In July 2026, the semiconductor sector underwent a sharp correction. In the early hours of July 8 (UTC), all three major US stock indices closed lower: the Nasdaq fell 1.16% to 25,818.69, the Dow Jones dropped 0.25% to 52,925.15, and the S&P 500 declined 0.45% to 7,503.85. The Philadelphia Semiconductor Index, which tracks overall chip stock performance, plunged 4.65% to 12,300.52, breaking below its 50-day moving average and hitting its lowest close since June 10. On the individual stock level, Intel tumbled over 9%, AMD fell more than 6%, Micron dropped over 4%, while Nvidia bucked the trend, rising 0.71% to $196.93.
This sell-off was not an isolated event. On July 1, the Philadelphia Semiconductor Index plummeted 6.27% in a single day; on July 2, it fell another 5.44%, resulting in a cumulative decline of over 11% in just two trading sessions. South Korea’s KOSPI Index also plunged about 7.9% on July 2, prompting the Korea Exchange to activate its sell-side circuit breaker. Despite Samsung Electronics reporting a Q2 operating profit that surged more than 18 times year-over-year to 89.4 trillion KRW on July 7, the result only slightly beat market expectations of 87.3 trillion KRW, falling short of investors’ lofty hopes.
Yet, as market panic spread, major Wall Street institutions—including Goldman Sachs, JPMorgan, Bank of America, UBS, and Morgan Stanley—almost simultaneously issued statements, sending a unified message: the semiconductor correction does not mark the end of the AI rally, but rather opens a new window for strategic positioning.
Nature of the Correction: Profit-Taking and Valuation Adjustment, Not Demand Collapse
Understanding the nature of this correction is essential for predicting future trends.
Fundamental data shows the AI demand cycle is far from over. In a July 7 industry research note, JPMorgan revealed that global semiconductor sales reached $131.9 billion in May 2026, up 16.1% month-over-month—significantly higher than the historical average seasonal increase of 4.5%. Year-over-year, industry sales grew 118.8%. Even if growth in the second half of the year merely follows historical seasonal patterns, global semiconductor revenue for 2026 is still expected to rise over 90% year-on-year, reaching $1.5–1.6 trillion. According to the latest forecast from the World Semiconductor Trade Statistics (WSTS), the global semiconductor market is projected to hit $1.51 trillion in 2026.
Demand remains robust. JPMorgan strategist Mislav Matejka stated in a July 6 client note that the semiconductor upcycle is far from over, and "meaningful new supply is unlikely to arrive before 2028." Memory chip manufacturers—including Micron, SK Hynix, and Samsung—have already sold out their high-bandwidth memory (HBM) supply through 2026, with new wafer capacity expected to come online only after 2028. AI data centers are forecast to consume about 70% of global memory chip output this year.
Bank of America’s July 8 semiconductor industry report noted that the recent correction in semiconductor stocks is a normal market adjustment, not a sign of weakening AI demand. History shows that semiconductor stocks often consolidate in the summer, with profit-taking and valuation corrections paving the way for a new rebound in the fall. The bank remains optimistic about the long-term AI semiconductor cycle, believing the industry is still in the middle of an 8- to 10-year growth period.
UBS Asset Management commented on July 6 that, despite ongoing volatility, semiconductor-related stocks are not in a bubble. UBS cited strong demand signals—such as an eightfold increase in weekly AI token consumption since the start of the year—as key evidence supporting its positive outlook.
Overall, the current correction is driven by three main factors: outsized gains earlier in the year (the Philadelphia Semiconductor Index surged over 100% in the first half), profit-taking (unwinding crowded trades and deleveraging), and a market reassessment of valuations (the PHLX Semiconductor Index is up over 80% year-to-date, raising earnings expectations). These forces are at play, rather than any structural deterioration in AI demand.
From "Buying the Sector" to "Stock Selection": A Fundamental Shift in AI Investment Logic
This marks the most significant change in institutional perspectives this cycle.
