By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel consensus is bearish on SOXX's recent rally, citing fragile catalysts, stretched valuations, and potential supply responses that could reverse momentum. Key risks include a faster-than-expected supply response, hyperscaler diversification, and geopolitical export controls on critical materials.
Risk: A faster-than-expected supply response to AI demand could reset memory and CPU pricing sooner than 2029, pressuring margins even with solid demand.
Opportunity: None explicitly stated.
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 →
Industry shortages are pushing chip stocks higher.
Micron and AMD are now the two biggest components in the ETF.
Broadcom's report is testing the limits of the semiconductor rally.
May was another strong month for the semiconductor sector as chip stocks continued to rally on impressive earnings reports, excitement over the SpaceX IPO, and continued signs of shortages in components like memory chips.
Additionally, CPU stocks soared as it became clear that the shift to inference would drive surging demand for CPUs, leading to another bottleneck in semiconductors.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
The gains across the industry were strong enough that it sent the iShares Semiconductor ETF (NASDAQ: SOXX) up 23% for the month. As you can see from the chart below, the gains came in two stages.
The top holdings on the SOXX have changed quickly as the momentum in the chip sector has shifted from Nvidia to newer bottlenecks in memory and CPUs.
As of June 3, the top holdings on the ETF were Micron, followed by Advanced Micro Devices and Marvell, which jumped early in June after Nvidia CEO Jensen Huang called in the next trillion-dollar chip stock.
Micron stock nearly doubled last year, driven by ongoing reports of a shortage in memory, which some expect to last through at least 2029. At an investor conference in May, management said that its financial outlook had improved since its last earnings report, a sign that it will likely beat estimates when it reports earnings later this month. A price target hike to $1,650 from UBS also sent Micron surging, as it showed the growth potential in memory.
AMD was another big winner last month, gaining 45% after it popped early in the month on a strong earnings report, and then followed the broader gains in chip stocks in the second half of the month. AMD's revenue in the first quarter rose 38% to $10.3 billion, and adjusted net income was up 45% to $2.27 billion. Both numbers beat expectations, and its guidance topped estimates as well.
Chip stocks have gone parabolic since the start of April, as the chart below shows, and that could put pressure on future gains.
On Thursday, the SOXX fell 6% as Broadcom's earnings report sent chip stocks sliding. While the industry leader beat estimates, investors sold the stock off as management acknowledged that Google would diversify away from its custom chips, and its AI chip sales for the third quarter were projected to be $16 billion, below expectations at $17.2 billion.
Still, if competition and diversification are to blame for that, it shouldn't be a problem for the broader chip sector. If the overall narrative in the semiconductor sector remains positive, then the SOXX should keep moving higher.
Before you buy stock in iShares Trust - iShares Semiconductor ETF, 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 iShares Trust - iShares Semiconductor ETF 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 $439,632! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,316,532!
Now, it’s worth noting Stock Advisor’s total average return is 959% — a market-crushing outperformance compared to 210% 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 June 4, 2026. *
Jeremy Bowman has positions in Advanced Micro Devices, Broadcom, Micron Technology, and Nvidia. The Motley Fool has positions in and recommends Advanced Micro Devices, Broadcom, Marvell Technology, Micron Technology, Nvidia, and iShares Trust-iShares Semiconductor ETF. 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
"The current rally may be overextended risk-wise, and a normalization in AI demand or memory pricing could trigger a meaningful pullback even if fundamentals remain positive."
SOXX's 23% May surge mirrors a shift from AI mega-cap leadership to bottleneck plays in memory and CPUs, with MU and AMD driving much of the gains. Yet the rally rests on fragile catalysts: Broadcom’s softer guidance hints not all chip segments are thriving; memory shortages cited through 2029 could loosen if supply responds faster than expected; AI capex optimism may have baked in a high bar for durable earnings upgrades. Valuations look stretched for a cycle that could see a demand normalization. If data-center demand cools or pricing for memory chips improves, the momentum could reverse quickly.
The rally could prove durable if AI-related capex remains robust and memory shortages persist longer than peers expect; the risk is underappreciated by the article that treats the move as broadly solid across semis.
"The semiconductor sector is transitioning from a scarcity-driven bull market to a competitive diversification phase that will compress margins for high-flying components."
The SOXX 23% rally in May reflects a classic 'bottleneck' trade, where capital rotates from GPU-monopolist Nvidia into secondary beneficiaries like Micron (memory) and AMD (CPUs/inference). While the narrative of structural shortages through 2029 is compelling, the market is currently pricing in perfection. Broadcom’s recent guidance miss—specifically the $1.2 billion shortfall in AI chip revenue—is a flashing red light for the sector's valuation multiples. When hyperscalers like Google begin diversifying their silicon sourcing, the 'indispensable' moat of these chipmakers narrows. I expect significant volatility as investors realize that supply-side constraints are being met with aggressive capex-driven supply expansion, risking a cyclical glut by 2026.
