
- Samsung To Earn $58 Billion in 90 Days -Why the Market Still Sold Off
- The 8.3x Profit Ratio: Why Samsung’s Margin Boom Is a Warning
- The Profit Paradox: Why Peak Margins Can Signal Risk For Memory Stocks
- Why Stocks Like Micron, SNDK and SK Hynix Are Under Pressure
- What Wall Street Says About MU, SNDK and Memory Stocks
- Supply Risk: How New Capacity Could Hit DRAM and NAND Prices
- Key Risks for Memory Stocks: AI Capex, China and Antitrust Pressure
- Bottom Line: What Investors Should Watch in Memory Stocks Next
Samsung Electronics just reported the highest single-quarter operating profit forecast any technology company has ever posted: At $58.4 billion, it beats giants like Nvidia and Apple. It was 19 times higher than Samsung's operating profit in the same quarter a year ago.
However, within hours of the announcement, Samsung's stock in Seoul fell up to 10%. In the United States, Micron dropped approximately 4.5% in overnight trading, SanDisk fell further, and the Roundhill Memory ETF (DRAM) extended what had already been a sharp week-long pullback for the sector.
Let's break down the single most important number in Samsung’s report, the four structural forces pressing on memory stocks simultaneously, and investors should track the revenue line more carefully than the profit headline.
Samsung To Earn $58 Billion in 90 Days -Why the Market Still Sold Off
Samsung’s Q2 2026 guidance shows a sharp earnings reset, led by strong AI-related memory demand. Revenue more than doubled year-on-year to KRW 171 trillion, or around $111 billion, while operating profit jumped to KRW 89.4 trillion, or about $58 billion. Reuters reported that revenue rose 129% YoY, while operating profit increased nearly 19x.
| Metric | Q2 2026 Guidance | USD | Key Takeaway |
| Revenue | KRW 171 trillion | ~$111 billion | More than doubled YoY |
| Operating Profit | KRW 89.4 trillion | ~$58 billion | Nearly 19x YoY |
| Operating Margin | 52.3% | — | Massive margin expansion |
The key point is that Samsung is not just selling more chips; it is earning much more on each unit. Memory semiconductors in mid-2026 are at a structurally similar moment: DRAM prices have risen approximately 700% over four years, reflecting a supply-demand reset driven by AI data centres, HBM demand, and constrained memory capacity.
That is why Samsung’s profit growth has far outpaced revenue growth. The full earnings release will be important to confirm how much of this profit surge came from the semiconductor business.
The 8.3x Profit Ratio: Why Samsung’s Margin Boom Is a Warning
Here is a number that is actually worth paying attention to. Samsung's profit grew 8.3x faster than its revenue year-on-year.
| Metric | Q2 2025 | Q2 2026 | YoY Change |
| Revenue | KRW 74.57 trillion | KRW 171 trillion | 2.29x |
| Operating Profit | KRW 4.7 trillion | KRW 89.4 trillion | 19.02x |
| Operating Margin | 6.3% | 52.3% | +46.0 percentage points |
In a hardware company that designs better products, sells into new markets, and gains share, faster profit growth than revenue growth reflects operating leverage: costs stay fixed as revenue scales. That is genuinely healthy.
In a commodity memory company, however, this ratio means something different. It means nearly all of the profit expansion came from pricing alone. Volume (as measured by revenue) grew 2.3x. Everything above that, meaning the additional 8.3x of profit momentum, came from charging more per unit.
| Samsung Financial Snapshot | Q2 2025 | Q2 2026 | Change |
| DRAM ASP Change (Q2 QoQ) | — | +44% | Citi Research estimate |
| NAND ASP Change (Q2 QoQ) | — | +53% | Citi Research estimate |
Sources: Samsung Q2 2026 earnings guidance; Samsung Q1 2026 annual report; Citi Research
For context, during the 2017-18 semiconductor supercycle, widely considered memory's previous "super" moment, Samsung's quarterly operating profit grew roughly 4 to 5 times faster than revenue at the cycle's peak. The current 8.3x ratio is approximately double that historical high-water mark.
Samsung's guidance implies an operating margin of approximately 52.3%, which confirms exceptional pricing power but also sets a much higher bar for July 30. Samsung's 2018 full-year operating margin was approximately 24.2%, based on KRW 58.89 trillion in operating profit and KRW 243.77 trillion in revenue.
The Profit Paradox: Why Peak Margins Can Signal Risk For Memory Stocks
This is the framework that matters: The Profit Paradox. In commodity cycles, peak margins are not a signal of safety. They are the cycle's most reliable warning. When a commodity reaches maximum pricing power, three structural responses become inevitable, usually within 12 to 18 months.
