Memory Is Still Short, but Less Short: Checking for the Peak This Week

Citi trimmed its 2027 DRAM and NAND shortage forecasts together. Server pricing remains strong, while consumer markets show weakness. This week's numbers show where the memory boom may peak.

Memory Is Still Short, but Less Short: Checking for the Peak This Week

The 20-second read

1
Citi cut its 2027 DRAM shortage forecast from 8.7% to 3.8%, and its NAND forecast from 6.1% to 3.3%.
2
Contract prices for server DRAM and enterprise SSDs are still strong, but buyers in the consumer spot market are waiting on the sidelines.
3
The peak in the rate of price increases may have passed, the peak in earnings has not yet been confirmed, and stock prices can move before earnings do.

The 20-second brief

Citi cut its 2027 DRAM shortage forecast from 8.7% to 3.8%. The shortage continues, but it is smaller than expected.

AMD's data center revenue surged. AI demand is alive, but consumer memory prices and the stock market are telling a different story.

This week's verdict: the pace of price increases has likely peaked, earnings have not been confirmed to have peaked, and there is a risk that stock prices peak first.

The biggest change this week is the size of the shortage

2027 DRAM shortage forecast: 8.7% → 3.8% (Citi revision)

2027 NAND shortage forecast: 6.1% → 3.3% (Citi revision)

The question investors ask about memory has changed. Through last week, the focus was "how long will it stay short?" This week, "how much less short will it get?" matters more.

In its 2027 outlook released Tuesday, Nov. 3, Citi cut the DRAM shortage from 8.7% to 3.8% and the NAND shortage from 6.1% to 3.3%. This is not an announcement that the shortage is over. It says supply is catching up with demand faster than expected.

Even if a store is still short of goods, it is hard to keep raising prices once the line of customers shrinks quickly. Memory works the same way. The change in the size of the shortage may be a more sensitive signal for future prices and earnings than whether a shortage exists.

That does not mean AI demand has cracked. AMD reported Tuesday, Nov. 3, that third-quarter data center revenue rose 97% from a year earlier. Demand for chips going into AI servers is strong. Yet the Nasdaq posted a weekly decline even after Friday's (Nov. 6) rebound. Good earnings and a good stock price are already separate matters.

The price list holds two different markets

Counterpoint said Thursday, Nov. 5, that the pace of third-quarter DRAM price increases had slowed. In the spot market tracked by TrendForce, buyers of some DDR products stood aside rather than chase prices. Consumer NAND wafers also showed weakness.

Contract prices for server DRAM and enterprise SSDs, by contrast, remain strong. TrendForce's fourth-quarter forecast calls for conventional DRAM contract prices to rise 10% to 15% and NAND 15% to 20%. Claims that prices have collapsed blend these two markets together.

Another shift shows up here. Splitting the market into HBM and everything else is no longer enough. Even ordinary DRAM can see strong demand once it goes into an AI server. Memory in smartphones and PCs, on the other hand, faces demand suppressed by high prices themselves. Who is buying now matters more than the product name.

This week's peak verdict: there is more than one peak

First, the peak in the rate of price increases. It has likely already passed. That does not mean prices are falling. They are still rising, only more slowly. It is like a car that slows from 100 km/h to 80 km/h: it still moves forward.

Second, the peak in corporate earnings. Not yet confirmed. Server demand and long-term supply agreements are supporting results. Micron said it has allocated or committed more than 75% of its fiscal 2027 volume to customers. But committed volume is not all revenue locked in at the same price.

Third, the peak in stock prices. It can come before earnings. The market prices not only what companies earn now but also what they will earn in the next cycle. Even while the boom lasts, lower expectations for future earnings can shake shares first.

The overall verdict this week is that the peak is one step closer than it was last week. That does not mean all three peaks have been confirmed.

Two illusions: capacity additions and low P/E ratios

The reasons for supply to grow have also increased. Citi raised its 2027 capex growth forecasts for DRAM and NAND to 55% and 35%, respectively. China's CXMT and YMTC are also expanding capacity. Price competition in commodity products in particular could intensify.

Still, money spent on factories today does not turn into memory supply tomorrow. Micron's expanded investment starts with construction and securing cleanroom space. Installing equipment and raising yields take time before real product ships. Higher investment is a seed of future risk, not evidence of oversupply next quarter.

There is also an illusion of cheapness. Barron's cited Micron's forward P/E at 3.9 on Thursday, Nov. 5. But when earnings are at a peak, P/E ratios can look very low. A stock at 4 times earnings is at 8 times if earnings halve while the price stays put.

So the better question is not "how cheap is it now?" but "how much can the company earn if prices normalize?" BofA stresses the strength of memory supply and demand, while Morgan Stanley warns about low multiples near a cycle peak. There is reason to watch both.

Next week, Nov. 9 to 13: three things to watch

First, whether spot-price weakness spreads to contract prices for server DRAM and enterprise SSDs. If both weaken together, this week's warning gets one notch heavier.

