TMTB Morning Wrap
Good morning. QQQs +30bps on a relatively quiet August morning. Yields and Crude are flattish.
Overnight, Asia was mixed: TPX +0.51%, NKY +0.59%, Hang Seng -1.1%, HSCEI -1.02%, SHCOMP +0.01%, Shenzhen +0.33%, Taiwan TAIEX -0.46%, Korea KOSPI +2.42%. Hynix +6% and Smasung +4.5% strong on the back SNDK’s analyst day yesterday which focused investors on FCF and potential EPS in 2030 (we dove into the topic in our EOD wrap here)
In Tech, RDDT +11% after getting added to SP500, SNDK +6.6% as Harlan at JPM finally upgraded, AEHR +8.5% on a positive initiation. AMAT -5% after numbers missed elevated bogeys and GM guide was flat. HDDs strong with memory. Semis +0.5% as Hardware +1.5% continues to rally while Software is flat
Lots of good stuff to get to, so let’s get to it. We’ll cover AMAT earnings first, then onto the usual…
AMAT: Ok #s with F3Q revenue/EPS of $9.115B/$3.50 vs. Street $9.020B/$3.42 and guided F4Q to $10.25B/$4.02 vs. Street $9.617B/$3.72, but flat GM and below buyside growth the nits as investors wanted LRCX-like sequential growth and more GM leverage.
Overall numbers looked fine and mgmt said ustomer visibility has improved and expects another strong growth year in 2027, supporting the longer-term WFE/semicap bull case, but #s came in a bit light of buyside which was closer tot $10.5B for the Oct guide. Net/net a bit weaker, but nothing negative thesis changing.
The #s:
F3Q revenue grew +24.8% y/y versus +11.4% last quarter, with record Foundry/Logic, strong NAND and AGS, and DRAM below Street largely on timing.
Mgmt raised its CY26 Semi Systems outlook beyond the prior >30% target and expects JanQ revenue to grow q/q, making >40% systems growth plausible but not formally guided.
Leading-edge Foundry/Logic, DRAM and advanced packaging should drive roughly 80% of WFE growth in both 2026 and 2027; advanced packaging is now expected to grow >70%, AGS >20%, and PDC >50% in CY26.
F4Q GM is guided flat at 50.4% despite +12.5% q/q revenue growth, as display mix and capacity-ramp hiring offset the favorable Systems mix and volume.
Key Quote:
Duration of Cycle:
“...the large customers, we actually have visibility to the roadmap. So we have a perspective on probably five years of visibility for our largest customers. We asked them for the detail at a detailed level for the eight quarters that we've been speaking about so we can aggregate that and get it to our supply chain. And other things that have changed, we get longer lead-time POs from our customers, so that the details are agreed from a longer lead-time perspective..So there has been a number of changes, and I think visibility is significantly increased from prior periods.”
Key Takeaway
Demand. The underlying demand message was materially stronger than the headline guide. Customer CapEx forecasts are rising, most leading-edge logic and DRAM fabs are running at full capacity, and customers announced more than 10 additional fab projects during the quarter. Mgmt raised CY26 Semi Systems growth beyond the prior >30% outlook, reiterated that AMAT should outgrow WFE and gain share, expects sequential growth in JanQ, and described 2027 as another strong year.
Bar. F4Q revenue guidance is objectively strong at +50.7% y/y and +12.5% q/q, but it fell short of heightened buyside expectations after LRCX had guided to roughly 20% sequential growth. The combination of slower relative growth, no precise CY26 systems target and flat GM explains why the stock sold off MSD.
AI mix. The composition of WFE growth is favorable for AMAT. Leading-edge Foundry/Logic, DRAM and advanced packaging are expected to generate approximately 80% of industry growth in both 2026 and 2027, and mgmt sees AMAT holding leadership positions across all three. Advanced-packaging growth was raised to >70% in CY26, while PDC is expected to grow >50%, driven by additional eBeam and optical inspection intensity.
DRAM. F3Q DRAM revenue of $1.830B, +6% q/q and roughly +50% y/y, was well below Street $2.150B. Mgmt nevertheless expects a “very significant” step-up in 2H26 as customers obtain cleanroom capacity, and continues to see strong growth into 2027. AMAT is also positioned across HBM packaging, epitaxy, conductor etch, wiring and patterning, with longer-term exposure to 6F², 4F² and 3D DRAM architectures.
Margins. F3Q non-GAAP GM of 50.4% and OM of 34.0% both beat Street, but F4Q GM is guided flat despite the substantial revenue increase. Mgmt attributed this to lower-margin display mix, customer-support and manufacturing hiring, and other ramp costs. The important nuance is that mgmt expects only slow improvement over the next few quarters, although the hiring headwind should recede as revenue scales.
