia and AMD. Per the chart below, TSMC owns 60% of the foundry market.
TSMC is able to supply advanced nodes at high volume — such as 7nm or smaller — due to ASML’s EUV lithography technology. TSMC plans to introduce High-NA technology from ASML for the first time in 2024 to produce 1.8 nm chips.
Samsung:
Samsung is the world’s second largest foundry. According to BusinessKorea, Samsung ordered 18 of ASML’s 55 EUV machines in 2022. The company is investing over $4 trillion in EUV lithography equipment. Samsung has roughly half the EUV machines that TSMC has.
Samsung produces memory chips, plus GPUs for Nvidia, CPUs for IBM, smartphone processors for Qualcomm, and is also working with Baidu on AI chips for its cloud data centers. In 2021, the company used EUV technology to mass produce a 14 nm double data rate (DDR5) memory chip. This was the smallest chip to be produced in the memory industry. This memory chip is capable of powering AI/ML applications. The DDR5 chip increases wafer productivity, lowers power consumption, and doubles data processing speeds. At the time, a third-party analyst stated DDR5 would account for 41% of the world’s global DRAM market, up from 0.1% in 2021.
Micron and SK Hynix compete with Samsung on memory chips and are also ASML customers.
Intel:
Intel is building two fabrication plants worth a combined $20 billion, located in Ohio and Arizona. In 2022, Intel placed its first order for an advanced chipmaking tool “worth more than $340 million.” The Q4 2021 results indicated a customer had ordered five EUV lithography machines plus one order for the High-NA machine, due in 2025. This order was generally understood to be Intel, per the Reuters report.
Side note: I’ve had fun throwing shade on Intel on this site from a design perspective, however, I would not write Intel off as a foundry on American soil.
Computational Lithography & cuLitho Announcement at GTC:
The need for accuracy increases with smaller nodes. Computational lithography can help by optimizing the scanner, masks and processes. Rather than only relying on the design, AI lithography improves the process by using algorithmic models of the manufacturing process. ASML is able to produce a replica of the desired chip patterns on the wafer, which helps to meet the accuracy required for nanoscale chips.
At GTC, Nvidia announced a partnership with ASML and TSMC to launch a library called “cuLITHO.” The goal of cuLitho is to use software to optimize the lithography process. There will be increasing demand pressure on ASML to produce their coveted machines into the foreseeable future. cuLitho allows computational lithography to improve the process to make the most of each machine.
Each device requires a mask set of photomasks that mask the architect of the design to physical features. The cost of a mask set is becoming very expensive. For example, a 10 nm requires about 76 individual photomasks and this number increases as the chip gets smaller. According to SemiAnalysis, “at 90 nm to 45 nm, mask sets cost on the order of hundreds of thousands of dollars. At 28nm it moves beyond $1M. With 7nm, the cost increases beyond $10M, and now, as we cross the 3nm barrier, mask sets will begin to push into the $40M range.”
cuLitho offers what’s called inverse lithography to speed up the photomask process and drive down costs. By using software and algorithms, photomasks that took two weeks to process can be generated overnight. This will be especially important as designs move from 3nm to the 1.8nm. For Nvidia, the goal is to sell more GPUs as the need for lithography increases. The company advertises that 500 Nvidia Hopper GPUs that run CuLitho can do the work of 40,0000 CPUs.
For ASML, the company is dependent on CPUs and can now start to support GPUs. ASML has stated computational lithography is especially important for the High-NA EUV lithography due for production at TSMC in 2024 and at Intel in 2025.
TSMC’s motivation is to stay at the cutting edge as a foundry and to drive down costs by moving the expensive operations from CPUs to GPUs.
According to Nvidia’s Advanced Technology Group VP Vivek Singh: “If a silicon foundry has three data centers, it'll need 100 data centers by the end of this decade if the trend of the last 15 years continues – that's not feasible. And what about power? 45 megawatts might be okay, but 45 gigawatts? Something's gotta give."
By moving the patterning, lithography process to GPUs, companies can make 3-5X more photomasks using 9X less power.
Geopolitical Export Controls
ASML is not allowed to sell EUV technology to China. This began in 2018 when ASML signed a Chinese customer for its EUV lithography machines. The Trump Administration worked with the Dutch government to block the sale and to enforce export controls. The Dutch government did not renew ASML’s export license, which effectively stopped the sale. Rumors are that the customer was SMIC, China’s biggest chip-making foundry.
The export controls on ASML have continued under Biden by requiring “advanced computing semiconductors or related manufacturing equipment” to apply for a license if they want to export to China.
Financials and Q4 Earnings Call:
ASML’s earnings are this week. We don’t own the stock yet so those we do own have a higher priority – “i.e., bird in hand.” In addition to this, TSM is reportedly lowering its capex right now. Therefore, ASML’s timing for entry may be a bit better in H2.
Revenue:
Revenue can be lumpy for ASML. The company is expected to report revenue growth of 80% in the upcoming quarter for a total of $7.08 billion, up from 15.30% last quarter.
The next two quarters – June and September – are expected to be at 26.3% and 27% revenue growth. This will be $7.09 billion and $7.25 billion, respectively.
Pictured Above: ASML has lumpy revenue (but a strong backlog alleviates this concern)
Annual revenue is expected to track at 25% for FY2023 and 13.40% for FY2024.
There was negative revenue in the September 2022 quarter. This happens because ASML must source many parts to deliver one machine. The company is especially susceptible to supply chain issues. However, the company has a lot of visibility due to the backlog (see below) and a quarter of lumpiness is not a big concern to the longer-term story.
Earnings:
Earnings have been trending up nicely for ASML. Quarterly EPS was in the $2.00 range in 2021 and is now in the $3.00 to $4.00 EPS range. This bottom-line growth trend is expected to continue into the foreseeable future, hitting $5.00 Quarterly EPS in 2023 and up to $8.00 EPS by 2025.
Margins:
Gross Margin in the 50% range. This can be affected if ASML impacted by the Deep UV mix effect versus higher margins from EUV.
Operating Margin in the 33% range. R&D Spending is around 15% of sales and this has been consistent.
Net margin of 29% leaving net income at $1.94 billion. Net profit has been consistent and the company has been buying back shares.
Cash and Buybacks:
Free cash flow for the year 2022 was €7.2 billion or $7.56 billion. The company had announced a new share buyback program in November 2022 to be executed by December 31, 2025, for an amount of €12 billion.
The company had cash and investments of €7.4 billion or $7.91 billion and debt of €4.57 billion at the end of the December quarter.
Key Metrics:
Backlog:
Backlog is a key metric for ASML. At the end of 2022, ASML grew backlog by 67% for a total of $40.4 billion Euro, or $43 billion USD.
Net bookings were $6.3 billion Euro, which was $3.4 billion EUV and $2.9 billion Non-EUV. Logic drove 66% of bookings and memory 34% of bookings.
Product Mix:
The company shipped 54 EUV systems in FY2022, up from 40 systems.
Per the earnings call, the product mix for EUV bookings is about 75% logic and 25% memory.
DUV systems grew 13% to $7.7 billion Euro
Revenue mix is as follows:
Logic System Revenue contributes $10 billion and grew at a rate of 4%
Memory System Revenue contributes $5.5 billion and grew at a rate of 34%
Installed Base Revenue contributes $5.7 billion and grew at a rate of 16%.
Earnings Call:
Comments on Strong Backlog:
“We've experienced several quarters of very strong bookings, which now provides backlog coverage significantly beyond 2023, which is almost twice the expected 2023 system sales. Based on discussions with our customers and continued improvements in the capability of our supply chain, we are planning to increase our output capability this year. We're planning to ship around 60 EUV systems and around 375 deep UV systems in 2023, with around 25% of the deep UV systems to be immersion. We still plan a significant number of fast shipments this year, which under the current way of working will result in a similar amount of delayed revenue out of 2023 that came into 2023.”
Comments on Expected Growth of Product Mix and ASPs:
“And for the people that weren't really carefully listening, what that really means in terms of ASP for EUV, we talked about it before. Originally, we were looking at €160 million. We've then been talking about €165 million to recognize also increased functionality, I think, with the increases on ASP on the inflation. I think it's good to go somewhere between €165 million and €170 million. I think that's, on average, I think, the right way to go.”
There was a second comment on the call about revenue growth:
“Also presented during our Investor Day last November, we see an opportunity based on different market scenarios to reach an annual revenue in 2025 between €30 billion and €40 billion and in 2030, an annual revenue between €44 billion and €60 billion.”
More on the Demand and Backlog:
“Having said that, of course, last year, we kept informing you that the demand on us significantly exceeded our build capacity, sometimes to 40%, 50% […] I think — and that gives us a lot of visibility into 2024 also. Customers give us orders throughout the year, very significant levels of orders, which actually have over €40 billion in the backlog, which is almost twice the system sales that we expect to have in 2023.”
Quick Note on TSMC Capex:
We are on the eve of ASML’s earnings report and TSMC will officially report the next day. There has been a pre-announcement which we covered here. However, per a Barron’s report, TSMC may be lowering its capex. Right now, 2023E for capex is for $32 to $36 billion, or $34 billion at the midpoint.
Pictured above: Management guidance for 2023 is $32B to $36B, representing a YoY decline of 6.3% at the mid-point
Per an article behind Barron’s paywall, one analyst is expecting TSM to further reduce its capex in the upcoming earnings report. We will see what happens and plan an entry into ASML accordingly. Although ASML serves other foundries, it’s likely TSM has to participate for renewed stock price action as it contributes 40% of revenue.
Conclusion:
The main takeaway is that the future of chips will be powered by ASML’s full monopoly on EUV lithography machines and there is no competitor at this moment (literally not one competitor). With a deep and undeniable moat, the restraint on stock price, if you will, is that production capacity is limited as ASML can only produce a certain number of machines every year.
We feel it’s best to buy this stock when we see capex increase across leading foundries. Right now, we are seeing some indication foundries are going to reduce their capex in the near term, yet this AI semiconductor train is headed in only one direction, which is up. Some finesse on entry now will pay off, coupled with strong due diligence on the entries we are seeking.
