CXMT Jumped 466% on Debut to $487 Billion. Why Does Morningstar Call It 68% Overpriced?

Chinese memory chipmaker CXMT (ChangXin Memory Technologies) listed on Shanghai's STAR Market on July 27 and surged 466% on debut. Its closing market cap hit roughly 3.3 trillion yuan, about $487 billion, the highest of any mainland China-listed company. The IPO priced at 8.66 yuan (about $1.27) a share and raised at least $8.6 billion.

Where does CXMT actually come from

Per Meru, the story starts with the 2018 US-China trade war. Xi Jinping began personally pushing chip self-sufficiency, calling semiconductors "the heart of manufacturing," and Hefei, the capital of Anhui province, took the lead building CXMT into China's memory champion.

Meru argues Hefei was running a playbook it had already proven. In 2002, when Hyundai Electronics (now SK Hynix) sold off its struggling LCD unit Hydis, buyer BOE got roughly $2.2 billion from Hefei, nearly half the city's annual budget then, and grew into today's largest LCD maker. In 2016, Hefei ran the same play, funding most of CXMT's $2.7 billion first-phase project.

Is CXMT really a private company

Officially, CXMT has no single controlling shareholder, a framing independent sources confirm too. But Meru notes its top three shareholders are all Hefei-linked entities or the state semiconductor fund, and even after this IPO's dilution, their combined stake stays around 38%.

Whether to define CXMT by its official status or its actual ownership math depends on that gap.

Where did the technology come from

CXMT's DRAM technology traces back to Qimonda, the German memory maker that went bankrupt in 2009. Spun out of Siemens and later an Infineon subsidiary, Qimonda was once Europe's largest DRAM maker with a large patent portfolio. Canadian patent firm Polaris (a WiLAN subsidiary) bought roughly 7,000 Qimonda patents for about 30 million euros in 2015, and CXMT signed a patent licensing deal with Polaris in December 2019.

That piece confirms the December 2019 Polaris-CXMT deal and the roughly 30-million-euro 2015 patent purchase, but doesn't independently confirm the specific engineer hires Meru describes.

Analysts are split on the same closing price

Nomura's Donnie Teng set a price target of 116 yuan, 12 times the IPO price and nearly triple the first-day close. That rests on AI-driven hyperscaler capex sustaining memory demand and CXMT's share growing from about 8% to 18% by 2028. Morningstar's Wei Jingjie takes the opposite view, pricing fair value at 14.90 yuan, 72% above the IPO price but 68% below the first-day close. His case: without EUV lithography, CXMT's multi-patterning DUV process leaves its cost per bit over 30% above Samsung and SK Hynix.

What's the US wildcard here

Per CNBC, export controls on chipmaking tools remain CXMT's core challenge, and one analysis says it needs at least 15% global share to stay competitive long-term, above its projected 11% by 2028. Some US lawmakers separately want to block American purchases of CXMT chips.

So what actually settles this

It comes down to one question: can CXMT close its cost-per-bit gap without EUV access, or does that gap persist and undercut today's valuation. If export controls hold and CXMT's share still climbs toward 15%, Nomura's case holds up. If the cost gap doesn't close and the price cycle cools, Morningstar's case holds up. Either way, day one's 466% is the start of the question, not the answer to it.

Why it matters · CXMT is officially described as having no controlling shareholder, yet state-linked holders combine for roughly 38%, and Nomura's tripling target sits against Morningstar's fair value 68% below the first-day close. That gap shapes how to read China's chip buildout inside the AI memory supercycle.

Worth asking · State-linked shareholders combine for roughly 38%, yet CXMT is officially described as having no controlling shareholder. Should this be read as a state enterprise or a private one? And do you put more weight on Nomura's tripling target or Morningstar's 68%-below-open fair value?