Over the past two years, investors largely adopted a "basket buy" strategy for the semiconductor sector—buying GPU companies, chip manufacturers, and semiconductor equipment firms, all of which delivered substantial excess returns. However, Goldman Sachs’ July 7 report made it clear that AI chip trading has entered a more selective phase, and investors should no longer simply buy the entire sector.
The rationale is straightforward: the PHLX Semiconductor Index has risen over 80% this year, far outpacing the S&P 500 and Nasdaq. This strong performance raises the bar for future earnings and makes the risk-reward profile ahead of Q2 earnings season more differentiated. In other words, the market has shifted from "concept trading" to "earnings trading."
Goldman Sachs puts it bluntly: a blanket "buy the dip" strategy carries risks. The bank favors CPU, ASIC, memory, and equipment stocks aligned with AI growth, naming AMD and Applied Materials as top picks, but remains cautious on smartphone supply chain players and semiconductor companies with high valuations or weaker demand.
JPMorgan takes a slightly different view. The bank maintains an "overweight" stance on the semiconductor sector, believing that AI accelerated computing, memory, and networking equipment supply chains will continue to benefit most directly from the cycle. However, JPMorgan warns that the valuation gap between AI semiconductor makers and major cloud service providers has reached unsustainable levels. Semiconductor stocks are up 87% this year, while the "Magnificent Seven" tech giants have dropped 7% from their yearly highs—a divergence the market must address.
Bank of America is more focused on undervalued opportunities. Analyst Vivek Arya points out that memory chips have a forward P/E of just 10x, which is severely undervalued given their growing share in AI infrastructure spending. The bank recommends overweighting industry leaders like Nvidia, Broadcom, Lam Research, and KLA.
UBS offers another warning: increased capital expenditures could pressure the cash flow of hyperscale cloud providers in the second half of 2026. As investors demand greater capital discipline, this may exert downward pressure on semiconductor and AI hardware valuations.
Quick Take: Four Major Wall Street Institutions’ Views
| Institution | Core View | Preferred Direction |
|---|---|---|
| Goldman Sachs | AI chips have entered a selective phase; "basket buy" not recommended | CPU, ASIC, memory, equipment; names AMD, Applied Materials |
| JPMorgan | Correction offers entry opportunity; upcycle to last at least through 2028 | Maintains "overweight" on semiconductors; AI accelerated computing, memory, networking equipment |
| Bank of America | Industry is mid-cycle in an 8–10 year growth period; correction is healthy | Nvidia, Broadcom, Lam Research, KLA; reiterates "buy" on Micron |
| UBS | Short-term volatility offers long-term entry; sector not in a bubble | Positive on semiconductors; recommends selective AI investment |
| Morgan Stanley | AI’s long-term trend intact, but capital may rotate from chip stocks to cloud computing | Hyperscale cloud providers (Microsoft, Amazon, Meta) |
Source: Compiled from July 2026 research reports by each institution
Where Are the Next AI Investment Opportunities?
If the market is shifting from "sector-wide rally" to "selecting leaders," what are the criteria? Synthesizing institutional views, opportunities can be dissected across three layers of the value chain.
Chip Segment: Compute Power Remains the Most Certain Direction
Goldman Sachs’ favored areas—CPU, ASIC, and memory—are essentially the foundational layer of AI compute infrastructure. GPUs generate intelligence, HBM and DRAM enable high-speed data transfer, while enterprise NAND and SSDs handle hot data and caching. Institutions like Goldman Sachs believe the AI compute arms race led by cloud giants is turning memory chips from cyclical products into scarce strategic assets. DRAM and NAND price hikes in 2026 are not the end, but likely the start of a supercycle.
TrendForce forecasts Nvidia will hold about 64% of the global AI chip market in 2026, with AMD at roughly 8.6%. On July 5, Goldman Sachs raised AMD’s 12-month price target from $450 to $640, maintaining a "buy" rating—a clear endorsement of the long-term value of the AI chip sector.