ASML’s estimates are likely based on capex estimates. If the AI war heats up, ASML may be able to charge higher ASPs, while doubling its bottom line. Overall, ASML offers lower risk than most tech stocks while serving overwhelming demand. Should we continue to see demand dry up in other tech verticals, ASML’s rare, bottlenecked backlog will stand out in the current macro environment.
We are expecting the company to confirm a rebound in the second half of the year. Even though 2023 will be a challenging year, the company will continue to grow faster than the industry. The company’s CEO, C.C. Wei, mentioned in the Q4 earnings call, “For the full year of 2023, we forecast the semiconductor market, excluding memory, to decline approximately 4%, while foundry industry is forecast to decline 3%. For TSMC, supported by our strong technology leadership and differentiation, we will continue to expand our customer product portfolio and increase our addressable market and we expect 2023 to be a slight growth year for TSMC in U.S. dollar terms.”
We like TSM as it has developed market leadership in the foundry industry, particularly with advanced nodes, which are nodes defined as 7nm and below. The advanced nodes have strong demand by top design companies, such as Apple and Nvidia, particularly in high-performance computing and smartphones. The company should also benefit from the recent generative AI trend and help withstand any slowdown in the macroeconomy.
Please note: Tensions with China are outside the scope of a fundamental analysis. It’s important to acknowledge that TSMC’s $40B Arizona plant will increase tensions, we prefer to use technical analysis to gauge sentiment around this issue as it’s impossible to predict China’s reaction. We include the price levels we are watching below.
What we are watching:
H2 rebound confirmation. In the earnings call, C.C. Wei said, “We forecast the semiconductor cycle to bottom sometimes in first half 2023 and to see a healthy recovery in second half this year. In the second half of 2023, we expect our revenue to increase over the same period last year in U.S. dollar terms.”semiconductor cycle to bottom sometimes in first half 2023 and to see a healthy recovery in second half this year. In the second half of 2023, we expect our revenue to increase over the same period last year in U.S. dollar terms.”
The June quarter is expected to report $16.5 billion and the September quarter is expected to report $19.8 billion. This QoQ growth is important to watch and for the analyst estimates to hold here, particularly, for the H2 rebound.
The management mentioned in the last earnings call that N7 and N6 chips demand outlook was weaker than expected due to end-market weakness in smartphone and PCs.
The management updates on 3nm chips. The company expects sizable N3 revenue contribution to start in Q3 and contribute to single-digit percentage to the wafer revenue for the year 2023.
The margins will be lower in 2023, as the management mentioned in the last earnings call. “We have just guided our first quarter gross margin to be 54.5% at the midpoint mainly due to a lower capacity utilization rate as customers further adjust their inventory levels and a less favorable foreign exchange rate. In 2023, our gross margin faces challenges from lower capacity utilization due to semiconductor cyclicality, the ramp-up of entry, overseas fab expansion and inflationary cost.” We will look in the call for more insights. The company has leading margins, so it is not a major concern.
Financials
The company reported the March monthly numbers on April 10th. March revenue declined by (15.4%) YoY to NT$145.4 billion. The consolidated Q1 2023 revenue grew by 3.6% YoY to NT$508.63 billion. In US dollar terms, the Q1 2023 revenue comes to approximately $16.74 billion by using the average exchange rate of 1 US dollar to 30.39 NT dollars. (Please note that the company’s official exchange rate could vary slightly). The revenue declined (4.7%) YoY and came at the lower end of the management guidance. We will look for more details when the company reports its results on April 20th.
The weakness in the consumer market could be one of the reasons for the company to report at the lower end of the guidance. According to the preliminary results by IDC, there was a YoY decline of (29%) in the shipments of traditional PCs in Q1 2023 due to weaker demand and excess inventory. IDC expects growth after 2023.
Q4 revenue grew by 26.7% YoY to $19.93 billion. The management revenue guidance for Q1 is $16.7 billion to $17.5 billion, representing a YoY decline of 2.7% at the mid-point of the guidance. The company’s CFO Wendell Huang said in the Q4 earnings call “As overall macroeconomic conditions remain weak, we expect our business to be further impacted by continued end market demand softness and customers’ further inventory adjustment.”
Below is a screenshot of the analyst’s revenue estimates which will vary due to the foreign exchange calculation. We use the estimates to understand the trend and use the company’s IR revenue USD figures in our reports. The analysts expect revenue to grow 1.8% YoY to $17.19 billion in Q1 and to decline (7.7%) YoY to $16.46 billion in Q2.
Source: Seeking Alpha
On a yearly basis the analysts expect revenue to grow 0.8% in 2023, 19.8% in 2024, and 14.7% in 2025.
The adjusted EPS is expected to decline this year and then rebound next year. The analyst expect GAAP EPS of $1.21 for Q1 and $1.11 for Q2.
Source: Seeking Alpha
The gross profit in Q4 came in at $12.4 billion compared to $8.29 billion in the same period last year. Gross profit margin improved to 62.2% from 60.4% in Q3 2022 and 52.7% in Q4 2021. It was higher than the management guidance of 59.5% to 61.5%. It was higher due to cost improvements and favorable foreign exchange rate, partially offset by lower utilization rates. The Q1 2023 management guidance for the gross margin is between 53.5% to 55.5%. The company reported 55.6% in Q1 2022.
The company has industry-leading operating margins. In addition, the company is working on cost improvements and in the past has been able to negotiate better prices with its customers.
Source: YCharts and Company IR
The operating income was $10.36 billion compared to $6.56 billion in the same period last year. Operating margin improved to 52% from 50.6% in Q3 2022 and 41.70% in Q4 2021. It was higher than the management guidance of 49% to 51%. For Q1 2023, the management guidance is between 41.5% to 43.5%. It is lower than the operating margin of 45.6% in Q1 2022.
The management mentioned in the last earnings call, “R&D expenses accounted for 7.2% of our net revenue in 2022. In 2023, as we increase our focus on technology development and add more resources, we expect R&D expenses to increase by about 20% year-on-year and account for 8% to 8.5% of our net revenue.”
In addition, the management also answered to an analyst’s question on the reason for the rise in R&D expenses. “We’re the technology leader, and we intend to continue to maintain the leadership. Therefore, we are devoting more and more resources in R&D, including people and other kind of resource. That’s the reason why our R&D expense will increase in 2023 and probably beyond.”technology leader, and we intend to continue to maintain the leadership. Therefore, we are devoting more and more resources in R&D, including people and other kind of resource. That’s the reason why our R&D expense will increase in 2023 and probably beyond.”
So, we understand that the margins will be lower due to higher R&D expenses along with lower utilization due to inventory adjustments, ramp-up, overseas fab expansion, and inflationary pressures.
The net income was $9.43 billion compared to $5.97 billion in the same period last year. Net profit margin improved to 47.30% from 37.90% in Q4 2021. The EPADR (Earnings per American Depository Receipt) came at $1.82 compared to $1.15 for Q4 2021.
The company has good free cash flow. The free cash flow was $4.78 billion with a free cash flow margin of 24%, compared to a free cash flow of $4.84 billion (free cash flow margin 24%) in Q3 2022 and $5.12 billion (free cash flow margin of 33%) in Q4 2021.
The company has also cut capex for 2023 due to the anticipated slowdown, which is a positive. The company had spent $36.3 billion in capex in 2022 and for the year 2023 it expects between $32 billion to $36 billion.
The company has a stable balance sheet. The company has cash & marketable securities of $50.84 billion and debt of $27.8 billion.
The company is trading at a P/E ratio of 14.10 and is lower than the average five-year P/E ratio of 22.97. The P/S ratio is 6.31 and below its five-year average P/S ratio of 8.4. This is positive that the company is trading below its historical average. The forward P/E ratio is 16.5 and forward P/S ratio is 6.34.
Noteworthy:
Analyst Gokul Hariharan in the Q4 earnings call, asked a question on the outlook for 2023. “Could we have some more color on what is that gives you the confidence for such a strong rebound in the second half of the year to get us back to like a flattish revenue growth for the year?”
The company’s CEO, C.C. Wei, replied, “The inventory correction actually began last year. And at the peak of the third quarter, and we think the inventory has been picked in third quarter last year and gradually reduced in the fourth quarter, and we did see some inventory reduced sharply recently, and it will continue to be so to first half of this year. So that’s why we say we have confidence that in the second half, the business will rebound. But is that a very strong V shape? We didn’t know yet, but certainly, it’s not a U shape for the business to recover in the second half.”But is that a very strong V shape? We didn’t know yet, but certainly, it’s not a U shape for the business to recover in the second half.”
The 3-nanometer process technology is the current most advanced chip production technology. The company is expected to have strong demand in the coming years driven by HPC and smartphone applications. The management mentioned in the earnings call, “Our 3-nanometer technology is the most advanced semiconductor technology in both PPA and transistor technology, thus, we expect customers a strong demand in 2023, 2024, 2025 and beyond for our 3-nanometer technologies and are confident that our N3 family will be another large and non-large node for TSMC.”we expect customers a strong demand in 2023, 2024, 2025 and beyond for our 3-nanometer technologies and are confident that our N3 family will be another large and non-large node for TSMC.”
Recent Headlines
Nvidia is expected to benefit from the strong demand for Artificial Intelligence chips and the company has recently increased orders for AI chips with TSMC.
DigiTimes reported that Microsoft has approached TSMC to use the company’s CoWoS packaging for its own AI chip.
Apple is expected to use TSMC’s 3nm technology for its first self-made 5G modem chips to be used in iPhone 16 series. Previously, Apple used to purchase 5G modem chips from Qualcomm.
Berkshire Hathaway slashed its stake in the company in February. We have proactively trimmed our position due to geopolitical tensions related to China and due to the Warren Buffet sale.
What Analysts are Saying/Channel Checks
Bernstein analysts highlighted the 1Q23 revenue miss but noted the results still "met the low-end of the guide." The analysts reiterated an Outperform rating on the TSM stock.