Pricing. AMAT’s systematic value-based pricing remains a meaningful structural lever. Mgmt now assesses and reprices every tool based on customer value, has higher pricing and margins across both new and existing products, and has expanded company GM by approximately 300 bps over three years. Semiconductor Systems GM is already above 55%, although continued product improvement and pricing are currently being partly absorbed by higher input and ramp costs.
Services. AGS revenue was $1.781B, +22% y/y and above Street $1.762B; mgmt now expects >20% CY26 growth and reiterated a sustainable mid-teens long-term rate. More than 37,000 chambers are connected to AMAT’s AIx capabilities, supporting predictive maintenance, chamber matching, yield optimization and higher-value contracts. The caveat is that this year also benefits from utilization moving toward 100%, so spares growth will naturally normalize, although mgmt remains unusually positive on both AGS top-line growth and profitability.
ICAPS. This was one of the more important narrative improvements. Mgmt now expects ICAPS to grow in both 2026 and 2027, versus the previous digestion narrative, supported by 28nm investment in China, rising utilization and improving power-semiconductor and photonics demand outside China. NAND should also grow, but mgmt expects it to remain the slowest-growing major end market next year because customer activity is still focused more on layer upgrades than greenfield wafer capacity.
China. China represented 26% of Semiconductor Systems plus AGS revenue and approximately 28% of consolidated revenue, depending on the denominator. Mgmt now expects China revenue to grow in both 2026 and 2027, led by differentiated 28nm Foundry/Logic products. That is incrementally positive for estimates, but it also increases scrutiny around export controls and emerging Chinese competitors in mature-node deposition and related ICAPS applications.
Visibility. AMAT’s largest customers are providing detailed rolling eight-quarter forecasts, longer-lead-time purchase orders and, in some cases, cancellation and expedite charges. Mgmt has approximately five years of roadmap visibility for its largest customers, with technology co-innovation discussions extending even further. This is much better planning visibility than in prior cycles, although roadmap discussions beyond eight quarters should not be treated as firm orders. No new customer-concentration metric was disclosed.
Capacity. AMAT has nearly doubled manufacturing space over the past several years, added more than 1,500 manufacturing and AGS-support employees during F3Q, and is preparing to double quarterly Systems output capacity by 2028, with further optionality for 2030. Mgmt explicitly cautioned that this is manufacturing readiness, not a 2028 revenue forecast. CapEx should remain above normal next year as EPIC and manufacturing investments continue, but decline as a percentage of revenue.
Products. AMAT announced six new systems spanning DRAM and advanced packaging, expanded its eBeam position into packaging process control, and now has 11 announced EPIC engagements. Panel-level packaging revenue should grow meaningfully in 2027 and ramp further thereafter, while hybrid bonding provides another multi-year opportunity across Foundry/Logic, DRAM and HBM. The October EPIC opening and investor breakfast could provide the market with a more explicit long-term financial framework.
Bull vs. Bear Debate
The broad bull thesis is that AMAT is not simply riding a cyclical WFE recovery; the mix of industry spending is shifting toward the areas where AMAT has its highest process intensity and strongest positions. Leading-edge Foundry/Logic, DRAM and advanced packaging are expected to account for roughly 80% of WFE growth in 2026 and 2027. AMAT participates across deposition, etch, CMP, epitaxy, eBeam, HBM packaging, hybrid bonding and emerging panel architectures, giving it multiple ways to increase content per wafer and per advanced package. Its large installed base adds a recurring AGS stream, while PDC and AI-enabled services provide a second source of higher-value growth beyond new equipment shipments.
This quarter strengthened the bull case because visibility, breadth and duration all improved. The CY26 Systems outlook was raised for a third time, customer plans added more than 10 new fabs, ICAPS shifted back toward growth, advanced packaging moved above 70%, and mgmt described 2027 as another strong year. Five-year customer roadmaps, eight-quarter tool-level forecasts, longer-lead-time orders and cancellation protections reduce planning risk and give AMAT greater confidence to build capacity. If the October investor event provides credible long-term revenue and margin targets, the market could begin underwriting the capacity build as evidence of a durable multi-year cycle rather than late-cycle overinvestment.
The other major bull lever is earnings conversion. Semiconductor Systems GM is already above 55%, value-based pricing is operating across both new and installed products, and the hiring burden should fade as revenue catches up with newly added resources. A mix shift toward Systems, advanced packaging, PDC and higher-value service offerings could lift company GM toward 51%–52% and drive stronger operating leverage than the F4Q guide currently implies. In that scenario, estimate increases would come not only from a higher WFE denominator, but also from share gains and better incremental margins.
