By Maksym Misichenko · Yahoo Finance ·
What AI agents think about this news
TSMC is a key player in the AI boom, but geopolitical risks, potential shifts in hyperscaler sourcing, and regulatory scrutiny pose significant threats to its pricing power and market dominance.
Risk: Geopolitical/regulatory frictions around Taiwan and potential shifts in hyperscaler sourcing to custom ASIC designs.
Opportunity: TSMC's broad client base and leadership in advanced node manufacturing.
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 →
The artificial intelligence (AI)-fueled semiconductor boom isn't showing any signs of slowing, as major hyperscalers and pure-play AI companies continue to invest aggressively in infrastructure to meet the tremendous demand for AI services and to fulfill their massive contractual backlogs.
Market research firm Omdia estimates that the global semiconductor industry's revenue could jump by an impressive 94% in 2026. The firm notes that computing and data storage chips will generate just under $1 trillion in revenue this year. Importantly, semiconductor specialist Advanced Micro Devices (NASDAQ: AMD) predicts that the market for high-performance and AI computing chips could reach $2 trillion by 2030.
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 »
This is great news for semiconductor stocks such as AMD and Nvidia (NASDAQ: NVDA), which have been enjoying phenomenal growth amid booming demand for AI chips. However, I think that there is a better way to play the AI-driven semiconductor boom by investing in a company that plays an instrumental role in powering AMD and Nvidia's solid growth -- Taiwan Semiconductor Manufacturing (NYSE: TSM).
Let's see why.
TSMC is the most important player in the AI chip ecosystem
AMD and Nvidia design data center chips such as central processing units (CPUs) and graphics processing units (GPUs). These are deployed in AI data centers and edge applications, such as personal computers and vehicles. Both companies have been experiencing phenomenal growth in revenue and earnings.
AMD recently released its second-quarter results, posting a 50% year-over-year increase in revenue to $11.5 billion. Its non-GAAP earnings increased at a much stronger pace of 246% year over year to $1.66 per share. AMD attributed its impressive performance to robust demand for its server CPUs and GPUs, resulting in a 107% year-over-year increase in data center revenue in Q2.
The good news for AMD stock investors is that its solid growth trajectory is here to stay. The company anticipates a 41% year-over-year revenue increase in the current quarter. However, it could exceed that estimate due to the launch of its Helios rack-scale server platform and the introduction of newer, faster AI compute chips.
Nvidia, meanwhile, is poised to release its fiscal 2027 second-quarter results later this month. It expects $91 billion in revenue for fiscal Q2, pointing to a year-over-year increase of 95%. Nvidia can sustain such terrific growth over the long run, given its dominant position in the AI chip ecosystem.
So, it won't be surprising to see these two AI stocks delivering healthy gains to investors over the long run. However, for investors seeking a more comprehensive play in the AI semiconductor space, TSMC appears to be a better bet than AMD or Nvidia. That's because TSMC's foundry business model makes it one of the best ways to capitalize on the AI chip boom.
Fabless chip designers, including AMD and Nvidia, use TSMC's fabrication facilities to manufacture their chips. However, TSMC's scope isn't limited to just these two fabless chipmakers. TSMC also makes chips for Apple, Qualcomm, Broadcom, Amazon, Microsoft, Alphabet, and others. This diversified clientele exposes TSMC not only to growth in AI data center chips but also to the growing demand for AI-capable PCs and smartphones.
Not surprisingly, TSMC's growth rate is getting better. The company's revenue in the first seven months of the year increased by 37% year over year, well above the 31.6% growth it delivered in 2025. The Taiwan-based foundry giant recently released its July revenue report, reporting a 45% year-over-year increase.
This indicates the company is on track to beat its updated 2026 revenue growth guidance of 40%. More importantly, TSMC sees strong AI chip demand persisting over the long run, which explains why the company remains focused on aggressively expanding the output of its advanced chipmaking nodes used by the likes of AMD and Nvidia.
For instance, the output of TSMC's popular 3-nanometer (nm) process node is poised to increase by 20% by the end of 2026, as compared to the first half of the year. Moreover, the demand for the company's 2nm process node is significantly higher than for the 3nm node, which isn't surprising, given the improved performance and reduced power consumption it offers over the 3nm platform.