KGI Securities resumed coverage of TSMC with an Outperform rating and NT$603 price target. The analysts said that the demand for high-performance computing, fueled by increasing adoption of AI in various applications, has led to a significant increase in silicon dollar content per chip or socket and TSMC stands to benefit.
Bank of America had a positive note on the company due to AI. "We think the generative AI should act as one of the greatest drivers, thanks to the substantial computational requirements for running and training the AI models," the analysts wrote.
"Datacenter (including supercomputing) related revenue currently accounts for ~10% of the revenue, and we estimate that CPU/GPU/accelerator upside for generative AI could potentially contribute 1%-2% initially and likely up to 8% in a bull case," they added.
They also believe there are structural long-term opportunities for the company.
"With rising computing power demand to shorten the time to market and provide better service quality with a faster response time, we believe the pursuit of leading-edge technology will not decelerate. Advanced packaging adoption will also grow. TSMC, with tech leadership in both, will ride on the structural uptrend," they concluded.
Our Plan
The recent escalation between China and Taiwan is something we are monitoring and will use TA to help navigate. This will hit TSM especially hard, but it will have effects on our semis like NVDA and AMD, plus AAPL. We reduced our position in TSM around $87 in late February. We currently hold it in a low allocation in our portfolio, and are targeting better entries. We believe TSM should, at minimum, go back into the $70s, and possibly lower.
Until we get clarity on the path TSM will take, we plan to hedge our current allocation and wait for clarity so that we can get a better entry in this excellent company.
Last Friday, we decided to log gains in both NVDA and AMD, while keeping NFLX on hold. We logged a 24.45% gain in NVDA and a 20.78% gain in AMD. Our conviction on NVDA and AMD is as high as any stocks we’ve covered in our service, so it may seem odd that we are closing these positions today.
The simple explanation was best expressed by the famous investor, Howard Marks, when he stated, “It’s not what you buy, it’s what you pay for it that determines whether something is a good investment.” Because of the current macro environment, coupled with stretched valuations in both stocks, we believe that we will be able to get much better prices in the coming months. In order to confirm a return to good times, we would like to see favorable price action in the markets, a new liquidity cycle start in the economy, and avoiding the looming credit cycle. As of now, all three of these factors are not suggesting a new bull market is starting.
On December 30th, we alerted our readers to the strength in the semiconductor sector. We stated that, “we believe that semiconductor stocks are signaling that they are ready to resume their leadership role going into 2023.” Since then, not only was the semiconductor sector the strongest tech sector, but it has turned out to be one of the best places to invest in for 2023.
This call has proven to be very timely, both as a general theme and in how we decided to play this trend. Our starting cost basis in NVDA was $212.65 on 2/11/23, and $76.37 for AMD on 11/15/22. However, in our premium service, we have a much lower cost basis in both. We started buying NVDA at $108.51 on October 13th, while our starting cost basis for AMD was $57.34 the day after.
We mention this to our readers because we believe that through our process, we can help navigate when to add/trim/sell/start a position in great tech stocks. Our goal is to own great companies at great prices for the long-haul. However, we believe the time to get defensive is a prudent position to take for following three macro reasons:
The Market Price Action Is Risky
The structure of the broad market since the October low is a clear 3 wave bounce. This is a risky structure, as it tends to suggest we are in a corrective bounce in a larger downtrend. Since February, we have seen many major banks stocks suggest a top is in, while many tech names appear to be putting in a topping pattern. Our primary case is that the broad market tops out over the next couple of weeks/months, and begins its push towards the 3000 level SPX. There is a chance that we see a push to new highs first, but it is uncertain whether that happens or not.
Note in the above chart the Relative Strength Index (RSI) below the price chart. This is a momentum indicator that produces very important patterns. Note how the post-COVID bull market held the black trendline as support. It then broke this support, which has been resistance in the following bear market. As of now, we are not seeing internal momentum that is suggesting a new bull market is developing. This is a further warning to the bulls, as the technical picture is not as healthy as some make it out to be.
Discount Window Borrowing Is Not a New Liquidity Cycle
Liquidity can be measured in the economy based on the Federal Reserve’s Policies as well as the Treasury’s general account balance. When you measure the liquidity in the economy against the S&P 500, you can see a clear correlation between the two.
As the FED starts a new liquidity cycle (black line going up), stocks eventually follow. Some may be encouraged by the sudden rise in the FED’s balance sheet as a sign that a new liquidity cycle is starting. This assumption would be a mistake.
Yes, the FED’s balance sheet expanded, but the FED’s balance sheet consists of many facets beyond quantitative easing (QE). In QE, the FED actively seeks out to buy long-duration bonds in exchange for bank deposits (not money). Their intention is to control the yield curve on the long end, to encourage economic activity, and it is an intentional action.
The part of the balance sheet that just expanded is what's called primary credit, or short term loans through the FED's Discount Window. The Discount Window is one of the FED's original purposes – to be the lender of last resort to banks that need liquidity now. This is a stigma for banks when they have to use it, because it means they are in dire straits and desperately need liquidity to maintain daily operations. These loans must be backed by collateral and have a very high yield that must be paid, which is about 4.6% right now. This is very expensive money that no bank takes unless it's absolutely necessary.
That being said, if we isolate the part of the FED's balance sheet that is primary credit, here is what we get.
These are expensive loans that banks do not want to hang onto for long. So, they typically get paid back quickly. Also, note when these spikes happen, and it is not during good times. They are the result of liquidity drying up and banks needing expensive injections to survive daily operations. The largest bump in weekly primary credit happened a few weeks ago – $152B vs. the prior high in 2008 at $111B.
Most importantly, we can see that during the same time the FED continued to engage in quantitative tightening (QT) while continuing to raise rates. Even though banks are taping the Discount Window, the FED is practicing restrictive monetary policy, and thus not starting a new liquidity cycle.
Don’t Forget About the Credit Cycle
It has been nearly 14 years since we have experienced a true credit cycle, so many investors have not experienced one. Due to low inflation, the FED has been able to flood the markets with liquidity in several periods of weakness, allowing the economy to avert a credit cycle. This resulted in relatively short, while sometimes deep, corrections that were bought up quickly.
However, due to inflation being at a 40 year high, the FED does not have this convenience anymore. They have thus engaged in one of the most aggressive rate hiking cycles we’ve seen since the 1970s.
What many forget is that rate hikes cause damage to the economy, and by some measures, it can take up to a year for those rate hikes to actually effect the economy. The bulls are betting that the rate hikes that are still happening, will somehow not affect the economy in a material way, which history does not support.
In the chart below, I compare the Fed Funds Rate (BLUE) to the S&P 500. The below gray bars show the lending standards for banks. As banks get more concerned about the economy, the less loans they provide, causing a cascade of defaults.
Note how the FED starts a new liquidity cycle usually around the top in equity markets. They start lowering rates, knowing that it will also take time for these lower rates to filter into the economy. However, once rates go up too high, the damage is done, and the economy as well as the markets must go through a credit cycle before a new expansion period can start. What’s concerning is that the FED is still hiking rates, meaning that this credit cycle could take longer to cycle through than most think.
In conclusion, bull markets do not happen in vacuums. They require expanding credit and expanding liquidity, both of which are not happening now. The price action in the broad market is also not favorable to the bulls, which has us playing defense. NVDA is up over 150% off its low, while AMD is up over 75% off its low. These two, at best, are due for a pullback. But, considering the macro environment, we believe taking gains now will provide us more cash to buy these great stocks lower in the coming months.
Aehr has been a wild ride since it’s last earnings report. The orders that came in were substantial, including a $25 million order from ON Semi. This order alone is 50% of last year’s $50M in revenue. The company was in the crosshairs of Tesla’s comments about a reduction in silicon carbide in their lower-tier models, and was also in the crosshairs of the failure of Silicon Valley Bank. When it seemed the stock simply couldn’t go any higher, it defied the odds, and marched higher.
Now, the market is shaking the stock loose on an excellent earnings report. Welcome to the world of small caps. The headlines are pointing toward the CFO leaving, but it’s not a real concern as he’s retiring and not moving onto a new company.
In a nutshell, the company beat on the top line and the bottom line plus reported strong margin expansion year-over-year. Compared to fiscal Q2, the margins were softer by 220 bps on gross margin, 150 bps on operating margin and 150 bps on net margin. This would be nitpicking the report, because on a YoY basis, the margins have expanded nicely (more below).
AEHR is not raising full year guidance despite beating on the top line. This implies $17.3 million to $27.3 million in revenue for next quarter. This compares to $17.2 million this quarter and $20.2 million in the year ago quarter. The high end of this guidance is not a problem given it would represent YoY and QoQ growth. However,given Aehr’s valuation, next year’s fiscal guide from management is where the market may be a touch nervous.
Bookings and backlog were very healthy this quarter (best in company history), which helps in the absence of management pulling forward the current quarter’s revenue beat. One thing to watch is that customer deposits were down, I quote the CEO on this below.
I do think it’s “all eyes” on the next fiscal year guide in July for a few reasons.
There are only two analysts covering the stock. Management’s input on what to expect is outsized, in this case. The fiscal year consensus for the 2 analysts is $102.3 million, for growth of 56.8%. This is a sizable hurdle to clear (will be up from roughly 29% growth this year), and with 1 quarter to go before management gives it’s guide, the market likely wants confirmation they can clear this expectation.
Management has referenced a strong H2 2023/2024 and this will help quantify those comments.
The valuation is quite high, and to support this, a fiscal year guide from management is very much needed come July.
Financials:
Aehr reported revenue of $17.2 million, for growth of 13% year-over-year and sequential growth of 16%. As stated above, the company did not raise full year guidance, rather reiterated “total revenue of at least $60 to $70 million, representing growth of 18% to 38% YoY, with strong profit margins similar to last year.”
EPS of $0.16 came in slightly higher than expected. This is up from $0.14 EPS from a year ago. The margins saw a turnaround in CY2021 (post-Covid) when margins were negative. This quarter was aligned with the new trend toward margin expansion for the company.
