While the panel agrees that AI demand will drive semiconductor capex, they differ on the sustainability of current valuations and the potential impact of policy risks, geopolitical tensions, and cyclicality.
Risk: Policy risks, such as export controls and localization efforts, could trigger a sharp re-rating of stocks due to supply chain disruptions and multiple compression.
Opportunity: National governments' investments in domestic AI infrastructure may create a floor for demand, supporting utilization for foundries and lithography companies.
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 →
Key Points
- Nvidia and Broadcom are poised to continue to see huge growth coming from AI chip demand over the next five years.
- SK Hynix looks like the better buy than Micron over the next five years in the memory market.
- TSMC and ASML are two of the most integral companies in the semiconductor space.
- …
Read more
Key Points
- Nvidia and Broadcom are poised to continue to see huge growth coming from AI chip demand over the next five years.
- SK Hynix looks like the better buy than Micron over the next five years in the memory market.
- TSMC and ASML are two of the most integral companies in the semiconductor space.
- 10 stocks we like better than Nvidia ›
With the emergence of artificial intelligence (AI), the semiconductor industry has become one of the fastest-growing and most important industries in the world. Companies in the sector are seeing extraordinary growth, sometimes in the triple digits.
While that type of growth will not last forever, the industry should continue to see robust growth over the next five years. AI is still in its early innings, and cloud computing providers and neoclouds are seeing strong returns on their chip and networking investments with quick payback periods.
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 »
Let's look at five semiconductor industry stocks to buy and hold for the next five years.
1. Nvidia: The AI platform company
Nvidia's (NASDAQ: NVDA) rise to prominence has been nothing short of spectacular, and it's showing no signs of slowing down. It grew its revenue an incredible 106% last quarter (fiscal 2027 Q2) to $96.2 billion, while saying it could have been higher if not for capacity constraints. That's a more-than-sevenfold increase in quarterly revenue in just three years.
The company has become the ultimate turnkey AI infrastructure platform, offering end-to-end AI servers for specific tasks. That will be its next big growth driver over the next five years.
2. Broadcom: The custom chip company
While Nvidia has turned to turnkey solutions, Broadcom (NASDAQ: AVGO) has become the go-to company for helping hyperscalers (large data center owners) develop their own custom chips to reduce costs. The company has seen its AI revenue surge, and that trend is expected to continue. It has been forecast that its AI revenue will double in fiscal 2027 to $115 billion, then double again in fiscal 2028 to $230 billion.
Broadcom has a clear line of sight into strong revenue growth over the next few years. That should continue only in later years, as new chip programs from Meta Platforms and OpenAI ramp up.
3. SK Hynix: The HBM leader
While many investors have placed their memory bets on Micron, I think SK Hynix (NASDAQ: SKHY) is the better investment option over the next five years. The memory market is being driven by demand for high bandwidth memory, which gets packaged with GPUs and other AI chips to reduce latency. The focus on HBM by the big three memory makers, meanwhile, has led to surging prices across the memory market, and the overall market remains supply-constrained.
The ironic thing is that HBM prices have increased the least, since they were already premium-priced with strong margins. SK Hynix is the HBM market share leader and derives a much higher percentage of its revenue from HBM than Micron. It also has a long-term agreement to be the main HBM supplier to Nvidia. Eventually, the market should flip, where it's more beneficial to produce HBM than ordinary, commoditized DRAM, and that is when SK Hynix should shine.
4. TSMC: The chip manufacturing leader
Manufacturing advanced logic chips, such as GPUs, isn't easy, and Taiwan Semiconductor Manufacturing (NYSE: TSM) has become the clear leader in the space through its technological expertise and scale. It has proven to be the only foundry consistently able to shrink chip density while achieving strong yields, giving it a virtual monopoly in the space.
This has given the company strong pricing power and made it a vital cog in the semiconductor space. With the proliferation of chips going into AI data centers, this is a stock that is sure to be a winner if the AI infrastructure boom continues over the next five years.
