Goldman Sachs Raises CXMT Target Price to 130 Yuan; Q3 Revenue Guidance Beats Expectations by 30%
nashnova research
Goldman Sachs maintained its Buy on CXMT (长鑫科技) and raised the 12-month target to ¥130; Q3 revenue guidance midpoint came in roughly 30% above Goldman's prior estimate, implying 225% upside from the current share price of about ¥40.
How much did Q3 guidance actually beat by?
Goldman estimates CXMT's Q3 revenue at ¥115–120 billion, up 15–20% quarter-on-quarter. The midpoint tops Goldman's prior forecast by roughly 30%.
Net-profit guidance is ¥53–55 billion, a modest QoQ gain and about 17% above Goldman's old estimate.
This means → Revenue beat by a far wider margin than profit, signaling that CXMT is in a rapid volume ramp — margins have not yet caught up with the top line.
Where is the upside coming from?
Goldman attributes the beat to three drivers: AI adoption keeps pulling server DRAM — the mainstream memory chip in computers and servers — and HBM (high-bandwidth memory, a high-speed memory built specifically for AI processors) demand higher.
Downstream customers are actively diversifying supply chains, and CXMT, as China's core DRAM supplier, captures that shift directly.
DRAM prices remain elevated, so both volume and pricing are lifting revenue together.
What does the five-year growth path look like?
Goldman raised FY2026 revenue to ¥387.5 billion (up ~9%) and net profit to ¥191.8 billion (up ~4%).
Projected 2026 gross margin: ~81.4%; net margin: ~49.5%.
By 2030, Goldman forecasts revenue at ~¥1.39 trillion and net profit at ~¥747.9 billion. In plain terms = Goldman expects CXMT to grow revenue roughly 3.6× and net profit roughly 3.9× within five years.
What does G5 mass production signal?
CXMT announced in September 2026 that its fifth-generation DRAM platform (G5) entered mass production, shrinking the active-area half-pitch — the narrowest repeating unit width in the chip's core — to 11.95 nm.
The core functional area height dropped to 6,762 nm, delivering higher efficiency, lower cost, and greater output.
This means → G5 mass production proves CXMT can keep pushing into advanced process territory — the technology backbone behind Goldman's Buy rating.
How did Goldman arrive at the ¥130 target?
Goldman starts from a 2030 EPS forecast of ¥11.2, applies a 16.6× target P/E, then discounts back to 2027 at a 12.7% cost of equity to reach ¥130.
The 16.6× P/E maps to ~21% average YoY net-profit growth in FY2030–2031; the target implies a 2027 P/E of roughly 24×.
Cost of equity of 12.7% = beta of 1.5 × equity-risk premium of 6.5% + risk-free rate of 3.0%. In plain terms = Goldman views CXMT's risk as 50% higher than the broad market, so demands a correspondingly higher required return.
With such a wide gap between price and target, what is the key test?
CXMT trades at ~¥40; Goldman's target is ¥130 — implied upside of 225%. The stock has sold off sharply, with market cap falling below ¥3 trillion.
This means → A huge gap separates the market's pricing from Goldman's expectation. The core test is whether Q3 and subsequent quarters can keep delivering.
Goldman expects Q4 revenue to continue growing QoQ, but whether this valuation gap narrows hinges on the interplay between production ramp progress and the DRAM pricing cycle.
市场有风险，内容仅供研究参考，不构成投资建议。