Margins:
Gross margin of 51.6% up from 48.6% in the year ago quarter. Last quarter, the GM was 53.40% softer by 180 bps
Operating margin of 22% up from 14.8% in the year ago quarter. Last quarter, the OPM was 23.50% softer by 150 bps
Net income margin of 23.8% is up from 14.7% in the year ago quarter. Last quarter, net profit margin was 25.30%.
On the margins, management pointed toward a slight increase in SG&A and R&D for the QoQ change. On net income, the GAAP includes a $1M adjustment for excess and obsolete inventory.
Cash:
Free cash flow will be available in the 10-Q that is filed over the next two weeks.
The August quarter reported cash flow margins of 50% which offset the other quarters:
In the year ago quarter ending in February, the company reported operating cash flow of (19.2%) and FCF margin of (19.7%). This was roughly ($3M)
Last quarter, the company reported operating cash flow of (1.3%) and free cash flow of (1.4%). This was roughly ($200K).
The cash has increased to $42.8 million, up from $36.6 million last quarter. This is due to the agreement for a $25 million at-the-market offering, of which $7.3 million was sold last month at a share price of $34.78. This leaves $17.7M remaining. There is dilution of 2.4%.
Key Metrics:
This earnings report stood out in terms of key metrics. Bookings were the highest in company history at $33 million, up from $10.8 million last quarter. This puts the three quarters of fiscal year to-date at $72 million, compared to $62.2 million for the entire fiscal year last year.
The backlog is at $31.6 million with an effective backlog of $41 million. The backlog in the year ago quarter was $26.9 million, and the effective backlog was “over $30 million.”
Inventories are ticking up, which can often be seen as a negative (company like Micron participating in a cyclical slowdown). For Aehr, it’s good to have inventory on-hand for any spikes in demand. Inventories were at $21.6 million, up from $3.6 million last quarter and up $6.5 million in the year ago quarter. This is only 1/5th of next year’s fiscal consensus estimate of $102 million.
“This quarter, AEHR discussed ramping inventory by an additional $5 million (so far) year-over-year in Q2: “We are increasing inventory to support our expected growth in the second half of fiscal 2023, and we continue to purchase inventory to ensure adequate supply to meet current customer and future customer market demand.”
“The company mentioned “momentum into 2024” in this call: “And as we had — if you look at the amount of capacity that everybody’s talking about to hit in 2025 calendar-wise, most people are just really focused on second half 2023 and into 2024 is where just a lot of capacity is coming online and so it may be less to do with the timing of us as the timing of that silicon carbide ramp. And our goal is to get qualified before that ramp happens and have a ton of capacity and material on hand to be able to address it.”
This quarter, Aehr discussed ramping inventory by an additional $5 million (so far) year-over-year in Q2: “We are increasing inventory to support our expected growth in the second half of fiscal 2023, and we continue to purchase inventory to ensure adequate supply to meet current customer and future customer market demand.”
Earnings Call:
Current Fiscal Year Guide:
Here’s a question from one of the analysts that cover the stock on the current fiscal year guide still looking conservative and he also notes the wide range:
“Jed Dorsheimer
So I guess first question, Gayn or maybe Ken, maybe you want to take either one. But the guide and kind of reiterating the numbers suggest a pretty wide variance at this stage in the game, $17 million to $27 million […] And I know that there were two tools with the — that we weren’t — that you weren’t sure were not you get the rev rec to fall into the quarter. But I am wondering, is that the only thing that sort of kind of the difference of that $10 million or is there something else that you can probably provide a bit more color on?”
Gayn Erickson
“Aehr Test along with most, I think, all capital equipment companies have revenue recognition policies related to when you can score revenue and that is different than when you get paid by the way. Our policy, I think, is very conservative. If we have a new product, particularly to a new customer, but if we have a brand new product that has never been proven or installed and accepted by the customer, we simply don’t take revenue for it until that milestone, even though we know it’s working here, it’s been completely proven out, et cetera, but until the customer actually signs off on it, we won’t score revenue recognition. And we gave that as a pretty big heads-up going in. That’s why a lot more detail than normal, and candidly, we will probably be pulling back on detail related to things. It’s just to make sure that our shareholders understand that we have got some pretty large revenue number of things that are shipping during the quarter, but may or may not score revenue.
And you have several multimillion-dollar tool that misses by a few days and it’s pretty easy. What I want to make sure that and I will be explicit even though it’s just been implied, we are just talking about whether it comes in, in Q4 or Q1. So that’s the bulk of it.”
Jed Dorsheimer
No. The color is helpful. So thank you. I guess if you could just help me reconcile just two moving parts. Inventory, not surprisingly picked up as you talked about in terms of ramping some of these products, but customer deposits dropped off on the balance sheet. I was wondering if you could just provide a bit more color there. Is that a timing issue or how should we read those two vectors, if you will?
Gayn Erickson
Yeah. That’s a good observation and good to move in [ph]. So we actually have taken with some specific terms and conditions with customers. There are circumstances where we do not take down payments. It’s a pretty good threshold contractually for them to actually do that.
I have also at times on a brand-new product with a new customer, waived the down payment to begin with, because it’s a little odd to tell them we guarantee it’s going to work and then at the same time, we holding their money.
And candidly, people are pushing back harder and harder on some of those deposits. Ken, I think, a lot of it is that they can earn a lot more money on that too. But there’s a little bit of examples where some of the backlog is not all out of deposit and that’s what you are seeing.
New Customers, New Products and New Markets:
Aehr has four committed customers for silicon carbide. Per the CEO: “We have actually announced a total of four customers in silicon carbide so far. We expect production orders from all of them during the next fiscal year.” One of these customers has not announced it’s in the silicon carbide market yet.
Gallium Nitride is getting more air time on the earnings call. From the sound of it, this will be the next market Aehr participates in a meaningful way.
“In addition to our momentum in silicon carbide, we are now engaged with several gallium nitride semiconductor suppliers ranging from radio frequency or RF devices to power devices. Since our last call, we also received a firm commitment from a very large multinational semiconductor supplier to move forward with a full wafer level evaluation of gallium nitride devices. This evaluation includes our new high voltage option for doing the critical HTRB stress needed for gallium nitride MOSFETs and amplifiers.
We believe gallium nitride will be a significant market, driven by some of the very — some very high volume applications such as RF amplifiers, consumer, electronic power converters and chargers, solar power inverters and charger and converter applications in both standard and electric vehicles. Feedback from companies has been that several of these applications will require production burn-in to meet the application’s critical quality and reliability needs.”
On the topic of silicon photonics … the CEO said “while we believe this transition is still several years out”, yet did state they have 6 potential customers: “Aehr currently has systems installed at over half a dozen customers for 100% test and burn-in of silicon photonics devices used in 5G infrastructure, data and telecommunication transceivers and a few additional applications yet to be introduced.” They also specifically name dropped Intel, Nvidia and AMD for plans to “integrate silicon photonics transceivers into their microprocessors, graphics processors and chipsets.”
China was talked about quite a bit.
The far majority of Aehr’s business comes from outside of China, but Aehr believes over time, the company will serve this market for silicon carbide.
“We are also talking to suppliers in China, as well as OEMs in China. So we are kind of making our way up the food chain, if you will, with several conversations with Tier 1s and OEMs, which as people that are close to this realize that is a completely new thing.
Prior to COVID, none of the automotive guys talk to the semiconductor guys, right? They all worked with Tier 1s and then the Tier 1s bought from the semiconductor guys. But with all the craziness that went on supply chain, automotive guys who realize they need to go directly to and talk to the semiconductor guys. Well, we are taking a step further, they are talking to us […] Having said that, we are very confident in next year and how things are going and candidly, without trying to be in a significant portion of it, I would say that would be upside to our plans.”
Please note:
These new markets, new products and new customers are exciting, yet we have emphasized many times that the majority of Aehr’s revenue comes from one customer today — On Semi. There is customer concentration risk. We want to weigh what drives revenue at the company today alongside what can move the needle over the next year. We are comfortable with this risk, each I/O Fund Member will need to decide for themselves what their risk tolerance is around high customer concentration.
Conclusion:
Aehr is a company where technicals are going to be of utmost importance as we surf the silicon carbide wave. Per our last earnings report write-up: “We took gains in AEHR recently because we felt it was the responsible thing to do. The small cap had grown to be the top leading position in our portfolio. However, we’d like to build back at key times as the company is doing all the right things.”
Some rough numbers: The stock is up 68% since the last earnings report when I stated we took gains – that 68% includes today’s (15% pullback). We have been reducing our position since late November, not out of lack of conviction, but because we feel it’s the responsible thing to do. We then attempted a 1% breakout.
My point is that small caps are one where investors should determine how they want to play this. Going long and strong with a company that has this kind of TAM is understandable. SiC wafer market is expected to grow 35X by 2030 – that’s not a typo. “Forecast from William Blair estimate that the silicon carbide market for devices in electric vehicles alone, such as traction inverters and onboard chargers is expected to grow from 119,000 6-inch equivalent silicon carbide wafers for EVs in 2021 to more than 4.1 million 6-inch equivalent wafers in 2030, representing a compound annual growth rate of 48.4%. This equates to almost 35 times larger in 2030 than in 2021.”
Notably, Aehr’s wheels fell off last year and it dropped (70%) with no real notable change in the story. Many investors will look for a “why” but it’s the nature of small caps facing macro headwinds, which results in a risk-off appetite. Even if an investor thinks they are mentally prepared for this, it’s extremely uncomfortable when a selloff in small caps (and other larger tech stocks) actually happens.
We prefer to be in the middle with this stock – we are going to participate heavily at times (it grew to be our number one position) but also try to take gains when we can. The company is trading at a forward P/S of 13. The stock is much safer under a 10 forward P/S and safer yet under a 6 forward P/S. If we get into this range, you’ll probably see quite a few buy alerts.
The Forward PE Ratio is similar – not much history holding at this level.
