By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
TSMC's strong Q2 results and Q3 guidance signal robust demand for advanced nodes, particularly 2nm, driven by AI demand. However, geopolitical risks, capital intensity, and potential margin volatility pose significant challenges.
Risk: Geopolitical risks, including Taiwan tensions and potential US export controls, could strand advanced-node IP and experienced engineers in the US, making Arizona fabs uncompetitive.
Opportunity: TSMC's strategic role in the US supply chain could lead to guaranteed margins via 'Buy American' mandates for defense and critical infrastructure.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Taiwan Semiconductor Manufacturing (NYSE:TSM) reported second-quarter earnings on Thursday, again demonstrating the dominance of the globe’s leading foundry. Revenues were up 33% from a year ago to $40.2 billion.
But I think investors should be even more excited about the future -- TSMC, as it’s known, projected that its revenue growth will accelerate in the third quarter. Management’s guidance forecasts revenue between $44.6 billion and $45.8 billion in the third quarter. At the midpoint of that range, that would represent 37% growth over the $31.1 billion TSMC recorded in the third quarter of 2025.
Missed Nvidia in 2009? This Rare Signal Is Flashing Again. In 2009, a "Double Down" signal flashed for a little-known chipmaker called Nvidia. For the first time in years, that same "Total Conviction" signal is flashing for a company 1/100th the size of Nvidia. Continue »
And full-year revenue would increase by 40% from a year ago, management said.
It’s really hard for a company the size of TSMC to be accelerating revenue growth. After all, we’re talking about a company with a market cap of $2.1 trillion, making it the sixth-largest publicly traded company in the world.
Image source: TSMC.
But rather than slowing after a solid second quarter, TSMC expects revenue growth to quicken in Q3, suggesting that the demand for advanced chips -- and TSMC’s expert foundry capabilities -- remains exceptionally strong.
TSMC has tremendous reach -- the company produced more than 12,680 different products in 2025, using 305 different process technologies. It has an estimated 73% market share in the global foundry market and makes chips for the biggest chipmakers on the planet, including Nvidia, Advanced Micro Devices, and Broadcom.
Nearly two-thirds of TSMC’s revenue comes from manufacturing chips 5 nanometers and smaller -- and that business is changing dramatically. In the first quarter, TSMC got 36% of its revenue from 5 nm chips and 25% from 3 nm chips. But in this quarter, TSMC saw its first revenue from even more advanced 2 nm chips, reportedly priced at $30,000 per wafer. Revenue from 3 nm chips increased from 25% to 30%, and 5 nm chip revenue dropped to 33%.
“Our business in the second quarter was supported by strong demand for our leading-edge process technologies,” Chief Financial Officer Wendell Huang said. “Moving into third quarter 2026, we expect our business to be supported by continued strong demand for our leading-edge process technologies, including the steep ramp-up of our 2-nanometer technology.”
Management announced it will invest another $100 billion in the company’s Arizona foundries to increase 2 nm production there, bringing its total commitment in the state to $265 billion.
While artificial intelligence has dominated the stock market over the last three years, this year is seeing greater emphasis on agentic AI -- programs that can operate independently to execute tasks without human intervention.
TSMC CEO C.C. Wei told analysts during the company’searnings callthat creating the hardware to power agentic AI will be profitable for his company. “The emergence of agentic AI is leading to a resurgence in the role of CPUs in AI data centers, which drive more silicon demand in addition to AI accelerators,” he said. “We believe this is positive for TSMC, as no matter what CPU approach is taken, whether it's an x86, Arm-based, or RISC-V architecture, they are almost all TSMC's customers.”
The biggest takeaway here is that TSMC had a solid quarter, but management expects things to get even better. The $2-trillion company is projecting accelerated revenue growth, fueled by the ramp-up of its 2 nm chips. The growth of agentic AI provides another tailwind for TSMC stock, as demand for its manufacturing capacity will remain strong.
Investors have plenty of reasons to be bullish about TSMC stock through 2026.
Before you buy stock in Taiwan Semiconductor Manufacturing, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Taiwan Semiconductor Manufacturing wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $397,351! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,304,257!
Now, it’s worth noting Stock Advisor’s total average return is 937% — a market-crushing outperformance compared to 211% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of July 16, 2026. *
Patrick Sanders has positions in Nvidia. The Motley Fool has positions in and recommends Advanced Micro Devices, Broadcom, Nvidia, and Taiwan Semiconductor Manufacturing. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
Four leading AI models discuss this article
"TSMC's accelerating growth into 2nm and agentic AI tailwinds justify a re-rating above 28x forward P/E through 2026 despite geopolitical overhang."
TSMC's Q2 revenue of $40.2B (+33% YoY) and Q3 guidance implying ~37% growth signal robust demand for 3nm/2nm nodes, with 2nm wafers priced at ~$30k driving margin expansion. Agentic AI boosting CPU silicon alongside accelerators is a credible tailwind given TSMC's 73% foundry share and exposure to NVDA, AMD, AVGO. The extra $100B Arizona capex (total $265B) diversifies from Taiwan risk. At a $2.1T market cap, sustaining 40% full-year growth is impressive but valuation at ~28x forward P/E already prices in much of the AI cycle.
Geopolitical risk around Taiwan remains acute; a China conflict or export curbs could halt production regardless of Arizona ramps. 2nm yields may slip versus 3nm, while AI capex from hyperscalers could moderate in 2027 if ROI on agentic systems disappoints, exposing TSMC's high fixed costs.