Second, whether customer behavior changes. Beyond simple waiting in the spot market, watch for rising inventories, delayed deliveries and changes to long-term contract terms.

Third, whether AMD's strong results lead to upward earnings revisions across the semiconductor sector. At the same time, watch capex and free cash flow at the big tech companies investing in AI data centers. Buying a lot of chips is different from making money on that investment.

Insight Times view

This week's news is not a death notice for the memory boom. But the easy formula that "as long as there is a shortage, stocks keep rising" has cracked. The shortage persists and earnings are strong. But if the shortage narrows and capacity additions speed up, stocks may price in ordinary future earnings first.

Long-term investors do not need a panicked sale of everything, nor an all-in purchase just because prices fell. It is reasonable to keep a long-term approach built on non-leveraged core ETFs, check for overconcentration in semiconductors and deploy new money in stages.

Next week the question is still one: not "is it still short?" but "can earnings last longer than the pace at which the shortage is shrinking?"

FAQ

Prices keep rising. Why worry about a peak?

Stock prices can reflect the future pace of increases and the durability of earnings before they reflect the price level.

Did Chinese suppliers' share surge this week?

No. By Counterpoint's measure, CXMT's DRAM revenue share slipped from 10% in the second quarter to 9% in the third. The concern is future capacity expansion.

Do heavy long-term contracts make oversupply impossible?

No. Contracted volume, price terms and the possibility of delivery adjustments are separate matters. Customer orders and actual shipments need to be checked together.

Has the memory chip peak already passed?
  • The peak in the rate of price increases has probably passed. That does not mean prices are falling.
  • The peak in corporate earnings has not yet been confirmed.
  • The peak in stock prices can show up before the peak in results. The three peaks need to be viewed separately.
How did Citi change its DRAM shortage forecast?
  • Citi cut its 2027 DRAM shortage forecast from 8.7% to 3.8%.
  • It cut its NAND forecast from 6.1% to 3.3%. This does not mean the shortage is going away. It means the shortage is getting smaller.
Can we say memory prices have collapsed?
  • The numbers so far make that hard to argue. Weakness has shown up in the spot market and in consumer NAND.
  • Contract prices for server DRAM and enterprise SSDs are still strong. TrendForce expects fourth-quarter contract prices for commodity DRAM to rise 10% to 15%.
Is Micron's forward P/E of 3.9 a cheap price?
  • It may look cheap, but it is hard to say for sure. When earnings are at a peak, the P/E can look low.
  • A stock with a P/E of 4 would see its P/E rise to 8 if earnings were cut in half, even with the share price unchanged.
What should we watch in the memory market next week?
  • Watch whether weakness in spot prices spreads to contract prices for server DRAM and enterprise SSDs.
  • Also watch for rising inventories, delayed deliveries, and changes to long-term contract terms.
  • We need to check whether AMD's strong results lead to higher earnings forecasts across the semiconductor sector.

What to watch

Weakness spreading to contract pricesWatch whether contract prices for server DRAM and enterprise SSDs weaken the way spot prices have during the week of 11/9 to 11/13.
Changes in customer behaviorCheck for rising inventories, delayed deliveries, and changes to long-term contract terms.
Citi's capex outlookWatch whether Citi's 2027 forecasts of 55% growth for DRAM and 35% for NAND are echoed elsewhere.
AI investment cash flowWatch whether Big Tech capital spending and free cash flow continue to support chip results like AMD's.
Memory is still in shortage, but the gap is narrowing, so stocks may price in a peak before earnings do.

Insight Times Editorial Desk

More by topic

Starting with this story's topics, then what readers are following now

Investor psychology, decision-making and books on investing

HBM

This story's topic77 storiesSee all →

Memory stacked next to the GPU.

The Real ASML Number Isn't the 3% Pop, It's the 30% Capacity Hike

ASML's rally isn't just an AI trade. Customers are already locking in 2027 tool slots, and ASML is raising Low-NA EUV capacity by roughly 30% to keep up.

Investing

A 3x P/E is not a discount. It is a verdict on how long memory profits last

Nomura's buy case on Samsung Electronics and SK Hynix does not stop at higher memory prices. It argues that if AI has turned memory from a cyclical component into strategic infrastructure, the formula the market uses to value both companies has to change too.

Investing

2027 Isn't the End: Citi's Case for a Memory Supercycle Running to 2031

Citi says the memory shortage timeline is stretching again. If AI shifts from "train once" models to continual learning, HBM, server DRAM and enterprise SSDs could all fall short at the same time.

Tech

AI ran out of memory before it ran out of GPUs

The HBM bottleneck is spreading into DRAM prices, server costs and the price of the next laptop or phone.

Tech

Has Memory Peaked? Prices Are Cooling, but the Shortage Looks Set to Run to 2028

Price growth has slowed and consumer demand is already breaking. Server and HBM supply remains short, and Micron says more than 75% of its 2027 volume is already committed to customers.

