China's SMIC reports Q2 revenue up 21.8% YoY to ~$1.9B, above est., net profit down 59.1% YoY to ~$164.6M, above ~$103.8M est., and $2.25B capital expenditure
Context & Ripple Effects
SMIC entered this quarter after a weaker demand period: its prior Q1 revenue decline in 2023 paired with a steep profit fall. The return to year-over-year sales growth therefore marks a recovery in top-line demand rather than a clean return to prior earnings levels.
The comparison is especially notable because the same quarterly revenue level in 2022 came with far higher profit, while that earlier period was already shaped by lockdowns and harsher US sanctions. The current result adds a large capital outlay to an unresolved revenue-versus-margin gap.
First-order effects
- SMIC exceeded revenue and profit expectations, but its year-over-year profit decline shows that higher sales did not restore prior-period earnings.
- The $2.25B capital expenditure commits substantial cash to expanding or upgrading manufacturing capacity, increasing near-term investment intensity.
Second-order effects
- The combination of recovering revenue and heavy capex makes the timing of capacity coming online more consequential: spending precedes output, so the financial payoff can lag demand improvement.
- For customers, SMIC’s investment signals a supplier preparing for more production, but the profit decline leaves limited evidence that added volume will translate directly into stronger margins.
Third-order effects
- If revenue recovery repeatedly requires outsized investment while profit remains volatile, China’s chip-manufacturing buildout may become more capital-intensive and slower to monetize than sales growth alone suggests.
- This is a case of recurring sales resilience alongside pressured profitability: the durable question is whether new capacity can improve utilization and earnings once the investment cycle matures.
The trend: SMIC’s results fit a semiconductor-capacity cycle in which demand recovery can trigger heavy fab investment well before margins normalize.