"TSMC's ability to capture the entire spectrum of AI compute—from accelerators to CPUs—makes it the only pure-play infrastructure bet for the agentic AI era."
TSMC’s 33% revenue growth and bullish Q3 guidance highlight its near-monopoly on high-end logic. The shift toward 2nm and the resurgence of CPU demand for agentic AI architectures provide a clear runway for margin expansion, as 2nm wafers command significant pricing power. However, the market is severely underpricing geopolitical risk and the massive capital intensity required for the Arizona expansion. While the revenue growth is impressive, the $265 billion total commitment in Arizona creates a massive drag on free cash flow and introduces significant execution risk regarding labor costs and operational efficiency compared to their Taiwan-based fabs.
The massive $265 billion capital expenditure in Arizona could lead to long-term margin compression if domestic production costs significantly exceed the efficiency of TSMC’s Taiwan-based ecosystem.
"Revenue acceleration is real but the article mistakes supply-constrained growth for demand-driven growth—TSMC's massive capex suggests they're racing to meet bottlenecks, not riding a structural wave."
TSMC's 33% YoY revenue growth and management guidance for 37% Q3 growth is genuinely impressive at $2.1T market cap. The 2nm ramp is real—first revenue this quarter, priced at $30k/wafer, with mix shift showing 3nm climbing to 30% of revenue. Agentic AI demand for CPUs alongside accelerators is a legitimate second-order tailwind. However, the article conflates acceleration with sustainability. Q3 guidance assumes continued 37% growth, but TSMC's own capex ($100B Arizona alone) suggests management sees supply constraints as the limiting factor, not demand. The article ignores geopolitical risk: Taiwan tensions, US export controls on advanced nodes to China, and potential fab redundancy if competitors succeed in catching up. Also missing: gross margin trajectory under 2nm ramp (historically new nodes compress margins initially) and whether $265B Arizona investment dilutes returns.
TSMC's guidance could reflect peak demand before AI capex normalizes and margin compression from 2nm production ramp erodes the earnings growth story beneath the revenue headline.
"TSMC could sustain above-market growth if the 2nm ramp and AI-driven capex persist, but execution risk and pricing power will determine whether this is a true multi-year upcycle or a temporary burst."
TSMC's finger on the AI compute pulse suggests a mid- to high-teens growth path for leading nodes into Q3, with the 2nm ramp as the linchpin. The article is bullish, but the path is hinge-laden: 2nm yields, customer mix, and the cost of a $265B Arizona program. Even if Q3 hits guidance, margins remain murky as wafer pricing (~$30k) and heavy capex press FCF in the near term. Key risks: AI demand normalization, competitive pressure from Samsung/Intel, geopolitical/regulatory frictions involving Taiwan and the US, and concentration risk with marquee clients. A durable upcycle requires not just demand but efficient ramp and pricing power.
The strongest counterpoint is that the acceleration could be a back-end backlog unwind or favorable mix rather than a durable demand surge, and the massive Arizona capex may weigh on cash flow if AI spending cools.
"Arizona capex adds subsidy-tied execution risk that amplifies—not mitigates—geopolitical exposure."
Nobody has flagged that Arizona's $265B isn't pure diversification—US CHIPS Act subsidies are clawed back if export controls tighten further. A real Taiwan contingency would still strand advanced-node IP and experienced process engineers. 2nm yield risk compounds this: even modest slippage could make the subsidized US fabs permanently uncompetitive versus any surviving Taiwan capacity.
"The massive Arizona investment transforms TSMC from a pure foundry into a protected strategic utility, likely securing long-term pricing power through US government mandates."
Grok and Gemini are fixated on the $265B Arizona capex as a financial drag, but they ignore the weaponization of the supply chain. This isn't just about diversification; it's about TSMC becoming a strategic US utility. If the US government subsidizes this, they essentially guarantee a floor for TSMC’s margins via 'Buy American' mandates for defense and critical infrastructure. The risk isn't margin compression—it's the regulatory capture of TSMC’s entire global pricing power.
"Arizona's value to TSMC is inversely correlated with geopolitical stability—the safer the world becomes, the worse the investment performs."
Gemini's 'regulatory capture' framing is sharper than the drag narrative, but both miss the core tension: US subsidies *require* export restrictions to justify politically. TSMC's margin floor depends on sustained geopolitical friction—the moment US-China tensions ease, those subsidies evaporate and Arizona becomes a stranded asset. The real risk isn't margin compression; it's margin volatility tied to policy whiplash, not fundamentals.
"2nm ramp is supply-chain constrained, which could cap margins even if demand remains robust."
Claude makes a solid point on margin volatility, but the bigger overlooked lever is the supply-chain/tooling bottleneck for 2nm. Ramp timing hinges on EUV tool availability, mask capacity, and specialty chemistries. If ASML/AMAT capacity or lead times push 2nm yields and output later than guidance, ASPs and gross margin could disappoint in the near term, even with strong demand and subsidies.
TSMC's strong Q2 results and Q3 guidance signal robust demand for advanced nodes, particularly 2nm, driven by AI demand. However, geopolitical risks, capital intensity, and potential margin volatility pose significant challenges.
TSMC's strategic role in the US supply chain could lead to guaranteed margins via 'Buy American' mandates for defense and critical infrastructure.
Geopolitical risks, including Taiwan tensions and potential US export controls, could strand advanced-node IP and experienced engineers in the US, making Arizona fabs uncompetitive.