The bear thesis starts with expectations and relative performance. AMAT delivered a very strong absolute guide, yet +12.5% q/q growth was not enough after LRCX’s roughly +20% cadence, and the stock’s reaction shows that the debate has moved beyond whether WFE is strong. Investors now need evidence that AMAT can materially outgrow WFE, not merely keep pace with it. The F3Q DRAM miss, refusal to explicitly commit to >40% CY26 Systems growth and lack of formal JanQ guidance leave room for AMAT to finish the year only modestly ahead of the market. Its broader portfolio also means less relative exposure to a strong NAND recovery than LRCX, while AMAT’s GM and FCF profile remains less attractive.
Margins are the most immediate bear evidence from the quarter. F4Q revenue is rising more than 12% sequentially, but GM is flat, and mgmt said ramp-related hiring will continue for several quarters with only slow improvement. Display growth is dilutive, AGS is receiving a temporary benefit from customer utilization moving toward 100%, and the larger manufacturing footprint is driving elevated CapEx and working-capital needs. If WFE growth normalizes before that capacity is fully utilized, AMAT could be left with a higher fixed-cost base and less operating leverage than bulls expect.
China creates a second structural risk. The near-term return to growth helps estimates, but AMAT remains exposed to future export restrictions, geopolitical volatility and increasingly capable domestic equipment competitors in legacy PVD, CVD and other ICAPS markets. The underlying WFE cycle also remains dependent on concentrated customer CapEx, memory profitability and cleanroom availability. A slowdown in AI infrastructure spending, customer project delays or stronger-than-expected domestic Chinese share gains could hit both revenue and the valuation multiple while AMAT is still spending heavily to add capacity.
What Changed This Quarter
The long-term bull case strengthened. Customer visibility became more tangible, the 2027 outlook improved, advanced packaging and PDC targets increased, ICAPS turned positive, China improved, and AMAT committed resources to support substantially higher output. The call also provided stronger evidence that value-based pricing and AI-enabled services are durable, company-specific earnings levers rather than merely WFE beta.
The near-term bear case also gained evidence. The guide did not clear the buyside bar, the DRAM result missed Street, mgmt declined to quantify CY26 Systems growth, and GM conversion was underwhelming relative to the revenue acceleration. The capacity build and associated hiring also create a larger execution burden.
TECH RESEARCH/NEWS
RDDT +11%: Will be added to S&P 500 prior to theopen on 8/18, replacing AVB
SNDK: JPM Initiates Overweight with $2,250 PT as LTAs Reset NAND Economics and AI Inference Expands the TAM
JPM initiates SNDK at Overweight with a Dec-27 price target of $2,250, saying SanDisk’s New Business Model contracts have structurally lifted the margin floor and reduced NAND cyclicality. JPM says eight signed NBMs represent roughly $94B of contract value at floor pricing, with >50% of FY27 bits and about two-thirds of FY28 bits covered, supporting roughly 80% gross margins and ~50% adjusted FCF margins through the cycle. JPM also says AI inference, persistent KV-cache demand, BiCS10 and High Bandwidth Flash should drive a step-change in NAND demand, while 100% of excess FCF returned to shareholders could further amplify EPS growth.
SNDK: Morgan Stanley Reiterates Overweight as 100% FCF Return and LTAs Support a Structurally Higher Margin Floor
Morgan Stanley reiterates Overweight after SanDisk’s Investor Day, saying the biggest positive was management’s commitment to return 100% of excess FCF through buybacks while maintaining disciplined capex. Morgan Stanley remains somewhat more cautious than management on how structural the margin reset is, but says NAND shortages, stronger AI/data-center demand and long-term agreements should keep conditions favorable for several years. Morgan Stanley also highlights SanDisk’s capital-efficient BiCS roadmap and HBF as additional upside, while arguing NAND’s growing role in AI should complement rather than displace DRAM.
FOXA/ROKU: Wells Fargo Upgrades FOXA to Overweight and Raises PT to $80 as World Cup, ROKU Upside and NFL Clarity Improve the Setup
Wells Fargo upgrades FOXA to Overweight from Equal Weight and raises its price target to $80 from $65, saying the shareholder-return story is improving alongside stronger World Cup economics and upside from ROKU. Wells Fargo raises FY27 FOXA EBITDA to roughly $4.1B, +7% versus prior, and says World Cup revenue could reach ~$800M versus >$600M previously, while Wells Fargo also raises ROKU estimates and expects sales synergies from the FOXA integration. Wells Fargo says the lack of meaningful NFL contract changes until the 2030 season removes an important revision overhang and supports a higher valuation multiple.
NVDA: UBS Reiterates Buy and $280 PT as Blackwell/Rubin Demand Supports Another Estimate Raise
Keep reading with a 7-day free trial
Subscribe to TMT Breakout to keep reading this post and get 7 days of free access to the full post archives.