Advanced Micro Devices (AMD) stock plummeted over 17% last night (the 4th), marking the worst single-day decline since May 2017, with a closing price of $200.19. Ironically, this crash occurred just after the company reported a Q4 earnings report that exceeded Wall Street expectations.
What caused a “better-than-expected” earnings report to trigger a panic sell-off?
Earnings Performance: Impressive Numbers, But “Not Enough”
AMD announced its Q4 2025 financial results, with revenue reaching $10.27 billion, up 34% year-over-year, and earnings per share (EPS) of $1.53, surpassing the consensus estimates of $9.67 billion revenue and $1.32 EPS. For the full year, revenue hit a record high of $34.6 billion, with a gross margin maintained around 50%.
The data center segment was the biggest highlight, with Q4 revenue of $5.4 billion, up 39% year-over-year, and full-year revenue of $16.6 billion, up 32%. While these are all positive signs, the issue is that investors were expecting more than just “good”—they wanted “astonishing.”
Amid the ongoing AI capital expenditure boom, with giants like Microsoft, Meta, and Google announcing hundreds of billions of dollars in AI infrastructure investments, some analysts had bet that AMD’s outlook would be even stronger. When the actual figures only “beat consensus” rather than “significantly surpass,” the disappointment from unmet high expectations quickly turned into selling pressure.
Q1 2026 Guidance: Moaning Signs of Sequential Decline
The real trigger for the sell-off was AMD’s guidance for Q1 2026. The company projected revenue of about $9.8 billion (plus or minus $300 million), higher than the previous market consensus of $9.38 billion, representing approximately 32% year-over-year growth. However, compared to Q4’s $10.27 billion, this indicates a clear sequential decline.
Additionally, AMD forecasted non-GAAP gross margin for Q1 at around 55%, down from 57% in Q4. The margin decline further deepened market concerns.
Compared to Nvidia’s long-standing pattern of “raising guidance beyond expectations,” AMD’s conservative outlook raised fears that growth in AI chip demand might be slowing or that AMD’s competitiveness in the AI accelerator market is weaker than anticipated.
China Factor: Impact of MI308 Export Controls
Another detail in the Q4 report that’s easy to overlook but has a significant impact: AMD’s shipments of the Instinct MI308 AI chips to China contributed about $390 million in revenue. However, due to tightening US export controls, the company expects China-related revenue to plunge to around $100 million in Q1.
In other words, the revenue gap from China alone is about $290 million, accounting for most of the sequential revenue decline from Q4 to Q1. But if we exclude the China factor, AMD’s core business trajectory might actually be more stable than the surface numbers suggest.
It’s worth noting that AMD recognized approximately $440 million in inventory and related costs in 2025 due to US export restrictions on MI308. Although the release of a $360 million inventory reserve in Q4 provided some relief, the uncertainty surrounding export controls remains a sword hanging over AMD.
The Unrepeatable Wave of Major Orders in Q4 2025
In Q4 2025, AMD experienced a wave of large orders and collaborations. Orders from major clients like OpenAI and Oracle, along with shipments of MI308 to China, collectively boosted that quarter’s performance. These one-time or seasonal factors set unrealistic high expectations for future outlooks.
When the guidance for Q1 indicated these tailwinds could not be replicated in the short term, the market reacted violently. This is a classic case of “high base effect”: the better the previous quarter, the harder it is to maintain similar growth next quarter.
Macro Environment: AI Bubble Concerns and Tech Stock Rotation
AMD’s plunge was not an isolated event, but the decline exceeded expectations.
On February 4th, the entire semiconductor sector came under pressure. Broadcom fell about 3.8%, Micron plunged 9.5%, Nvidia dropped 3.4%. The Philadelphia Semiconductor Index fell 347.18 points (-4.36%), closing at 7,619.16.
Software stocks also suffered heavy sell-offs, with Oracle down 5.1% and CrowdStrike down about 1.51%. There was a pervasive panic that AI tools would disrupt traditional software business models, prompting capital to rapidly flow out of tech stocks into other sectors.
CEO Su Zifeng’s Response
In response to the sharp decline, AMD CEO Su Zifeng actively defended the company in an interview on February 4th. She emphasized:
- The pace of AI acceleration “exceeds her previous imagination”
- Customer demand continues to rise, surpassing current computing resources and deployment speed
- Data center business is “significantly accelerating” from Q4 to Q1
- Demand for CPUs is showing “explosive growth”
Su Zifeng aimed to shift market focus from short-term guidance volatility to the structural long-term growth of AI demand.
What Do Analysts Say?
Despite the sharp stock decline, Wall Street analysts remain generally positive:
| Institution |
Rating |
Target Price |
| UBS |
Buy |
$310 (down from $330) |
| BofA |
— |
$280 (upgraded) |
| Wells Fargo |
Overweight |
$345 |
| KeyBanc |
Overweight |
$300 (up from $270) |
| Mizuho |
Outperform |
$275 (down from $285) |
| Benchmark |
Buy |
$325 |
The average target price among analysts is about $289, implying over 40% upside from the current price around $200. Most believe the market overreacted to short-term guidance, and AMD’s long-term growth story in AI and data centers remains intact.
The Cruel Lesson of Expectation Management
The core contradiction behind AMD’s stock plunge is: the company’s performance isn’t bad, but market expectations were too high. In the current frenzy of AI investment, “meeting expectations” is no longer enough; investors want “beating expectations” and then some.
In the short term, factors like China export restrictions, the high base effect from Q4, and overall risk-off sentiment in tech stocks created a perfect storm. But from a medium- to long-term perspective, AMD’s data center business continues to grow rapidly, and the structural demand for AI remains unchanged. Su Zifeng’s description of “explosive demand” is supported by multiple data points.
For investors, this may serve as a classic reminder: in a narrative- and expectation-driven market, good performance doesn’t necessarily translate into good stock performance. What truly influences short-term price movements is often not the absolute numbers themselves, but the gap between these numbers and market expectations.