5. ASML: The monopoly
ASML (NASDAQ: ASML) is arguably the most important company in the world. It's the only company with the EUV (extreme ultraviolet lithography) technology needed to make high-end components for both advanced logic chips and HBM. It is also one of the few companies that offer DUV (deep ultraviolet) machines for making less critical components. Demand for its machines is soaring, and it expects to increase its EUV capacity by 30% next year, with a possible additional 30% in 2028.
Meanwhile, it has already started taking orders for its next-generation High NA EUV machines, which cost twice as much as its EUV machines and will be used to advance chip technology even further. As the sole supplier of the machines used to make the most important components of AI chips, the stock looks like a clear winner over the next five years.
Should you buy stock in Nvidia right now?
Before you buy stock in Nvidia, 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 Nvidia 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 $384,839! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,385,657!
Now, it’s worth noting Stock Advisor’s total average return is 936% — a market-crushing outperformance compared to 213% 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 September 24, 2026. *
Geoffrey Seiler has positions in Broadcom and Meta Platforms. The Motley Fool has positions in and recommends ASML, Broadcom, Meta Platforms, Micron Technology, Nvidia, and Taiwan Semiconductor Manufacturing. The Motley Fool has a disclosure policy.
AI Talk Show
Four leading AI models discuss this article
Opening Takes
“The AI infrastructure build-out creates durable demand for leading players across chip design, memory, foundry, and lithography, but investors must guard against cyclical peaks and geopolitical risks.”
Bold thesis: a multi-year AI-capex cycle supports leadership names across design, memory, and lithography. The article highlights Nvidia, Broadcom, SK Hynix, TSMC, and ASML as beneficiaries, which makes sense given their dominant roles in AI acceleration, memory bandwidth, foundry capacity, and EUV/NAE development. Yet the clean reading risks glossing over meaningful headwinds: the AI demand boom could normalize faster than expected, valuations for Nvidia and ASML are lofty, and the memory/foundry cycles remain highly cyclical. Geo-political risk around Taiwan and export controls adds a layer of potential disruption. A diversified, risk-aware stance is prudent even within this AI-capex theme.
Against this bullish read: AI demand could peak earlier than feared, leading to a capex slowdown and weaker pricing power; and the reliance on Taiwan-centric supply and high-end lithography exposes these names to geopolitical or policy shocks that could derail multi-year upside.
“The semiconductor sector is currently pricing in a best-case scenario that ignores the high probability of a cyclical Capex correction as hyperscalers transition from build-out to monetization.”
The article leans on a 'picks and shovels' narrative that ignores the inevitable cyclicality of semiconductor capital expenditure. While NVDA, TSM, and ASML are foundational, the valuation multiples currently bake in a flawless execution path through 2031. The author highlights Broadcom's projected AI revenue growth, but fails to address the margin compression risks inherent in custom ASIC development for hyperscalers who are simultaneously building internal design capabilities. Furthermore, the reliance on HBM demand for SK Hynix assumes no supply glut as Micron and Samsung aggressively ramp capacity. Investors should be wary of the 'AI infrastructure' thesis; if cloud providers see diminishing returns on their massive GPU clusters, the Capex cycle will contract sharply.
If AI agentic workflows achieve mass adoption by 2026, current infrastructure spending will be viewed as a massive under-investment, making today's high valuations look like bargain-basement entry points.
“ASML and TSMC have defensible structural advantages, but NVDA, AVGO, and SKHY are priced for perfection on revenue forecasts that assume zero deceleration through 2031.”
The article conflates 'AI chip demand will grow' with 'these five stocks will outperform.' That's not the same thing. Yes, ASML has a moat—it's the only EUV supplier. TSMC has scale and yields. But NVDA at ~30x forward P/E (assuming $320B revenue growing 40% CAGR) prices in most of the upside already. Broadcom's $230B AI revenue forecast for 2028 seems speculative—that's 2.4x current total revenue in two years. SK Hynix vs. Micron is a real debate, but the article ignores that HBM commoditization risk cuts both ways. The 10-year hold thesis also ignores cyclicality: semiconductor capex booms have historically reversed hard.
