Kioxia wants a $10B US listing as the memory trade cools

Share
Kioxia wants a $10B US listing as the memory trade cools

Tokyo's NAND maker is shopping for US capital at an awkward moment — the stock is down more than half from its peak, and the market's memory trade has turned.

Kioxia is weighing a US share sale of $10 billion or more, targeting an American Depositary Receipt listing as early as spring 2027. Bloomberg reported the plan on September 14, citing people familiar with the matter. The company has not commented, so treat the size and the date as reported, not confirmed.

The stated rationale is liquidity and access to US capital, not expansion. Kioxia has spent heavily buying back its own stock in Japan, and Bloomberg's sources describe the US listing as a way to widen the buyer base. The company is also considering a stock split at home, after its shares rose roughly 900% since the start of 2026.

That 900% is the part that needs a second sentence. Kioxia has fallen more than 54% from its peak in just over two months, giving back part of an eightfold run through the first half that had helped carry the Nikkei to records. The reversal tracks the same worry that has hammered the semiconductor complex all week: hyperscalers may not keep raising AI infrastructure spending at the current slope, and industry-wide capacity expansion could push memory prices down just as the new supply lands. Add last week's calls from Anthropic, OpenAI and xAI for a more measured development pace, and the memory trade is now fighting the market's slowdown narrative instead of riding it.

The mechanics of the listing matter more than the headline number. Kioxia's depositary shares already trade in the US, but unsponsored, on OTC Markets' Pink Limited tier — thin disclosure, higher risk for buyers. A sponsored Nasdaq listing would put the company formally behind the instrument and cut that risk substantially. SK hynix set the template by listing in New York in July, with the same pitch: own the memory layer of the AI buildout in dollars. Kioxia's shareholder register is shifting underneath it in parallel — Toshiba cut its stake to 12.84%, continuing a steady sell-down, while SK hynix holds convertible bonds that could become a 14.19% voting stake, the largest single block. SK chair Chey Tae-won floated closer manufacturing ties this month; Kioxia CEO Nobuo Ota shot that down immediately.

The honest read: this is a company trying to raise public money near the top of a cycle that the equity market has already started to discount. Kioxia's underlying case is not weak — the AI memory squeeze is real and we covered how it spread when AI demand sold out all of 2027's DRAM and HBM capacity. But a sponsored ADR does not fix the cycle. It just puts the cycle in front of a larger audience, priced in the currency those buyers actually hold.

What to watch: whether Kioxia files, and where a deal prices relative to that 54% drawdown. A listing that clears at a discount confirms the market thinks memory is peaking; a deal pulled for "market conditions" says the same thing louder.

Is the NAND side of the AI trade a buy at 54% off the peak, or is the memory cycle already past its top? Tell us in the comments.

Read more

Kling 4.0 goes live with 30-second native video at 10-bit HDR

Kling 4.0 goes live with 30-second native video at 10-bit HDR

Kuaishou's Kling AI has spent two years becoming the video model people actually finish projects with. Its new flagship, Kling 4.0, went fully live on October 15 — and this round is less about prettier frames than about the two things that make a generated shot usable in a real edit: length and control. Kling 4.0 generates up to 30 seconds in a single native pass, outputs 10-bit HDR at 4K, 1080p and 21:9, and takes up to 15 reference assets in one task. That budget splits across ten images, fi

Brain organoids classified Braille at 83% — in a trio, not alone

Brain organoids classified Braille at 83% — in a trio, not alone

Two stories today: human brain tissue on a chip decoded a tactile signal, and the data-centre company behind some of the largest AI contracts in Europe turns out to be six months behind on its flagship site. Human forebrain organoids grown on a microelectrode array classified Braille letters with 61 percent accuracy on their own and 83 percent when the responses of three organoids were combined, in a peer-reviewed study from the University of Bristol. The team took event-based tactile readings

Suleyman says Microsoft's base model beat OpenAI's on a quarter compute

Suleyman says Microsoft's base model beat OpenAI's on a quarter compute

Microsoft finished pre-training its biggest in-house model yet — then used it to make a point about the company it bankrolls. Microsoft AI completed pre-training of MAI-Base-2, a mixture-of-experts foundation model with 80 billion active and 2.8 trillion total parameters, and says the recipe is 4.2–12.3x more pre-training FLOPs-efficient than DeepSeek-V4-Pro-Base. The run itself is the number worth keeping: 16,000 Nvidia GB200 GPUs for 38 days across August and September, at 21% average model

Amazon's new Bee devices listen all day — for $11.99 a month

Amazon's new Bee devices listen all day — for $11.99 a month

Amazon turned its always-listening Bee line into a two-device family on Thursday, and NVIDIA opened a 2.8-trillion-token Indic corpus recipe. The connecting thread: both are bets that the next AI money is in memory — yours, or your model's. Amazon put two new always-listening devices on sale, and both now lead to a subscription. Bee Frontier, the next-generation wristband that also clips to clothing, ships now at $74.99 in Onyx, Lavender and Sage — up from the $50 predecessor — with seven-day