- Buyers reduce dependency: AI labs, PC makers, and smartphone companies start designing systems that use less memory per workload. NVIDIA’s inference software stack and Apple’s newer SoC architectures both point toward better memory efficiency. This demand destruction usually appears only after orders begin slowing.
- Suppliers expand capacity: Samsung is accelerating memory expansion, while industry capex is rising sharply. Goldman Sachs expects the four largest hyperscalers to spend $5.3 trillion on AI-related capex between FY2025 and FY2030, and Samsung has announced KRW 2,100 trillion of South Korea investment through 2040. If Samsung, Micron, and SK Hynix all add supply together, the 2027–2028 capacity wave could hit a market where demand is already moderating.
- Contract cycle turns against sellers: The current generation of long-term agreements (LTAs) being signed by SNDK, Micron, and SK Hynix includes floor pricing that locks in healthy margins. But the next round of negotiations for 2028–2029 will be negotiated by hyperscalers like Google, Microsoft that have just absorbed a 700% price shock. Their priority will be pricing protection, not supplier margins.
The selloff in memory stocks like Samsung, Micron, SK Hynix or an ETF like DRAM was not a rejection of the quarter. It marked the shift from recovery to repeatability. Rising memory prices are no longer enough. Samsung now has to defend a 52% margin in a cycle already showing signs of slower momentum.
Why Stocks Like Micron, SNDK and SK Hynix Are Under Pressure
The Samsung earnings selloff did not happen in isolation. It was the final push on a door that four other forces had already pushed open.
- SK Hynix’s Nasdaq listing is creating portfolio rebalancing pressure: With book building starting July 6, final pricing due July 9, and Nasdaq trading expected July 10, some institutions may be trimming existing memory positions to make room for the new ADR exposure.
- The DRAM antitrust lawsuit has added legal overhang: Samsung, SK Hynix, and Micron were sued on June 25 in California, with plaintiffs alleging coordinated DRAM supply restrictions and price inflation of nearly 700% over four years. The allegations are unproven, and the companies say they compete independently, but the case still weighs on sentiment given the sector’s past price-fixing history.
- Demand destruction risk is rising: Citrini Research warned that PC OEMs, hyperscalers, and Nvidia server partners may start reducing memory usage if prices stay elevated. That is the demand-side risk: buyers may begin engineering around the very shortage that created this profit boom.
- Q3 price momentum is cooling: Conventional DRAM contract prices are expected to rise 13%–18%, while NAND Flash prices are expected to rise 10%–15%. These are still strong gains, but far below Q2’s 44% DRAM and 53% NAND surge. That means Samsung’s Q2-style profit expansion is unlikely to repeat at the same pace.
What Wall Street Says About MU, SNDK and Memory Stocks
The Wall Street consensus on memory stocks remains broadly constructive. But the spread between the most bullish and most conservative targets is wider than almost any other sector, which itself reflects the core debate: structural cycle break or temporary pricing peak?
| Company | Analyst / Firm | Rating | Price Target | Core Thesis |
| SNDK | Mark Newman / Bernstein | Outperform | $3,000 | New LTAs have floor pricing at ~$0.29/GB; downside protection is structural |
| Blayne Curtis / Jefferies | Buy | $3,000 | Improved earnings stability from LTA structure through 2027-28 | |
| Wamsi Mohan / Bank of America | Buy | $2,500 | NAND supply imbalance persists through 2027; pricing power "strong for longer" | |
| Citigroup | Buy | $2,500 | NAND ASPs to rise ~186% YoY in 2026; supply tightness beyond 2027 | |
| Susquehanna | Buy | $3,250 | Highest target on the Street; strongest bull case on LTA durability | |
| Morgan Stanley | Equal Weight | $1,750 | Cautious; limited upside at current prices; monitors capex trajectory | |
MU
| Goldman Sachs | Neutral | $1,100 | AI capex moderation risk; near-term price exhaustion |
| KeyBanc / John Vinh | Overweight | — | June DRAM contract prices rose ~3% MoM; supply still tight through 2026 |
Sources: Bank of America; Bernstein, Citigroup, Jefferies, Susquehanna, Morgan Stanley, Goldman Sachs, KeyBanc
The divergence between Bernstein's $3,000 and Goldman's cautious $1,100 on Micron is the fundamental debate: are the new long-term supply agreements a structural break from memory's boom-bust pattern, or are they the kind of optimism-era contracts that always look solid until the cycle turns?
Micron disclosed 16 Strategic Customer Agreements with approximately $100 billion in minimum contracted revenue and $22 billion in upfront customer cash, covering 20% of its DRAM volume and one-third of its NAND volume through 2030. Micron guided Q4 fiscal 2026 revenue to approximately $50 billion plus or minus $1 billion, against analyst consensus of $43.6 billion, coming in roughly 15% above what Wall Street expected.