The fiscal year guide is very important because it will lay the foundation for the stock’s valuation. If the guide is higher than analysts are forecasting, then these valuations get cheaper overnight. The orders that are announced throughout the quarter help, in this regard. The risk would be a lower FY2024 guide (given only two analysts cover the stock) and valuations will be forced to get cheaper. Hence, there is some buildup going into the July call.
Additional Information on Orders and Recent Headlines:
Note: we’ve spoken about this throughout the quarter, listed here for reference purposes.
For a small cap, Aehr has had to weather quite a few headlines this past quarter.
Aehr holds a credit line and checking account at Silicon Valley Bank. The official statement from the company is the following:
“Aehr Test does have a checking account at SVB with a current balance of under $2.5 million, which is less than 6% of our total of $41.8M in cash and short-term liquid assets”
“Aehr has over $39.3 million in another financial institution which includes over $9.7 million in cash and $29.6 million in short term US Government backed Treasury Bonds.”
“Aehr has no outstanding balance on its line of credit with Silicon Valley Bank and foresees no need to draw on the line in the near future. Aehr believes that there is no impact to our operations, customers, vendors, or employees. We are taking all appropriate steps to prevent any impact on our operations.”
The second headline, which had a larger impact on daily price movement, was Elon Musk’s comments about reducing the need for silicon carbide. Management was quick to respond with the following:
“Tesla clarified that this will not impact the current high-performance model platforms including the Model S/X and Model 3/Y vehicles. Also, we believe that the new chips in the lower cost models will be 100 Amps per device versus 50 Amps per device today and likely 50% or more larger in surface area; therefore, the number of wafers required will be less impacted”
“This is important as Aehr’s total available market is primarily driven by the number of wafers required, not the number of devices”
“It is also important to understand that a 100A device using today’s generation of silicon carbide devices is approximately 50% larger than the devices used in the current Telsa inverters. So, while this new lower performance 800 kVa inverter only uses 12 die or 75% less than the current 48 die design, the die themselves are estimated to be about 50% larger, or require 50% more wafers for the same number of die.”
“In addition, during the Q&A session, Tesla further clarified that the new inverters would be made from a new Tesla-proprietary custom module package, and that Tesla would purchase the die from multiple manufacturers and package them in this Tesla-proprietary custom module. Again, Aehr sees this as a natural roadmap and consistent with the roadmaps stated by major manufacturers of silicon carbide where the electric vehicle inverters will migrate multi-chip modules to reduce power conversion losses, improve thermal performance, simplify design, and lower overall cost of the inverter system. As companies migrate to silicon carbide modules with multiple die in a single module package, the need for wafer level test and burn-in become critical to ensuring automotive quality and reliability as well as cost as the yield loss as a result of the stress test induced failures during burn-in become extremely expensive as a single die failure in a module results in throwing away the entire module including the other die in the module. Therefore, we believe the business use case for our solution actually increases. Wafer level test and burn in of 100% of die and extended burn in times will be required to earn Tesla’s business.
Orders Since the Last Earnings Report
In January, AEHR received a $25.1 million order for FOX-XP Test and Burn-in Systems. This will include a later order of the WaferPaks (the razor-razor blade model). The customer was ON Semi, per the CEO stating it came from their “lead silicon carbide customer”
A little more than a week ago, AEHR received a Volume Production Order from a “Major Silicon Carbide Customer” – this was not ON Semi as it was stated it came from Aehr’s “second major silicon carbide semiconductor customer.” The shipments are expected to begin March 1st.
The tech giant reveals cutting-edge AI advancements, a powerful cloud based Omniverse platform, and strategic collaborations in the automotive industry.
This year at the GPU Technology Conference (GTC) 2023, NVIDIA unveiled a series of groundbreaking AI innovations, talked about upgrades to their Omniverse platform, and highlighted new partnerships that promise to revolutionize the world of computing and automotive manufacturing.
AI Breakthroughs and Applications
NVIDIA showcased its first PCIe dual-GPU in years with applications to train the latest generative AI models, GPT4. NVIDIA has been involved with the training function of these models which have the potential to transform many industries, from language processing to visual content creation and even biology. The company also demonstrated AI Foundations, a new set of cloud services that enables the creation of custom language models and generative AI. Partnerships with Getty Images, Shutterstock, and Adobe highlight the potential of these new AI tools in creative fields.
Our first coverage of Nvidia’s AI thesis began nearly five years ago in the analysis “Holding Nvidia Stock Will Pay Off Due to Two Impenetrable Moats” Since then, the stock has returned 436%. More recently, in January, I discussed why Nvidia powering OpenAI and Microsoft is a great way to play this trend in my interview with Real Vision. Watch the Video here.
NVIDIA Omniverse: Unifying Global Industries
The conference spotlighted NVIDIA's Omniverse, a digital twin platform that allows industries to simulate, optimize, and plan their operations in a virtual environment. The platform has already been adopted by major automotive manufacturers like BMW, Toyota, and Mercedes-Benz to optimize their assembly lines, plan operations across multiple factories, and accelerate digitalization.
Collaborations with companies such as Rimac and Lucid Motors further demonstrate the versatility of the Omniverse platform in creating digital stores and facilitating virtual planning sessions for the automotive industry.
I spoke with VP of Omniverse and Development Platform, Richard Kerris, in a candid one-on-one interview last year. You can view the full-length interview here: “How to Value the Metaverse.”
In our conversation, I asked Kerris about CEO Huang’s well-publicized comment that the “Omniverse or Metaverse is going to be a new economy that is larger than our current economy.” – essentially, what I wanted to know is how can the Metaverse grow to this size considering the dominant influence the internet has on the global population?
Here is what Kerris said:
“[The Metaverse] is going to be many times bigger than the web because of what a virtual world can do for business, for education, for medical, for all sorts of things including entertainment; we’ve just begun to scratch the surface of these possibilities […] You’ve probably heard the term digital twin. One example is what it’s going to do to revolutionize the industrial market, design and manufacturing. Well, a digital twin is a true-to-reality twin in synthetic worlds of what happens in the physical world. We are seeing this transform these things because when you can make decisions in that synthetic world before you commit to it in the physical world, you have a lot of cost savings.” –Richard Kerris
New Hardware, Systems, and Cloud Services
NVIDIA unveiled its latest hardware, including the Ada RTX GPUs and the OVX servers, designed to run the Omniverse platform efficiently. The company also introduced new chips, Grace, Grace-Hopper, and BlueField-3, engineered specifically for energy-efficient focused data centers.
The tech giant announced the NVIDIA Omniverse Cloud, a fully managed cloud service in partnership with Microsoft Azure. The Omniverse Cloud will be integrated with Microsoft 365 and Azure IoT Digital Twins services, connecting hundreds of millions of Microsoft 365 and Azure users. This is a key partnership and a remarkable highlight from the event for public investors.
Spearheading the AI revolution, NVIDIA's DGX H100 AI supercomputer provides the processing power needed for mass-scale AI applications. The company expanded its business model with NVIDIA DGX Cloud, partnering with Microsoft Azure, Google GCP, and Oracle OCI to bring AI supercomputing to companies via a browser.
A few months back, I encouraged investors to stay long Nvidia through the crypto mining selloff with an editorial in September “Nvidia is Ready to Rumble with RTX 40 Series and H100 GPUs” – the stock has returned 113% in roughly 6 months and is the number one performing stock in the S&P 500.
With a focus on energy efficiency, strategic collaborations, and wide-ranging applications, NVIDIA's accelerated computing ecosystem is poised to play a significant role in shaping the future of industries globally. By combining the power of AI, digital twin technology, and strategic partnerships, NVIDIA continues to show that it will push boundaries of what's possible in technology and industry.
This year's GTC demonstrated NVIDIA's commitment to moving industries forward, improving energy efficiency, and driving innovation through AI and advanced computing technologies. As the entire world races toward digitalization, NVIDIA stands at the cutting edge, helping businesses tackle the challenges and opportunities that lie ahead.
The AI thesis is important for the long-term thesis. In the near-term, Nvidia needs to recover the $2.5 billion decline that occurred following Ethereum’s merge to Proof-of-Stake (PoS) last August. I wrote about this in an editorial “Nvidia Stock: Evidence Gaming Has Bottomed and Why It’s Important” with the conclusion “The company’s swift and concise answer to the crypto mining selloff helps illustrate why Nvidia stands apart from its peers – primarily, that its products are superior, end-market demand remains strong, and management has many levers it can pull to quickly reverse a bottom. Since this article was written, the stock has returned 67%
What’s Next for Nvidia
The I/O Fund is an actively managed portfolio. We sent a trim alert last week on Nvidia and took gains. This is a position we have managed for 5 years, building up to 15% allocation and taking gains near the top, while layering in at the bottom. You can learn more about our Research Services here and our Verified Returns here.
Given Nvidia’s recent rally after it’s Q4FY23 results, we’d thought it be helpful to provide a fundamental and technical view, as well as a brief follow up on the other two stocks we have discussed within this service – AMD, NFLX. Both of which are positioned to capture market share from their competitors. We wrote about it here, here, here and here.
As long-term investors, we firmly believe holding quality companies for at least 3 years is crucial. However, as you will see, there are times to buy, and times to take some gains. Last week we discussed why we are cautious right now based on our overall technical and macro view of the markets.
In summary, my current market outlook for 2023 has devolved since the start of the year. I have grown more bearish as time has progressed and remain rather cautious. Prompted by technical indicators that point to lower market index levels as the Fed’s fight against “supercore inflation” has proved to be difficult. Previously, I wrote about my concerns over the US consumer here.
I think that the market will provide me an opportunity to buy Nvidia, AMD and NFLX at lower levels.
First let’s take a look back at Nvidia. Below is a chart showing the buys and sells I’ve communicated to readers on a real time basis, which are moves we made in real-time. This is an example of how we actively manage a high quality position to help mitigate risk and boost returns. For reference, the percentage buys/sells are in reference to our portfolio value.
Fundamentally, we have written about the potential of AI in the past and potential beneficiaries of this secular trend. We identified Nvidia as a winner given its product offering and market position. The announcement of its H100 GPU in March 2022 to be available in 2h22 was a gamechanger. We wrote about that here.