Infrastructure Segment: Extending Value from Chips to Cloud
Morgan Stanley offers another critical perspective. Chief equity strategist Michael Wilson noted in a July 6 report that momentum in the semiconductor sector is fading, and investors are turning to AI hyperscale cloud providers that have lagged this year, including Microsoft, Amazon, and Meta. Wilson’s team argues that semiconductor growth ultimately depends on capital spending by cloud giants, and the current divergence between chip stocks and cloud leaders may not persist.
JPMorgan forecasts the global semiconductor equipment market will reach $159 billion in 2026, up 28% year-over-year, growing to $205 billion in 2027, and $237 billion in 2028. Long-term procurement plans by leading global firms lock in sustained demand, and the semiconductor equipment sector is set for a strong two-year upcycle from 2026 to 2027.
AI Applications Segment: Commercialization Loops Are Taking Shape
Morgan Stanley’s equity investment division highlights that the industry has now established a positive cycle: "continuous large model iteration → steady enterprise revenue growth → increased AI capital spending." Business models are gradually proving themselves. With explosive growth in token demand driven by coding and AI agents, large models and cloud providers are entering their next phase of accelerated revenue growth.
Bank of America expects global cloud and AI infrastructure capital expenditures to reach $1.5 trillion by 2027. This scale of spending means future AI winners may not just be chip companies, but the entire infrastructure ecosystem—from compute chips to storage devices, networking hardware to data center operations, and from foundational hardware to upper-layer software applications—all stand to benefit from this long-term trend.
Conclusion
In summary, Wall Street’s collective optimism about AI semiconductors rests on a clear logical chain: the AI demand cycle is far from over (global semiconductor market is set to surpass $1.5 trillion); supply constraints persist (new capacity unlikely before 2028); and capital expenditures continue to expand (cloud and AI infrastructure spending projected to reach $1.5 trillion by 2027).
However, this does not mean the semiconductor sector will repeat the broad-based gains of the past two years. The consensus among Goldman Sachs, JPMorgan, Bank of America, UBS, and Morgan Stanley is: the AI theme remains intact, but the market has moved from concept trading to earnings trading. Future differentiation will depend on companies’ ability to deliver profits, their share of AI revenue, and their strategic position within the broader infrastructure ecosystem.
For investors, the current correction may offer a chance to reassess portfolio structure—shifting from "buying the sector" to "stock selection," and from chasing beta to uncovering alpha. This is both a challenge and the starting point for a new round of strategic positioning.
FAQ
Q1: What are the main reasons behind the current semiconductor correction?
The correction is primarily driven by three factors: outsized gains prompting profit-taking, unwinding crowded trades and deleveraging, and a market reassessment of high valuations. Both JPMorgan and Bank of America note this is a normal market adjustment, not a structural change in AI demand. New capacity is expected to come online around 2028, and industry supply-demand dynamics remain healthy.
Q2: How do Wall Street institutions view the long-term outlook for AI chips?
Goldman Sachs, JPMorgan, Bank of America, UBS, and others generally agree the long-term trend for AI chips remains intact. JPMorgan expects the semiconductor upcycle to last at least through 2028; Bank of America sees the industry mid-cycle in an 8- to 10-year growth period; UBS says the sector is not in a bubble; Morgan Stanley believes the long-term AI trend persists, though capital may rotate from chip stocks to cloud computing.
Q3: What does the shift from "buying the sector" to "stock selection" mean for AI investing?
It means the market is no longer bullish on the entire semiconductor sector indiscriminately, but is starting to distinguish genuine AI beneficiaries from mere followers. Goldman Sachs explicitly advises against continuing the "basket buy" approach. Future investment opportunities will focus on companies with solid profitability, high AI revenue share, and those poised to benefit from ongoing AI infrastructure expansion.
Q4: What are the key directions for the next phase of AI semiconductor investment?
Institutions generally favor the following: on the chip side, CPU, ASIC, GPU, and HBM memory; on the infrastructure side, semiconductor equipment, data centers, and cloud computing; and on the application side, enterprise AI software and AI agents. Bank of America expects global cloud and AI infrastructure capital expenditures to reach $1.5 trillion by 2027.
