SMIC vs Hua Hong: which Chinese foundry captures more of Beijing’s chip push?

SMIC generated FY2025 revenue of $9.33B, nearly four times Hua Hong's $2.40B, and recovered EBITDA to $4.42B as utilization improved, positioning it as Beijing's principal advanced-node and fab-investment vehicle. Hua Hong posted record Q2 2026 revenue and 385% profit growth, but its FY2025 gross margin was only 11.8%, ROIC was -0.4%, and its valuation is elevated at 301.7x trailing and 155.8x forward P/E. The September 2026 electronics plan targets 30 trillion yuan of sector revenue by 2030 and supports both companies, though SMIC is better placed for large-scale advanced-wafer policy capture while Hua Hong remains a higher-risk specialty-node recovery play.

Analysis

The investable distinction is not simply advanced versus specialty nodes; it is return on incremental state-directed capex. SMIC (0981.HK) should capture a disproportionate share of subsidies, domestic-design wins and strategic customer allocation, but equipment restrictions can turn this into a volume-led, low-return growth model rather than a margin-inflecting one. Hua Hong (1347.HK) has greater operating leverage if mature-node utilization tightens, yet its equity already embeds a near-flawless conversion of new capacity into higher pricing and cash returns.

A long 0981.HK/short 1347.HK structure best expresses policy allocation while neutralizing much of China semiconductor and Hong Kong risk appetite beta. The spread should work over 1-3 months if implementation details identify recipients, procurement commitments, or financing terms; it can also benefit if investors rotate from momentum into companies with more credible strategic scarcity. The key near-term risk is that retail/policy momentum continues to reward Hua Hong's earnings acceleration regardless of valuation, while SMIC remains constrained by tool access and absorbs depreciation ahead of usable leading-edge output.

The second-order beneficiary of domestic mature-node substitution is not necessarily Hua Hong alone: Chinese analog, power-management and automotive semiconductor designers can gain supply assurance, but oversupply may transfer economics from foundries to customers. Conversely, a broad expansion of China fabrication capacity is structurally negative for mature-node pricing and therefore for global exposed foundries such as UMC (UMC) and Tower Semiconductor (TSEM), particularly if Chinese suppliers prioritize utilization over returns. A meaningful tightening in power-device lead times, sequential gross-margin expansion, or evidence that new capacity is being allocated to higher-value automotive/industrial wafers would falsify the Hua Hong short leg.

Consensus appears too willing to equate policy support with shareholder returns. State capital reduces financing risk but can encourage capacity additions that suppress industry utilization; the more durable bullish signal would be rising free cash flow after subsidies and lower capital intensity, not revenue growth alone. For SMIC, the upside case requires evidence that domestic tool and materials substitution is improving yield and throughput rather than merely replacing constrained imported equipment at lower productivity.

Key Decisions for Investors

• Initiate a 3-month market-neutral pair: long 0981.HK / short 1347.HK in equal HKD notional. Target 15-20% spread appreciation; cut if the spread widens 10% against the position or if Hua Hong delivers sequential margin expansion materially ahead of expectations.
• Do not chase 1347.HK outright after momentum-driven advances. Reassess for a tactical long only after two consecutive quarters of utilization and gross-margin improvement accompanied by positive operating cash flow after capex; absent that evidence, valuation leaves asymmetric downside on any guidance miss.
• Use UMC or TSEM as a 6-12 month watch-list short hedge against Chinese mature-node capacity rationalization failure. Trigger only if industry utilization, wafer pricing, or management commentary confirms renewed price competition; this is not yet a standalone recommendation without current utilization and valuation data.
• For 0981.HK, add on confirmation of funded project awards, domestic equipment qualification progress, or customer wafer commitments rather than on policy headlines. Thesis is impaired by new export-control measures, a material debt-funded capex acceleration without utilization gains, or failure of gross margin to improve over the next 2-3 reporting periods.