However, in 2022 its cyclical gaming business was a significant earnings headwind. Nvidia Q3FY23 missed expectations mainly due to China related inventory write-downs, higher compensation expenses and excess inventory in the gaming channels. However, Nvidia’s gross margin guidance for Q4 suggested that the first two were isolated to Q3 and stated that gaming related inventory would be worked down in Q4. Importantly, H100 was gaining acceptance faster than expected. We wrote about it here. We bought shares the first trading day after the earnings release.
As we entered 2023, we assessed Big Tech’s capex plans and all pointed to the prioritization of AI related capex. This gave us further confidence that Nvidia was well positioned, regardless of the macro headwinds.
Nvidia recently announced Q4FY23 and full year earnings. Gross margins improved sequentially, earnings beat expectations and the Q1FY24 revenue guidance was better than consensus. Gaming showed sequential growth and the inventory situation was no longer an issue. Clear signs that earnings had bottomed in Q3. Critically, management guided to sequential growth in all 4 of its businesses and talked about the benefits of AI related demand on the company. Pointing to their March GTC conference as a key event where they will update investors.
Fundamentally, we continue to like Nvidia. Technically, after its recent rally we would wait and look for better entry points.
Regarding the technicals, we are seeing a 3 wave uptrend, which is a warning. It’s pushing higher on weakening momentum (divergence). So, big warning. I do believe that we could see one more, push we believe the time to buy is not now. Our buy zone is in green, and we must hold $138 on any drawdown, or the odds start shifting towards us seeing a fresh low.
Taking a look at AMD and NFLX.
AMD (hold)
We are seeing the same symmetrical 3 wave pattern off the October low. I can see AMD making one more run higher, but it should not be bought. I believe AMD will retest the lows in the coming months, which will set up a great buying opportunity. (hold/trim)
NFLX (hold)
Like NVDA, I believe THE low is in. However, like most stocks, the time to buy is not at these levels, unless one plans to be nimble. I do believe NFLX could set up for one more push to new highs, but after that it is due for a large retrace. We expect lower prices as we move into 2023, which will set up a great buying opportunity.
Conclusion
Given the macro backdrop, Winners and Losers will emerge within the technology sector. From a fundamental stock perspective, the team has been focusing on companies exposed to secular rather than cyclical growth with strong competitive moats.
Recent commentary from Big Tech indicates a prioritization of AI related infrastructure capex through 2023 and beyond. We continue to look for companies benefiting from AI or other themes that can withstand these macro headwinds. I believe Nvidia will emerge as a winner.
Meanwhile, AMD and NFLX are well positioned to capture market share from their competitors.
I believe that the market will provide us with better entry points in all three.
The world today was engineered to be ephemeral and noisy. This is a terrible combination for an investor.
On Twitter alone, there are 456,000 messages sent every minute. On Facebook, there are 510,000 comments posted every minute and 293,000 status updates. Outside of social media, there are 16 million text messages sent every minute and 156 million emails.
For an investor, the antidote to noise is quality stock analysis. Due diligence requires dozens of hours per equity, and it takes hundreds of hours every year to produce a free newsletter with quality analysis. I/O Fund strives to offer some of the team’s best analysis for free, and we believe the consistency and depth of what we provide for free is hard to replicate.
We offer this in the most challenging sector for investors, which is hands-down the tech sector. The tech sector is unusually challenging because it involves many different verticals – consumer, media, cloud, artificial intelligence, electric vehicles, and more. It’s also the highest risk and highest reward sector in the market. Due to sudden price movements in both directions, the stakes are high. Perhaps we are biased, but quality analysis particularly in the tech sector can be hard to come by.
Below are highlights from our free newsletter during the grueling year that was 2022. Due to the broad market being in the driver’s seat, our first few highlights review the free broad market analysis we published followed by a few strong fundamental calls.
For more information on our premium services, please click here.
Top Broad Market Highlights from I/O Fund’s Free Newsletter
The August to September Pullback:
Portfolio Manager, Knox Ridley, warned our free readers in August in the article, “Levels to Monitor in the Coming Pullback,” that the broad market failed to make a new low despite bad news and that a pullback was on the horizon. He also detailed why weakness in the bond market was coinciding with the pullback he was forecasting.
The analysis stated, “In last week’s broad market webinar, we warned our readers that a pullback was imminent. We also laid out what levels need to hold in order to confirm a new uptrend is forming. We also showed that the bond market is simply not buying what the equity market is, and that the USD pushing to new highs along with equities. These markets are simply not aligned with the current uptrend in equities, and until they are, we will remain cautious.”we warned our readers that a pullback was imminent. We also laid out what levels need to hold in order to confirm a new uptrend is forming. We also showed that the bond market is simply not buying what the equity market is, and that the USD pushing to new highs along with equities. These markets are simply not aligned with the current uptrend in equities, and until they are, we will remain cautious.”
Despite a level of exuberance in the markets following a bear market rally that formed in July, our analysis clearly stated now was not the time to buy – rather it was better to wait for the coming pullback: “These markets are warning investors that are paying attention to not get too excited, yet. No matter what scenario plays out, we do believe there will be a better opportunity to get aggressive on the long side.”
Note: For a Limited Time, I/O Fund is offering a $99/year Premium Newsletter plan that provides one actionable stock tip per month and analysis from a top performing, audited team. Click here for more detailsNote: For a Limited Time, I/O Fund is offering a $99/year Premium Newsletter plan that provides one actionable stock tip per month and analysis from a top performing, audited team. Click here for more detailsClick here for more details
Perhaps most importantly, to help with risk management, the analysis provided a long-term prediction for investors. The analysis stated: “No matter what scenario plays out, we expect around a ~10% pullback in the coming weeks, which should be followed by one more large push above SPX 4400 before the next leg lower begins.”
Carefully Timing the October Bottom:
Knox followed up with another article, “The Pullback is in Effect – Broad Market Levels – 08/26/2022”, to reiterate to our free newsletter readers that the pullback we have been warning about is in effect. The analysis provided two long term bullish scenarios. In the bullish scenario, we advised investors to buy on the second wave retrace, sometime in October.sometime in October.
Knox provided a macro update in September and discussed the key economic data points in the article “Broad Market Update: The FED versus Inflation.” The analysis points out how the FED team of experts completely missed the warning signs of data from The NAHB Index, The Case-Shiller Home Price Index, The Bloomberg Commodity Index, Crude Oil, and M2 Money Supply, pointing that inflation was a concern in 2021.
The analysis stated, “Despite the numerous market indicators pointing towards growing inflation pressures in September of 2021, the FOMC ignored the signs, and instead continue to press their loose monetary policies. They ultimately waited a year after inflation showed up to begin addressing it, putting them much farther behind the curve than investors are used to.” The aggressive increase in interest rates led to the worst stock market on record in nearly 50 years.
The analysis pointed out the similarities seen during 2021 and we hedged most of September last year to protect the portfolio from the downside risk with real-time alerts sent to members. Knox also provides regular macro updates to our premium members.
He stated: “I do believe many stocks and some markets have bottomed, and those are the ones that tend to lead going into the next uptrend […] In conclusion, we are seeing the types of extreme sentiment readings as well as divergences that mark a reversal. We are also seeing the market shrug off horrible inflation data. Since the PPI and CPI numbers came in hotter than expected, the market is up 6.5%. The last time we saw these patterns was in mid-June, just before the market moved up 18% in less than 2 months.”
Since then, we have seen a 3-month plus bounce where some stocks in our portfolio are up over 100% since we bought them around those lows. The I/O Fund portfolio manager put cash to work based on the analysis he provided at the free level. He also used more advanced analysis that he provides to our premium subscribers to help guide entries.
For example, after raising cash in mid-late August, on October 13th, we went on a buying spree within the first hour of the market open, while removing half of our hedge. We followed this up with various buys between October 14th, 18th, 21st, November 4th, 7th and 9th.
Every trade the I/O Fund makes is done through real-time trade alerts. Learn More.Every trade the I/O Fund makes is done through real-time trade alerts. Learn More.Learn More.
In detailing the October low, the I/O Fund free newsletter stressed the importance of tracking divergences.
The analysis pointed out, “We are seeing [divergences] now across bellwether stocks, varying sectors, and global markets. Many risk assets as well as global markets did not follow the S&P 500 (SPY) to new lows last week. Instead, they are signaling that a new push higher is likely to follow.”
This was partly determined by the fact transportation stocks, high beta, and small caps have been leading the markets since 2021, and when the S&P 500 made a new low, these markets made a new high. This was unique analysis that informed a critical turning point in the tough market of 2022.
Note: For a Limited Time, I/O Fund is offering a $99/year Premium Newsletter plan that provides one actionable stock tip per month and analysis from a top performing, audited team. Click here for more detailsNote: For a Limited Time, I/O Fund is offering a $99/year Premium Newsletter plan that provides one actionable stock tip per month and analysis from a top performing, audited team. Click here for more detailsClick here for more details
There’s more — the above article was followed up with more evidence in the article – “The Bear Market Rally has Much Further to Go.” when Knox stressed that the market was ignoring the fact that small caps, financials, and industrials are in notable uptrends. Even within tech, there was new leadership developing and these new leaders tend to be stocks that outperform in new uptrends.
Knox talks about patterns of major indices like the Canadian S&P/TSX Composite Index, the Japanese Nikkei 225 Index, and the Australian S&P/ASX 200 Index. The Canadian Index has a history of leading the S&P 500 Index. He stated that if all the major global markets are moving in the same direction, you are in a very powerful trend. Knox also points to some of the key levels that he monitored to confirm an uptrend for the S&P 500 Index. He said three markets, namely the Canadian Index, the Japanese Index, and the Australian Index need one more move to go higher. This pattern is very bullish for equities.
Below is a chart that shows the I/O Fund’s accurate broad market calls, particularly the prediction of a market drop in August and the bottoming in October. These calls were used alongside an automated hedge signal developed by Vincent Duchaine of The Wealth Umbrella, who is an A.I. and Machine Learning engineer.
