U.S. stocks: IBM plunges 25%, nearly 70 billion in market cap wiped out—why are chip stocks surging across the board?

On July 14, 2026, the U.S. stock market saw an exceptionally divisive trading day. On one side, tech giant IBM suffered an epic sell-off—its share price plunged 25.21%, logging the largest single-day drop since the 1987 “Black Monday”; on the other side, chip stocks celebrated in unison—NVIDIA rose more than 4%, Micron climbed nearly 5%, and SK Hynix ADR surged by more than 27%.

In the same market, at the same time, the price action moved in starkly opposite directions. This wasn’t random emotional volatility, but a clear structural signal: the flow of AI capital expenditures is shifting from software and the services end, toward large-scale hardware infrastructure.

How a preliminary earnings report triggered IBM’s historic plunge

IBM’s plunge was driven by a second-quarter preliminary earnings report that was released early. A full week before the company’s scheduled release of its official earnings report, it unusually issued a performance warning to the market.

Based on preliminary data, IBM’s second-quarter revenue was about $17.2 billion, up only 1% year over year—far below analysts’ consensus expectations of $17.86 billion to $17.9 billion. Adjusted earnings per share were forecast at $2.93, also below the market expectation of $3.01.

Looking at business segments, software revenue grew 5% year over year, well below the market’s expected 11%; consulting was essentially flat, rising only 1% on a constant-currency basis; and the infrastructure segment became the biggest drag, with revenue down 7% year over year—far worse than the company’s earlier expectation of a “low single-digit decline.” Among them, the infrastructure segment that includes the mainframe business was hit the hardest, and sales of IBM’s next-generation z17 mainframe—an initiative the company had high hopes for—missed expectations.

After the news was released, IBM’s stock closed at $217.07, down $73.16 on the day, a 25.21% drop. Full-day trading value expanded to $14.8 billion. This decline not only set the company’s biggest single-day drop since it went public in 1915, it also surpassed the 23.7% plunge during the October 19, 1987 “Black Monday” stock-market crash. The company’s market capitalization evaporated by about $69 billion to $70 billion in a single day.

Why customer budgets suddenly shifted from software to hardware

In a letter to investors, IBM CEO Arvind Krishna unusually acknowledged the company’s difficulties. He said bluntly that in the last few weeks of June, corporate customers suddenly and significantly adjusted the direction of their capital expenditures—redirecting budgets that had originally been planned for software and mainframes on a large scale into hardware purchases such as servers, storage devices, and memory.

The driving force behind customer behavior is expectations. Corporate customers expect that AI-related servers, storage devices, and memory will rise in price and that supply will become increasingly tight, so they want to lock in access to constrained infrastructure before prices increase. Krishna said, “We expected some impacts on the supply chain, but we did not anticipate the magnitude of the reordering of capital expenditures.”

This is not a strategic mistake by IBM itself or a flaw in its products; rather, it reveals a structural change across the entire industry. With budgets limited, enterprise CFOs and CIOs make rational choices—when AI deployment becomes the top priority, money inevitably flows into AI servers, data center storage, and high-speed memory, while procurement of traditional software and mainframes gets squeezed out.

Morningstar analyst Luke Yang noted that this has formed a new trend of “hardware eating everyone’s lunch.”

Why chip stocks became the biggest winners of this “budget shift”

In sharp contrast to IBM’s dire situation, the chip stocks that IBM effectively “pointed at” enjoyed a collective celebration.

The Philadelphia Semiconductor Index closed up 2.54% on the day. By individual stock: NVIDIA closed up 4.06% to $211.80; Micron surged 4.92% to $983.12; SanDisk rose 5.01%; Intel rose 4.50%; and AMD gained 2.57%.

The standout performer was storage-chip giant SK Hynix. Its U.S. depositary receipts (ADR) jumped 27.29% in a single day to close at $193.92, hitting a new high since listing. The ADR premium of SK Hynix versus its domestically listed shares in South Korea soared to 51%, far above the roughly 3% issuance premium at last week’s IPO. Market capitalization reached $1.36 trillion.

SK Hynix ADR’s surge was also boosted by the launch of listed options. Its options began trading on the U.S. Options Exchange this Tuesday. In the near term, demand for call options was strong, with the most active contract being a call option with a strike price of $185. At the same time, SK Hynix announced that it has officially started mass production shipments to NVIDIA of 12-layer HBM4 high-bandwidth memory chips, and expects to expand shipments starting in September.

Software stocks face collective pressure as the logic of fund flows splits

IBM’s earnings warning was not an isolated case. On the same trading day, software stocks fell across the board—ServiceNow dropped 5.8%, Workday fell 3.5%, SAP slid 3.2%, and Salesforce fell 2.1%. Microsoft fell 1.55%, and Apple fell 0.77%.

The underlying logic behind this divergence is clear. Enterprise IT budgets are not increasing; they are being reallocated under limited funds. When a large share of money is forced into AI hardware infrastructure, the share of purchases devoted to software and traditional enterprise services must be squeezed out. IBM’s plunge was the most direct manifestation of this “squeezing effect.”

A Goldman Sachs Seoul team commented that the root cause of IBM’s earnings collapse is that enterprise customers are shifting massively to storage and memory purchases, confirming the arrival of a “storage super-cycle.” This shift in AI capital expenditures may further push a “software bear market.”

How the CPI cooldown beyond expectations provides a macro stage for a shift in market style

This market divergence did not occur in isolation. Before the market opened on July 14, the U.S. Department of Labor released June Consumer Price Index (CPI) data: up 3.5% year over year, a sharp drop from May’s 4.2%, and below the market expectation of 3.8%. The month-over-month figure recorded a decline of -0.4%, the first month-over-month decrease in six years.