These numbers are exceptional. The question Micron's bears are asking is: what happens to that $50 billion revenue trajectory when AI capex, currently approximately 52% of cloud provider spending, starts to plateau?
Supply Risk: How New Capacity Could Hit DRAM and NAND Prices
The memory shortage is real. Samsung is pushing PCIe Gen6 eSSD and plans to deliver its first HBM4E samples in H2 2026. KB Securities-Jefferies’ Jeff Kim expects the shortage to deepen in 2026 and 2027, with limited capacity growth and strong demand.
But supply is coming, and semiconductor cycles turn when future supply starts changing pricing power.
- Micron expects about $10 billion of capex in fiscal Q4, taking FY2026 capex to roughly $27 billion, with FY2027 quarterly capex expected to move even higher.
- Samsung has announced KRW 2,100 trillion of South Korea investment through 2040. South Korea also plans to double DRAM capacity within five years.
- SK Hynix’s Nasdaq listing could support further manufacturing expansion.
That matters because all three major memory producers are adding capacity while prices are still around 700% above 2022 levels. Even if new supply arrives only in 2027 or 2028, hyperscalers signing three-year contracts today know the seller’s leverage is nearing its peak.
History supports the caution. The 1993–95 PC DRAM boom ended after yield gains and new capacity drove prices down more than 60%. The 2017–18 mobile memory boom also started repricing once Samsung and SK Hynix signaled capacity additions, before new supply actually arrived.
This does not mean the cycle ends now. SemiAnalysis still expects memory profits may not peak before Q4 2027, with key peak-cycle signals still missing as of early 2026. The bull case remains real: no major new fab arrives before mid-2027, AI demand is structural, and long-term contracts now provide pricing floors that past cycles lacked.
Key Risks for Memory Stocks: AI Capex, China and Antitrust Pressure
Five risks could weaken the memory supercycle, increasing concerns around stocks like Samsung, SK Hynix, Micron, Sandisk, and others.
1. AI capex moderation: Memory demand depends on hyperscalers continuing aggressive AI infrastructure spending. J.P. Morgan estimates AI capex already makes up about 52% of cloud service provider spending and could exceed 70% by 2026. Any slowdown would reset demand faster than new supply can adjust.
2. Chinese competition: Chinese companies like CXMT and YMTC remain behind Samsung, SK Hynix, and Micron in leading-edge memory, but they do not need cutting-edge nodes to compete in conventional DDR5 and LPDDR5. Low-cost Chinese supply could pressure the commodity memory segment where prices are up nearly 700%.
3. Demand destruction: Apple reportedly raised Mac and iPhone prices due to higher memory costs, while CyberPowerPC cited a 500% rise in memory costs. These are early signs that end buyers may reduce volumes or shift to lower-memory configurations.
4. Legal overhang: Samsung, SK Hynix, and Micron face unproven class-action allegations around DRAM supply restriction and price inflation. Even if the case fails, any DOJ inquiry or settlement process could limit the pricing behavior currently supporting peak margins.
5. Bonus-related cash drag: Samsung’s wage agreement allocates 10.5% of semiconductor operating profit to special employee bonuses. Analysts estimate cumulative bonuses could exceed KRW 40 trillion, or about $26.1 billion, creating a recurring cash flow drag before shareholder returns.
Bottom Line: What Investors Should Watch in Memory Stocks Next
Samsung's Q2 2026 report was not a mixed result. It was a historically unprecedented profit print accompanied by a revenue miss and an implied operating margin of 52.3%, a level that has no precedent in memory semiconductor history. The selloff is not irrational. It is the market updating its probability estimate for where in the cycle we are.
Bull case: This cycle has unusual demand visibility. Micron has $100 billion of contracted revenue through 2030, SNDK has floor pricing near $0.29/GB, and Samsung plans HBM4E delivery in 2026. If AI capex stays strong, Goldman Sachs projects $5.3 trillion of hyperscaler spending through 2030, the memory shortage could last until 2027–2028.
Bear case: Samsung’s 8.3x profit-to-revenue expansion ratio is the warning. Pricing, not volume, is driving profits. Q3 DRAM price growth is expected at 13%–18%, well below Q2 levels. With new capacity due in 2027–2028, plus the antitrust lawsuit and SK Hynix Nasdaq listing, near-term risk-reward has weakened.
The One Number to Track: Watch Samsung's Q3 2026 revenue growth rate, due July 30. If revenue growth continues to decelerate while profit holds, the Profit Paradox will be confirmed. If revenue accelerates, meaning volume is picking up alongside price, the bull case gets a second wind.