Bitcoin:
We announced that we were buyers of Bitcoin around $16,000 and gave in-depth analysis on why a major low was likely from these levels in the article, “Bitcoin is Going to Rally Again – Here’s What You Need to Know.”
Per the article, “We are seeing more and more institutional investors, economies and businesses adopting Bitcoin. Though we are in the 4th bear cycle in Bitcoins history, the prior 3 cycles suggest where we are is a rare buying opportunity. There is ample evidence to support the $15,500 level is either a major either a major lovelow or very close to a major low. Both the technical and on-chain analysis support this.”
Bitcoin is up 36% since the article was published in December, and is just shy of the initial price target discussed in the article of $25,600. Here is what was stated: “our multifaceted analysis into Bitcoin is supporting the likelihood of a larger trend reversal. This is not confirmed from our end until we see price make that last high in the coming weeks towards the $25,600 region.”
As of now, it appears to be setting up for one more push before we see a deeper pullback. Knox updates our premium members in real-time on Bitcoin and all other portfolio positions the I/O Fund owns.
Source: YCharts
Top Fundamentals Highlights from I/O Fund’s Free Newsletter
Nvidia Stock:
The I/O Fund has an unusually strong track record on Nvidia. In fact, Beth Kindig began covering Nvidia’s product strength on artificial intelligence nearly five years agonearly five years ago. Considering it’s the best performing mega cap stock in the tech sector since the team initiated coverage – beating all FAANGs on returns — we think this is an important accomplishment.
In 2022, Beth Kindig encouraged her followers to stay long on Nvidia in August during an interview with Charles Payne on Fox Business News. Nvidia had pre-announced a Q2 2022 revenue miss of $2.5 billion due to gaming and crypto mining related weakness and the stock was tanking. The revenue miss caused the stock to sell off (8%) in one day on already weak price action of (40%) YTD.
Many pundits were questioning if Nvidia could overcome the gaming segment weakness, given Ethereum’s Merge to Proof of Stake would permanently reduce demand for gaming GPUs.
Charles Payne asked Beth Kindig if she still plans to hold the stock given the crypto mining surprise. Her answer was fairly simple: “It’s a tough day for Nvidia investors but in the long run it’s not going to matter. We hold the stock for its lead in artificial intelligence. Anything outside of that thesis is not important to us. To be contrarian, data center is going to be up 61%, so for AI investors such as myself, we are right on track.” AI investors such as myself, we are right on track.”
At the time of writing, the stock is up 39% since the interview compared to the negative (6%) for the Nasdaq-100 Index. However, most importantly, this conviction coupled with Knox Ridley’s broad market analysis caused the I/O Fund to enter Nvidia on the very day the stock bottomed for a price of $108 with a real-time alert sent to Premium customers. Below is the I/O Fund trading history on Nvidia which shows why it’s important to have conviction in tech stocks.
The Gaming Bottom:
Many thought it would take Nvidia a long time to recover from gaming, however, our analysis in September stated the company was “Ready to Rumble” and would stage a quick comeback. The free analysis stated: “Nvidia’s GeForce RTX 40 Series is perfectly timed” Nvidia’s GeForce RTX 40 Series is perfectly timed” and that the “timing of these releases is no coincidence as it’s a rapid two months following the crypto/gaming revenue miss. Suffice to say, Nvidia’s management team is prepared to rumble —- putting its very best release in gaming and its most powerful AI chip to-date up against the crypto mining selloff.”
This was important because it helped the team time the Nvidia entry at bottom, and it was this exact analysis the team depended to feel confident that the crypto mining sell-off would not take as long to absorb as many critics had forecast. Fast forward, and the most recent earnings report in February of 2023 confirmed that gaming had bottomed and was up 16% sequentially, which is what Beth’s analysis had called for a few months prior.
Notably, all of this analysis was provided for free in the I/O Fund newsletter.
Netflix:
Netflix is another hidden gem that Beth wrote about for her free newsletter readers in June. She highlighted that the market was focusing on the loss of subscribers for the stock selloff, which was a mistake, and that it was more important to look at Netflix’s plan to monetize the 100 million viewers who are sharing passwords.
Beth said, “I would argue the day that Netflix’s stock price dropped 35% was consequently one of the most important days in the company’s history in terms of its chances for a boost in revenue and a renewed uptrend. Patience, though, will be required, as Netflix has work to do (minimum one to two years for full global roll-out). Yet the path to adding more subscribers is finally clear for Netflix and will pay off long-term especially during times of inflation or muted consumer confidence as it drives down household costs across fragmented subscriptions.”
The company’s decision to start an ad-supported tier was a key highlight in the article that would drive the share price higher. “We think Netflix could set a new record on ad-supported ARPU due to its premium content and captive audience.”
The stock was down YTD 71% at the time the article was written in June for free newsletter subscribers. The stock is currently up 70% since the article was published.
Source: YCharts
While reviewing the financials of the company, Beth also noted to her readers to keep an eye on improving free cash flow as management was expecting $1 billion free cash flow in 2022 and
“substantial” free cash flow in 2023.
Note: For a Limited Time, I/O Fund is offering a $99/year Premium Newsletter plan that provides one actionable stock tip per month and analysis from a top performing, audited team. Click here for more detailsNote: For a Limited Time, I/O Fund is offering a $99/year Premium Newsletter plan that provides one actionable stock tip per month and analysis from a top performing, audited team. Click here for more detailsClick here for more details
In a premium note to I/O Fund subscribers, the team stated said the free cash flow for 2023 would likely be in the $2 to $3 billion range. Months later, the company beat its guidance by reporting a $1.6 billion free cash flow in 2022 and also provided a strong guide of a $3 billion free cash flow in 2023. This also contributed to Netflix’s strong price action of the 2022 low.
Cloud:
During the podcast Barron’s Live, Beth highlighted to her readers that cloud valuations were still trading higher than they did prior to Covid. She also pointed out that the Enterprise sector could be the next shoe to drop after the consumer sector due to budget constraints.
About two months later, this exact scenario was echoed in cloud earnings. In December, our free analysis highlighted that cloud growth rates were slowing very quickly. On average, analysts expected the top cloud companies to only grow 5% sequentially QoQ compared to 17% QoQ last year. The analysis was quite clear this was a red flag because Q4 is typically quite strong for cloud, and that this deceleration likely foreshadowed more slowing growth for 2023 once annual budgets were set in January.
Below is a chart that the I/O Fund published to premium members, however, there was coverage on the free side that pointed toward the same conclusion.
Using this analysis, the I/O Fund prudently decided to reduce the firm’s exposure to cloud. The Q1 guides would later report one of the slowest growth rates in the Cloud segment in the past decade.
Conclusion:
Last year marked the biggest destruction of wealth on record, and the tech sector was not immune to this. However, by dedicating to due diligence, the I/O Fund team was able to mitigate some of those losses with a few strong calls – not only in tech stocks – but also strong calls on where the broad market might go next.
Certainly, there were many lessons learned last year and this write-up is not intended to forego the puts and takes that all investors experienced in 2022. Rather, it’s a spotlight on how the I/O Fund strives to provide quality analysis to the community for free.
In addition, this write-up helps illustrate how the team operates behind the paywall. The team is proficient at not only product and fundamentals, but most importantly during a bear market, the team is capable of making accurate calls on what the broad market might do next. To raise the bar, the team partnered with The Wealth Umbrella on an automated hedge signal for their Premium Members. The hedge combined with buying a few high-allocation stocks near or at the lows last year is how the I/O Fund was able to mitigate losses in 2022.
Note: For a Limited Time, I/O Fund is offering a $99/year Premium Newsletter plan that provides one actionable stock tip per month and analysis from a top performing, audited team. Click here for more detailsNote: For a Limited Time, I/O Fund is offering a $99/year Premium Newsletter plan that provides one actionable stock tip per month and analysis from a top performing, audited team. Click here for more detailsClick here for more details
The I/O Fund is a publishing company. The analysis, strategies, reports, activity and all other features of our service is provided for informational and educational purposes only, and should not be construed as personalized investment advice. Hedging is an advanced method of trading stocks, sudden losses can occur, and hedging should only be pursued under the supervision of your personal financial advisor.
Note: We’ve written a lot about Nvidia this month for our Essentials Members. One thing that differentiates the I/O Fund from other services is that we are an actively managed portfolio that is audited and traded in real-time. This means, we will gladly discuss a stock many times in one month if it means we will make money.we will gladly discuss a stock many times in one month if it means we will make money.
Other services need to fill a meaningless, content pipeline. This means they will cover four, eight or even more stocks per month! This does not lead to making good money, it’s distracting and noisy, at best. Instead, we provide a strong, fundamental picture alongside thorough technicals. After a setup plays out, we will gladly tell you a stock we think is quite strong fundamentally may need to take a breather. This work is not quick, it’s not easy, and very sites are good at it. I wanted to illustrate for our Essentials Members in the Month of February how our site operates differently. We are here to make money; we are investors – not marketers. We look forward to providing more strong calls like we did with Nvidia prior to earnings. Notably, we may trim this position now. So, stay tuned next week of 2/26 for a Positions Update including a broad market stock tip from Knox. Click here for more information on our Advanced Plan. This work is not quick, it’s not easy, and very sites are good at it. I wanted to illustrate for our Essentials Members in the Month of February how our site operates differently. We are here to make money; we are investors – not marketers. We look forward to providing more strong calls like we did with Nvidia prior to earnings. Notably, we may trim this position now. So, stay tuned next week of 2/26 for a Positions Update including a broad market stock tip from Knox. Click here for more information on our Advanced Plan.Click here for more information on our Advanced Plan.
Nvidia Q4 Earnings:
Nvidia stock has reacted positively to an adjusted EPS Q4FY23 beat of $0.88 vs $0.80 and a Q1FY24 revenue guide of $6.5B vs $6.32 consensus. The most important on the call was the statement: “We expect sequential growth to be driven by each of our 4 major market platforms led by strong growth in data center and gaming”We expect sequential growth to be driven by each of our 4 major market platforms led by strong growth in data center and gaming” as it supports the H2 rebound.