Energy prices were the main drag item. The energy index fell 5.7% in June, including a 9.7% single-month plunge in gasoline prices. However, CPI year over year was still significantly above the Federal Reserve’s 2% inflation target, meaning inflation pressure has not fully disappeared.

This data substantially eased market concerns about persistently high inflation and directly changed expectations for the Fed’s rate hikes. According to the CME FedWatch tool, after the release, the probability the Fed would raise rates at the July meeting dropped from 42% the previous day to 17%. U.S. Treasury yields moved lower in response: the two-year Treasury yield fell 7.4 basis points to 4.19%, and the 10-year Treasury yield fell 2.5 basis points to 4.59%.

A rebound in macro sentiment pushed all three major U.S. stock indexes higher: the Dow rose 0.02% to 52,508.66; the Nasdaq rose 0.90% to 26,107.01; and the S&P 500 rose 0.38% to 7,543.89.

While cooling inflation gave high-valuation tech stocks some breathing room, divergence within the market reached an extreme level—funds did not flow evenly into all tech sectors; instead, they concentrated heavily along the upstream AI hardware industry chain.

The valuation reshuffle between old and new tech forces is only just beginning

IBM’s plunge and chip stocks’ rally, at their core, represent a reshaping of valuation frameworks for old and new tech forces.

In recent years, IBM has spent tens of billions of dollars acquiring software companies such as Red Hat and HashiCorp, attempting to transition from a traditional hardware manufacturer into a high-growth enterprise software and hybrid cloud services provider. However, in the AI era, the logic of capital expenditures did not play out as IBM had expected—enterprise customers did not prioritize purchasing software platforms to help integrate AI models, and instead poured directly into hardware such as servers, storage, and memory.

The market is now re-examining IBM’s AI transformation logic. Earlier this year, in February, after AI startup Anthropic launched AI tools that could help modernize IBM mainframe traditional programming languages, it temporarily sparked market concerns about IBM’s software business moat. This earnings warning further intensified those doubts.

By comparison, chip stocks are benefiting from the certainty of demand driven by AI infrastructure investment. From NVIDIA’s GPUs to SK Hynix’s HBM memory, and from Micron’s NAND and DRAM, the entire hardware supply chain is capturing the dividend from enterprise budgets being reallocated.

IBM said it is currently seeing delays in the timing of large project contract signings—not that customer demand has disappeared. Over the next few quarters, the market will closely watch whether these delayed orders can return gradually, or whether the structure of enterprise IT spending has already undergone irreversible, deep-seated changes. IBM is expected to release its full second-quarter earnings report on July 22.

Summary

The U.S. stock trading day on July 14, 2026 showcased, in the most extreme way, a structural shift in enterprise capital expenditures in the AI era. IBM plunged 25.21%, posting the biggest single-day drop since the 1987 “Black Monday,” with its market value erasing nearly $70 billion; meanwhile, chip stocks such as NVIDIA, Micron, and SK Hynix rallied across the board, with SK Hynix ADR jumping more than 27% in a single day.

The immediate trigger for this divergence was IBM’s weak preliminary earnings disclosure released ahead of schedule, but the deeper driver is that enterprise customers, amid the AI wave, are massively shifting their IT budgets from software and services toward hardware infrastructure such as servers, storage, and memory. The June CPI cooldown beyond expectations further reduced the likelihood of Fed rate hikes, providing a macro tailwind for tech stocks, but money did not flow evenly into every segment—AI hardware became the only broadly shared direction.

This may not be an isolated event, but the beginning of a new era. When every item of corporate capital expenditure is being reordered, and when the narrative that “hardware is eating everyone’s lunch” becomes the new market storyline, the valuation logic of traditional software and services providers, capacity planning for chip supply chains, and even the overall landscape of the tech industry may all face continued reshaping.

FAQ

Q: What historical level does IBM’s 25% plunge represent?

A: This is IBM’s biggest single-day decline since it listed in 1915, exceeding the 23.7% drop recorded during the October 19, 1987 “Black Monday” crash. IBM’s stock closed at $217.07, returning to a two-month low since May 13.

Q: What is the core reason behind IBM’s plunge?

A: The direct cause is that the company’s early-disclosed second-quarter preliminary earnings report came in far below market expectations—revenue of about $17.2 billion, roughly $17.9 billion below expectations. The deeper reason is that in the last few weeks of June, enterprise customers suddenly shifted capital expenditures at scale from software and mainframes to AI hardware such as servers, storage, and memory, causing multiple large deals to be recognized later.

Q: Why did chip stocks surge while IBM plunged?

A: IBM’s earnings warning precisely revealed the change in where enterprise IT budgets are going—customers are moving money from traditional software and services into AI hardware infrastructure. This means chip stocks (especially those related to AI compute and storage) directly benefit from this budget shift, while traditional software and services providers get squeezed out.

Q: What role did the CPI data play in this event?

A: In June, CPI rose 3.5% year over year, below the expected 3.8%, and month over month was -0.4%—the first decline in six years. After the data was released, the probability of a Fed rate hike in July fell from 42% to 17%, and U.S. bond yields retreated, providing a macro tailwind for tech stocks. But with the macro tailwind combining with the structural budget shift, it ultimately resulted in an extreme and highly concentrated sell-off and rally pattern, with funds concentrating in the AI hardware sector.

Q: What does IBM’s future look like?

A: IBM said it is currently seeing delays in the project signing timeline rather than the disappearance of demand. But the market will closely watch whether these delayed orders can come back in subsequent quarters, or whether the structure of enterprise IT spending has already undergone irreversible, deep-seated changes. IBM is expected to release its full second-quarter earnings report on July 22.