#高盛看多长鑫 Goldman Sachs initiates coverage of CXMT with a Buy rating and a RMB 129 target price: PEG is only 0.13—cheap or a geopolitical discount?

Goldman Sachs is formally initiating coverage of ChangXin Technology (ChangXin Storage / CXMT) for the first time, giving it a Buy rating and a 12-month target price of RMB 129. On the surface, a 24x 2027 P/E ratio is not exactly cheap, but Goldman calculates a PEG of only 0.13, putting the question back to the market: Is this genuinely cheap, or has a geopolitical discount been permanently priced into the valuation?

Goldman Sachs gives ChangXin Technology a Buy rating. This 24x figure is derived by discounting the expected 2030 P/E of 16.6x back to 2027 at a 12.7% cost of equity.
The report points out that CXMT accounts for approximately 50% of China’s DRAM capacity. Goldman lays out the valuation dilemma itself: Is this an undervalued low valuation, or a geopolitical discount? RMB 129 is the sell-side anchor, not a price that has already been realized.”

A low PEG does not necessarily mean cheap
A PEG of 0.13 is strikingly low. Back-solving from a 24x 2027 P/E implies extremely high profit growth in the model; the same is true when back-solving from the 16.6x 2030 P/E.
A low PEG can mean genuine cheapness, but it can also mean that high-growth assumptions have inflated the denominator, or that the market has continued to apply a geopolitical discount to China’s domestic DRAM industry. These three explanations cannot currently be separated.

Equipment and capital expenditure spillover
The year-on-year upward revisions to forecasts for 2026 through 2030 versus the previous estimate are 15%, 32%, 51%, 64%, and 79%, respectively. This is a forecast revision path, not the year-on-year growth of capital expenditure itself.
Wafer fab equipment spending is expected to grow 13%, 20%, and 15% year on year from 2026 to 2028, respectively, reaching approximately $53 billion in 2027. Deposition, etching, and lithography are the largest subsectors. Do not write these figures as orders that have already materialized.

Risk boundaries
The entire article is based on Goldman Sachs’ sell-side model, not company guidance. The figures of RMB 129, 24x, 16.6x, PEG 0.13, $82 billion, and $53 billion are all subject to the original report. The extremely low PEG may result from aggressive growth assumptions and does not automatically mean undervaluation. The 50% figure refers to China’s DRAM capacity share, not global share. Geopolitics, export controls, capacity expansion execution, and the memory price cycle could all rewrite this anchor.$CXMT