Nvidia’s current quarter weakness is already priced in, yet the anticipation is high that Nvidia nails the turnaround come the October quarter. This quarter — especially with the comment on sequential growth comment across all four quarters — makes this outcome a bit more likely.
We had recently written about Big Tech’s prioritization of AI related infrastructure and how Nvidia was well positioned to benefit from this secular trend here. We listened for further evidence of this from the conference call. Big Tech capex can’t be overestimated in terms of how Nvidia will perform, and the comments about re-allocating for AI investments was further reflected in Nvidia’s report.
Financials
$6.05B revenue came in line with prior guidance and consensus estimates, up 2% sequentially and down 21% year-over-year. For the full year, total revenues came in at $26.9B, flat year over year.
Next quarter’s revenue guide was $6.5B, better than consensus of $6.32B.
Adjusted eps came in $0.88 which beat consensus of $0.80.
Within the main business segments, Data centers came in at $3.6B, down 7% sequentially, and up 11% year-over-year. Gaming revenue was $1.8B, up 16% sequentially and down 46% YoY. Importantly, gaming’s sequential improvement showed further evidence that it had bottomed. We wrote about this in November here.
Gross and adjusted gross margins came in at 63.3% and 66.1%, in-line with guidance. Operating and adjusted operating margins came in at 20.7% and 36.8%, in-line with guidance.
Both gross and operating margins showed sequential improvement as China related inventory write-downs and higher compensation related expenses were limited to Q3.
Net income improved to $1.4B vs $0.7B in the previous quarter for a net income margin of 23.3% vs 11.5%. Adjusted net income improved to $2.2B vs. $1.5B in the previous quarter for an adjusted net margin of 35.9% vs 24.5%
For the full year, total revenues came in at $26.9B, flat year over year. Within full year sales Data center sales were up 41% and gaming sales was down 27% y/y. Gross margins were 56.9% vs 64.9% the prior year while adjusted gross margins were 59.2% vs 66.8% prior. Operating margins were 15.7% vs 37.3% prior while adjusted operating margins were 33.5% vs 47.2% prior.
Regarding Gaming, Nvidia added the following. “The year-on-year decline reflects the impact of channel inventory correction, which is largely behind us.” This is important because it is exposed to the consumer and was facing cyclical headwinds last year that impacted group earnings. Rather than a detractor, it should be a contributor to group earnings going forward. Management provided a Q1 outlook compared to Q4 for all business segments. ,
“Let me look to the outlook for the first quarter of fiscal '24. We expect sequential growth to be driven by each of our 4 major market platforms led by strong growth in data center and gaming.”“Let me look to the outlook for the first quarter of fiscal '24. We expect sequential growth to be driven by each of our 4 major market platforms led by strong growth in data center and gaming.”
Looking at the balance sheet. Cash, cash equivalents and marketable securities were $13.30B. Inventory increased, primarily to support the ramp of new products in Data Center and Gaming. Meanwhile, free cash flow was $1.74B compared to $2.74B a year ago and negative $156 million a quarter ago. Fiscal-year free cash flow was $3.76B, down from $8B a year ago.
Earnings Call:
The main write-up on Nvidia’s product side will come following GTC at the end of March. However, on the call, Nvidia’s AI-as-a-service was mentioned, so I want to provide that quote for you, as it was one of the most important parts of the earnings call.
I’ve discussed in the past that the H100 is an important leap forward for enterprise AI when stating: “the A100 GPU is what led the company’s gains since Q2 2020 (detailed here) and the Hopper H100 GPU is what will lead the company’s gains for the next two years (detailed here).
The company has stated the following in regards to H100 sales:
“Adoption of our new flagship H100 center GPU is strong. In just the second quarter of its ramp, H100 revenue was already much higher than that of A100, which declined sequentially.”
I feel like I’ve talked quite a bit about the H100 and its importance, so we won’t rehash that right now.
However, per our July write-up here, there is an important point to what was discussed on the call and what Nvidia investors can expect to hear about in the coming quarters in regards to software monetization. I’m repeating here what we wrote in July before I elaborate on what I think was the most important part of the earnings call:
“According to Nvidia, the H100 delivers 9X more throughput in AI training, and 16X to 30X more inference performance. The company also states in HPC application-specific workloads, the H100 is 7X faster. The goal of the H100 was not only to add more transistors and make the H100 faster, but to also offer function-specific optimizations. This is achieved through the transformer engine.
The architecture aims to answer one of the bigger challenges facing superfast compute, which is that moving data into traditional servers overloads the CPU and system memory and becomes bottlenecked by PCI-Express.
By improving the bandwidth issue, Nvidia’s goal is to create more demand for their DGX Pod and SuperPod Systems, which in turn, will create more demand for their software.”
The comments in the earnings call that pertain to the H100 and DGX Pods and SuperPods is this – it’s important because it can mark the beginning of Nvidia’s software revenue. So, I’m including this as a bigger quote from the earnings report:
“Generative AI's versatility and capability has triggered a sense of urgency at enterprises around the world to develop and deploy AI strategies. Yet, the AI supercomputer infrastructure, model algorithms, data processing and training techniques remain an insurmountable obstacle for most […]
We are partnering with major service — cloud service providers to offer NVIDIA AI cloud services, offered directly by NVIDIA and through our network of go-to-market partners, and hosted within the world's largest clouds. NVIDIA AI as a service offers enterprises easy access to the world's most advanced AI platform, while remaining close to the storage, networking, security and cloud services offered by the world's most advanced clouds […]
AI supercomputers are hard and time-consuming to build. Today, we are announcing the NVIDIA DGX Cloud, the fastest and easiest way to have your own DGX AI supercomputer, just open your browser […]
With our new business model, customers can engage NVIDIA's full scale of AI computing across their private to any public cloud. We will share more details about NVIDIA AI cloud services at our upcoming GTC so be sure to tune in.”
The takeaway is that not only will Nvidia begin to monetize through software on the DGX systems but accessibility will improve through CSPs, or cloud service providers. This is an attempt to democratize AI development while driving software sales.
In the call, management stated the following about CSPs, or cloud service providers:
“With cloud adoption continuing to grow, we are serving an expanding list of fast-growing cloud service providers, including Oracle and GPU specialized CSPs. Revenue growth from CSP customers last year significantly outpaced that of Data Center as a whole as more enterprise customers moved to a cloud-first approach. On a trailing 4-quarter basis, CSP customers drove about 40% of our Data Center revenue.”
This is important as it links back to the comment about Nvidia’s AI as-a-service and cloud service providers helping to move DGX Cloud. It also helps to illustrate how DGX Cloud can be successful, given the strong CSP partnerships and revenue growth in the data center segment.
Here is another quote in regard to DGX Cloud and why it’ll be important for a lower barrier to entry for AI development:
“The accumulation of technology breakthroughs has brought AI to an inflection point. Generative AI's versatility and capability has triggered a sense of urgency at enterprises around the world to develop and deploy AI strategies. Yet, the AI supercomputer infrastructure, model algorithms, data processing and training techniques remain an insurmountable obstacle for most. Today, I want to share with you the next level of our business model to help put AI within reach of every enterprise customer.
Moving along, this was the Q&A piece that is most important to Nvidia investors long-term:
“Timothy Arcuri
Jensen, I had a question about what this all does to your TAM. Most of the focus right now is on text, but obviously, there are companies doing a lot of training on video and music. They're working on models there. And it seems like somebody who's training these big models has maybe, on the high end, at least 10,000 GPUs in the cloud that they've contracted and maybe tens of thousands of more to inference a widely deployed model.So it seems like the incremental TAM is easily in the several hundred thousands of GPUs and easily in the tens of billions of dollars. But I'm kind of wondering what this does to the TAM numbers you gave last year. I think you said $300 billion hardware TAM and $300 billion software TAM. So how do you kind of think about what the new TAM would be?
Jensen Huang
I think those numbers are really good anchor still. The difference is because of the, if you will, incredible capabilities and versatility of generative AI and all of the converging breakthroughs that happened towards the middle and the end of last year, we're probably going to arrive at that TAM sooner than later.”
Today, Nvidia trades at less than 1X that TAM at $515 million compared to what this analyst believes will be an easily-achieved TAM of $600 billion. This would suggest the stock price does not yet fully reflect the future market opportunity.
Conclusion:
There are some upset investors today on social media who shorted Nvidia going into the print. This was based on Nvidia’s current weak financial profile coupled with its valuation. As pointed out in our Q1 webinar, we are entirely focused on the H2 rebound, which can arguably be easier to predict with semiconductors due to the longer-term supply chain visibility this industry has. At least for today, Nvidia proved it’s on track for the H2 rebound.
As you know, we track Nvidia very closely due to its leading allocation in our portfolio. We saw evidence of a gaming bottom in November, which we published about here. We also felt Nvidia had masterfully timed it’s RTX40 Series with the Ada Lovelace architecture plus the H100 release to drop exactly when the crypto mining selloff would be most felt. We discussed this here in September. These points were entirely overlooked by Nvidia critics.
Yes, that revenue miss in the Fall was crazy – but what was lying beneath the surface for chances of a quick recovery?
Basically, the devil is in the details and not a lot of investors or analysts care to look into Nvidia’s complex hardware products. Jim Cramer got 1M views on his tweet here that admitted it was tough to listen to this particular company’s earnings calls. It works in our favor that talking heads prefer to discuss consumer tech, and that the masses are collected around those who have not taken the time to get to know his company.
We know Nvidia is not pushing a buzzword to move stock, as we’ve been covering Nvidia’s AI angle for going on five years. It’s the headlines that changed; not Nvidia.
To remain balanced here, we agree that Nvidia is likely due for a pullback. The shorts were probably right in that regard.That is Knox’s territory. He had written here that $230 has a lot of resistance and also on the forum. He plans to update everyone on Nvidia in his webinar this afternoon.
Your bigger product update will come post-GTC as we begin to lay a strong foundation for 2024 and onward for this exciting company.