If AI inference demand scales exponentially as projected, the current supply bottlenecks will persist far longer than historical cyclical patterns suggest, justifying current premium P/E multiples.
"Broadcom's guidance miss on AI chips and Google's chip diversification signal that the AI-driven semiconductor shortage narrative is narrower and more competitive than the article suggests, making a 23% monthly rally unsustainable without proof of earnings growth to match."
The article conflates a 23% monthly rally with sustainable momentum, but the Broadcom miss—where Google diversifies away from custom chips and Q3 AI revenue guidance trails expectations—is the real story being buried. Yes, memory shortages and CPU demand are real, but the article doesn't address valuation. After a parabolic move since April, SOXX is pricing in years of flawless execution. Micron's $1,650 UBS target assumes memory stays scarce through 2029; any supply normalization breaks that thesis. AMD's 45% May gain on strong Q1 is impressive, but the article ignores whether Q2 guidance already bakes in the rally.
If AI inference truly is a multi-year bottleneck and memory scarcity persists through 2029 as claimed, then current valuations may be justified, and Broadcom's miss could be an outlier (Google is one customer) rather than a sign of broader demand destruction.
"Parabolic semiconductor gains since April plus Broadcom's guidance miss point to near-term downside risk for SOXX despite ongoing shortages."
The article frames SOXX's 23% May surge as driven by memory and CPU bottlenecks plus AI demand, with Micron and AMD now topping holdings after Nvidia's earlier run. Yet the parabolic move since April, capped by Broadcom's 6% post-earnings drop on $16B Q3 AI guidance missing $17.2B estimates, signals fragility. Google's diversification away from custom chips and UBS's $1,650 Micron target reflect stretched expectations. Shortages may linger to 2029, but rapid rotation among chip sub-sectors and any macro slowdown could trigger sharp reversals not addressed in the bullish narrative.
Persistent memory shortages and AMD's 38% revenue growth could sustain momentum into 2026 if AI inference demand accelerates faster than diversification fears.
"A faster supply response could break the scarcity thesis and force earlier multiple compression, despite current demand optimism"
Gemini flags a valuation risk, but the bigger overlooked risk is a faster-than-expected supply response to AI demand. If hyperscalers front-load capex and new fabs come online sooner, memory and CPU pricing could reset sooner than 2029, pressuring margins even with solid demand. Diversification hurts moats, but it also accelerates capacity growth and could trigger a cyclical downturn by 2026.
"Geopolitical supply chain constraints on raw materials pose a greater threat to the AI capex thesis than hyperscaler diversification or cyclical oversupply."
Gemini and Claude focus on hyperscaler diversification as a margin threat, but you're all ignoring the geopolitical risk premium. If China's export controls on gallium or germanium tighten further, the 'supply-side expansion' everyone expects will hit a wall regardless of fab capacity. The market is pricing in a frictionless supply chain that simply doesn't exist. We aren't just looking at a cyclical glut; we are looking at a potential raw material bottleneck that could break the entire AI capex thesis.
"Geopolitical material constraints matter less than demand-pull acceleration collapsing the shortage thesis faster than consensus expects."
Gemini's gallium/germanium export control risk is real but overstated as a near-term brake. SOXX exposure to rare-earth constraints is minimal—most AI chip bottlenecks are silicon, not specialty materials. The actual risk nobody's flagged: if Broadcom's miss signals hyperscalers are already pulling forward capex into Q2-Q3, the 'shortage through 2029' narrative collapses within 12 months, not 2026. That's the timing mismatch worth watching.
"Material export curbs could extend bottlenecks even if silicon supply accelerates."
Claude underplays the gallium/germanium angle by calling it niche. These materials feed directly into GaN and GaAs components for power delivery and RF in AI servers, where even modest restrictions could delay ramps that hyperscalers are already front-loading. That linkage turns Claude's 12-month normalization scenario into a higher-risk bet rather than a clean offset to memory shortages.
The panel consensus is bearish on SOXX's recent rally, citing fragile catalysts, stretched valuations, and potential supply responses that could reverse momentum. Key risks include a faster-than-expected supply response, hyperscaler diversification, and geopolitical export controls on critical materials.
None explicitly stated.
A faster-than-expected supply response to AI demand could reset memory and CPU pricing sooner than 2029, pressuring margins even with solid demand.