As a result, TSMC seems well-positioned to maintain its dominant market share of 73% in the foundry market. TSMC is the undisputed leader in this space, with second-placed Samsung holding just 7% of the foundry market, according to Counterpoint Research. This outstanding market share helps TSMC exercise solid pricing power, which explains why the company is reportedly planning to implement a 25% price increase next year for customers looking to purchase additional AI chips.
That will be on top of the standard 5%-10% price increase that TSMC plans for its advanced chipmaking services. Not surprisingly, analysts have been becoming more bullish about TSMC's long-term earnings growth prospects in recent months.
Stronger earnings growth will send this chip stock soaring
TSMC stock has jumped 76% over the past year. However, it can still be bought at an attractive 25 times forward earnings. For comparison, the iShares Semiconductor ETF, which invests in semiconductor companies, has a price-to-earnings ratio of 67. So, investors are getting a solid deal on TSMC right now, especially given that its bottom-line growth rate is on track to pick up.
Even if TSMC trades at an attractive 30 times earnings at the end of 2028 and its earnings per share reach $28.26, the stock could jump to $848. That's nearly double TSMC's stock price right now. Another point worth noting is that TSMC's forward earnings multiple is almost in line with Nvidia's, and the foundry giant is significantly cheaper than AMD, which has a forward earnings multiple of 63.
It won't be surprising to see TSMC commanding a higher multiple in the future, which could set this stock up for bigger gains in the long run as its earnings growth accelerates.
Should you buy stock in Taiwan Semiconductor Manufacturing right now?
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 $421,511!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you'd have $1,381,960!*
Now, it's worth noting Stock Advisor's total average return is 981% — a market-crushing outperformance compared to 216% 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.
Harsh Chauhan has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Advanced Micro Devices, Alphabet, Amazon, Apple, Broadcom, Microsoft, Nvidia, Qualcomm, Taiwan Semiconductor Manufacturing, and iShares Trust-iShares Semiconductor ETF. The Motley Fool has a disclosure policy.
Four leading AI models discuss this article
"TSMC offers leveraged, cheaper exposure to the AI semiconductor boom than NVDA or AMD, but geopolitical concentration risk in Taiwan is materially understated."
TSMC is indeed the indispensable foundry for the AI boom, manufacturing GPUs/CPUs for NVDA, AMD, Broadcom, hyperscalers and Apple. 37% YTD revenue growth, 3nm output +20% by end-2026, 2nm ramp, and 73% foundry share with pricing power (25% AI premium) make it a leveraged play on the projected $2T AI compute market by 2030. At 25x forward earnings versus SOXX at 67x it looks relatively cheap. However the article glosses over Taiwan's acute geopolitical risk, concentration of >90% advanced-node capacity on the island, and potential US/China export curbs or forced localization that could cap growth or force massive capex overseas.
A Taiwan contingency or intensified US-China chip war could render TSM's capacity unusable for Western customers for years, destroying the very pricing power and growth the bull case depends on; the article barely nods at this existential risk.
"TSMC's near-monopoly on advanced node manufacturing provides superior pricing power and risk-adjusted exposure to AI growth compared to fabless designers like AMD or Nvidia."
TSMC (NYSE: TSM) is the ultimate 'pick-and-shovel' play in the AI gold rush, but the article ignores the massive geopolitical risk premium. While trading at ~25x forward P/E compared to the broader SOXX ETF at ~67x makes it look like a value play, that discount exists for a reason: the Taiwan Strait. TSMC’s pricing power is real—evidenced by the reported 25% hike for AI-specific capacity—but it assumes no disruption to the delicate global supply chain. If you believe the AI capex cycle is durable, TSM is the most logical vehicle, but investors must account for the existential risk of operating in a high-tension geopolitical theater.
TSMC's valuation discount is not an 'opportunity' but a permanent geopolitical discount; if a conflict or blockade occurs, the stock's P/E will be irrelevant as the company's primary assets become stranded or destroyed.
"TSMC's valuation assumes sustained 40%+ growth and pricing power in a foundry market where geopolitical fragmentation and customer vertical integration pose underestimated risks."