Tech

AI Chip Peak Check: Demand Still Rising, Risk Building in 2027-2029 Supply and Funding

DRAM and NAND contract prices are still climbing and HBM suppliers hold the edge. Power shortages, AI financing costs and simultaneous memory expansion are the risks for later years.

Investing

AI Chips Haven't Peaked. The Real Risk Sits in 2028 Supply

HBM and server DRAM shortages are now spreading into enterprise SSDs. The question isn't today's strength, it's how much new capacity these prices are pulling in and when it lands.

Investing

A building that houses servers and networking gear and supplies them with powe…

Nvidia and Palantir Put AI Inside the Supply Chain - The Real Shift Is That It's Learning Judgment

A single Vera Rubin rack holds 1.3 million parts. The real breakthrough in Nvidia and Palantir's new supply chain AI isn't automating that complexity, it's capturing the judgment calls that used to live only in planners' heads.

Tech

Is Space Cooler for AI? Why Musk Wants Data Centers in Orbit

The next limit on AI may be power and cooling, not GPUs. Elon Musk is looking to Earth orbit for an answer, but the test is cost per hour of compute.

Tech

Musk Sees the Next Phase of AI Infrastructure: Who Can Switch GPUs On, Fast

If neoclouds struggle under financial pressure, AI demand may not shrink. It may simply favor the few operators with the capital and execution to serve it, a shift that could matter for SpaceX and Tesla.

Tech

White House Swaps AI for "SI." The Real Investment Signal Is Infrastructure, Not the Name

An executive order renames artificial intelligence "Super Intelligence" and six frontier labs sign a voluntary safety accord. What investors should watch is the data center, power, nuclear and chip policy now gathered under one national-competitiveness frame.

Tech

The Cheaper Tokens Get, the Bigger Your AI Bill Gets

As AI shifts from training to inference, the token is becoming an industrial commodity. Falling prices are not cutting demand, because agents and automation consume tokens at a scale no human chat session ever did.

Investing

Marvell's $115 Billion Lumentum Deal: Is AI's Next Bottleneck Light, Not Chips?

Marvell's purchase of Lumentum points to a contest beyond GPU speed, in data transfer and power efficiency. Whether the price is right is not yet confirmed.

Tech

Workers in hard hats beside a communications equipment cabin, with a Starship rocket on its launch tower at dusk in the background

SpaceX Forecast at About $730 Billion in 2028 Revenue: The Bigger Opportunity Musk Sees

Sequoia Capital made a bold forecast, and Elon Musk replied, "I think we can do better." The future SpaceX is looking at is far bigger than the rocket launch business. The plan is to link satellite communications, AI data centers and orbital computing into one giant business.

TechEXPLAIN

NVDA

96 storiesSee all →

Who Hits $1 Trillion Next: Reading AMD and ASML by Different Rules

One company is chasing Nvidia's GPU empire. The other sells the machine that draws the chips for every advanced chipmaker, including Nvidia. Both get floated as 2028 trillion-dollar candidates, but the numbers investors need to watch are entirely different.

Investing

GPU

69 storiesSee all →

A chip built to draw graphics, whose ability to run thousands of identical cal…

The Next AI Chip Race Is About How Chips Connect

Applied Materials will lift its stake in Dutch equipment maker BESI from about 9% to 20%. The bet is less about control than about a new way to link chips.

Tech

Musk Borrowed $40 Billion for GPUs. He Also Warned AI Cloud Firms Could Go Bust

SpaceX's giant financing shows a new dividing line in AI: who borrows matters as much as how much is spent. Firms with other revenue may outlast those that rent out GPUs alone.

Tech

US AI Data Centers Are Headed for 90GW. Gigawatts Now Matter More Than Dollars

Goldman Sachs expects US data center capacity to grow from 64GW at the end of 2026 to 90GW at the end of 2027. The AI race is moving from buying GPUs to building power plants and transmission lines alongside them.

Tech

The AI Bubble Question Has Changed: By Late 2026, It's ROIC Over GPUs

There is no sign yet that AI investment is slowing. If anything, hyperscaler spending forecasts keep climbing. But as the scale of that capital has grown, the market's question has shifted from "is AI growing" to "when, and how much, does this capital come back as cash."

Tech

2027 Investing Needs Alignment, Not Prediction

A new investing book argues the hard part was never picking the next winner. It's staying aligned with where value and bottlenecks are moving while keeping old human weaknesses in check.

Investor Mind

Why Nebius Is Keeping Some 2027 AI Capacity Off the Shelf

Nebius mixes long-term contracts, short-term premiums and auctions. Holding back capacity pays only if the servers stay busy, and it is not yet clear the pricing will last.

Tech

Amazon Goes After Nvidia's Strongest Moat, and It Is Software, Not GPUs

Amazon's open-source TorchNeuron aims to let existing AI code run on its Trainium chips. The contest comes down to the total cost of AI compute and the return on Amazon's $220 billion in annual capital spending.

Tech