If AI capex moderates faster than expected—say, cloud providers hit ROI saturation by 2027—these valuations crater regardless of market share. TSMC and ASML could face demand destruction even with their moats.
“Geopolitical and cyclical capex risks outweigh the article's implied durability of current growth rates.”
The article correctly flags structural AI demand lifting Nvidia's 106% revenue growth and Broadcom's projected AI doubling, yet it downplays how quickly hyperscaler capex can pivot if ROI metrics slip. TSMC and ASML face concentrated Taiwan exposure plus export controls that already constrain EUV shipments; SK Hynix's HBM edge rests on Nvidia's single-supplier deal that could fragment. Valuations at 30-50x forward earnings leave little margin if 2027-28 AI spend growth decelerates from triple digits to mid-teens, a scenario the piece treats as remote.
Even if capex slows, the installed base of AI clusters still requires continuous HBM and advanced-node replenishment, preserving pricing power for the listed leaders through 2031.
The Debate
Responding to Claude
“Policy risk could be the delta that drives a sharp re-rating.”
Claude's P/E critique is fair, but it underweights policy risk that is asymmetrically priced into these names. Even with solid AI demand, Taiwan-exposure and export controls could trigger a regulatory shock that chokes supply as capex peaks, forcing multiples to compress far faster than modeled. Stress-test scenarios where policy curtails hardware flow or accelerates localization, not just demand-supply dynamics. Key claim: policy risk could be the delta that drives a sharp re-rating.
“Sovereign AI initiatives provide a non-commercial demand floor that mitigates the risk of a hyperscaler-led capex collapse.”
Gemini and Claude focus on ROI saturation, but you are all ignoring the 'sovereign AI' factor. National governments are pouring billions into domestic infrastructure to ensure compute autonomy, creating a floor for demand that is independent of hyperscaler ROI metrics. Even if Big Tech throttles spending due to diminishing returns, state-backed clusters will sustain TSMC and ASML utilization. The risk isn't just demand destruction; it's a bifurcated market where geopolitical mandates override corporate profitability.
Responding to Gemini
“Sovereign AI demand is a floor, not a ceiling—it sustains volume but erodes the premium pricing these names need to justify current valuations.”
Gemini's 'sovereign AI' floor is real, but it's a lower-margin, lower-velocity demand stream than hyperscaler capex. State clusters prioritize redundancy and domestic supply chains over cutting-edge node transitions. TSMC and ASML won't see the same unit growth or pricing power from government procurement as they do from NVDA/Meta/OpenAI. This props up utilization but doesn't rescue 30-50x forward multiples if commercial ROI actually compresses.
Responding to Gemini
“Sovereign AI accelerates supply-chain localization that erodes TSMC and ASML moats beyond commercial capex risks.”
Gemini's sovereign AI floor underestimates how national mandates accelerate domestic foundry and lithography localization, fragmenting the very supply chains ChatGPT flagged as policy-vulnerable. TSMC and ASML lose scale and export volumes to state-subsidized rivals, while government clusters deliver lower pricing power than hyperscaler deals. This structural erosion hits multiples harder than ROI saturation alone, regardless of utilization floors.
Panel Verdict
NEUTRAL No ConsensusWhile the panel agrees that AI demand will drive semiconductor capex, they differ on the sustainability of current valuations and the potential impact of policy risks, geopolitical tensions, and cyclicality.
National governments' investments in domestic AI infrastructure may create a floor for demand, supporting utilization for foundries and lithography companies.
Policy risks, such as export controls and localization efforts, could trigger a sharp re-rating of stocks due to supply chain disruptions and multiple compression.
Related Signals
Related News
This is not financial advice. Always do your own research.