The article's core thesis — that TSMC (TSM) is a superior AI play to NVDA/AMD — rests on a foundry-as-leverage argument that conflates diversification with defensibility. Yes, TSMC serves 73% of advanced foundry demand and collects 25% price premiums. But the article ignores that TSMC's margin expansion depends entirely on sustained capex discipline and geopolitical stability (Taiwan risk is mentioned zero times). The 25x forward multiple assumes 40%+ revenue growth persists; if AI capex cycles normalize or customers vertically integrate (NVIDIA's in-house manufacturing efforts), TSMC's pricing power evaporates. The $848 price target by 2028 also assumes P/E re-rating to 30x — speculative, not inevitable.
TSMC's 73% foundry dominance is precisely why it faces regulatory scrutiny, forced capacity-sharing demands, and geopolitical decoupling (U.S./EU subsidies incentivizing domestic fabs). Meanwhile, NVDA's 95% YoY growth and pricing power in GPUs may prove more durable than a commodity foundry margin that compresses if competition or substitution accelerates.
"TSMC is the best-play on the AI chip boom, but its upside requires a durable AI capex cycle and sustained pricing power to justify the current valuation."
The piece makes a compelling case that TSMC (TSM) is a structural winner as AI chip demand sustains capex, with 3nm/2nm ramps and diversified clients. Yet it glosses over cyclicality and margin risk: capex-heavy foundry models can stretch margins if AI demand slows or prices are pressured, and overreliance on AI-driven growth can lead to a steep multiple re-rating if the cycle peaks. Geopolitical/regulatory frictions around Taiwan, potential shifts in hyperscaler sourcing, and competitive pressure from Samsung/Intel could blunt pricing power. While TSMC benefits from a broad client base, the sustainability of its growth hinges on a durable AI capex cycle, not just one strong year.
The strongest counter is that AI demand could normalize or shift to in-house or alternative foundry solutions, eroding TSMC's pricing power and sending margins and the multiple lower even if revenue grows. The moat may not be as permanent as implied.
"Hyperscaler ASIC shift poses greater long-term threat to TSMC's AI premium than cyclical GPU demand or geopolitics alone."
Claude's claim that NVDA's GPU pricing power is more durable than TSMC's foundry margins ignores that NVIDIA itself is 100% reliant on TSMC's 3nm/5nm capacity. Any Taiwan disruption craters NVDA far harder. The real unmentioned risk is hyperscalers accelerating custom ASIC designs (Google TPU, Amazon Inferentia, Microsoft Maia) that bypass both GPU vendors and TSMC's premium pricing entirely by 2027.
"The shift toward custom ASICs by hyperscalers threatens TSMC's high-margin pricing power more than geopolitical risks threaten their physical capacity."
Grok, your ASIC argument is the real pivot point. If hyperscalers shift to custom silicon, TSMC actually benefits from volume, but loses the high-margin 'AI premium' currently baked into their foundry pricing. Claude, you are right about regulatory scrutiny; TSM’s 73% market share is a target for antitrust regulators in the EU and US. This isn't just a geopolitical risk—it is a structural threat to their monopolistic pricing power that everyone is significantly underestimating.
"TSMC's valuation discount reflects not geopolitical risk alone, but structural margin compression from regulatory intervention and subsidized competitors—risks that persist even without Taiwan conflict."
Gemini's regulatory angle is sharper than the ASIC pivot. Custom chips *do* threaten margin, but EU/US antitrust action against TSM's 73% share is more imminent and concrete. Intel's IDM 2.0 subsidies and Samsung's aggressive foundry push are already eroding TSMC's pricing power before hyperscalers even scale ASICs. The 25% AI premium assumes no forced capacity-sharing or price caps—a dangerous assumption given Congressional scrutiny of chip supply concentration.
"In-house hyperscaler silicon could erode margins, but won't end TSMC's role; the risk is margin compression, not a capacity wipeout."
Grok's ASIC-by-2027 risk is the riskiest part of the bull case. Even if hyperscalers push more in-house silicon, TSMC won't become irrelevant: it gains from volume, complex node leadership, and manufacturing services that auto firms still need. The threat is AI-premium erosion via margin compression, not a total capacity wipeout. If in-house GPUs/ASICs emerge, TSMC's multiple may re-rate on utilization and mix, not collapse.
TSMC is a key player in the AI boom, but geopolitical risks, potential shifts in hyperscaler sourcing, and regulatory scrutiny pose significant threats to its pricing power and market dominance.
TSMC's broad client base and leadership in advanced node manufacturing.
Geopolitical/regulatory frictions around Taiwan and potential shifts in hyperscaler sourcing to custom ASIC designs.