Goldman Sachs Raises Hua Hong Semiconductor Target Price to HK$333
0xBroomberg
Goldman nearly doubled its target on Hua Hong Grace (01347.HK) from HK$174 to HK$333, maintaining a buy rating; this signals a bet not on near-term profit recovery but on the company's specialty-foundry scale-up along the AI power chain by 2030.
Target price nearly doubled — where is the valuation anchored?
The new HK$333 target is not based on 2026 or 2027 earnings. Goldman anchors it to a 2030 target P/E of 92×, then discounts back to 2027 at a 9.3% cost of equity.
This means → Goldman is effectively saying: on the next two years' profits, this stock cannot justify the price. The bet is on profit scale four years out.
In plain terms = this is not a "cheap right now" buy case. It is a growth-stock pricing that says "get on the bus now, wait for it to grow up."
How far do revenue and margins need to travel?
Goldman forecasts revenue rising from roughly $2.4 billion in 2025 to about $5.85 billion by 2029. Gross margin climbs from 11.8% to 22.5%; operating margin moves from near zero to roughly 12%.
The largest earnings revision is for 2029 — net profit raised 8% — driven mainly by AI power-supply and AI edge-device demand lifting 28 nm revenue, plus operating-expense leverage as scale increases.
This means → the "sweet spot" in Goldman's model is not next year but 2028–2029 — the window where depreciation absorption, product-mix upgrades, and AI order ramps converge.
Where is the 12-inch migration now?
In Q1 2026, monthly capacity hit 489,000 8-inch-equivalent wafers at 99.7% utilization. New-line revenue is already the majority of the top line.
High utilization ≠ high margins. The economics of 12-inch lines — moving processes from 8-inch wafers to larger ones so each wafer yields more chips — still depend on node mix, depreciation load, and customer pricing. Q1 capex was roughly $925 million, of which $886 million went to 12-inch.
Goldman splits the migration into three phases: ① fill capacity → ② migrate from 55/65 nm toward 40 nm and 28 nm higher-ASP products → ③ depreciation fully absorbed, margin amplification. The company is in early Phase 2 — revenue recovery is running ahead of profit recovery.
What does AI have to do with Hua Hong Grace?
Hua Hong Grace does not make GPUs or leading-edge logic chips. Its AI exposure runs through power and companion chips — AI servers need rising volumes of power conversion, power management, and analog chips.
The company's BCD process — a technique that puts bipolar, CMOS, and DMOS devices on one chip, specialized for power management — along with MCUs, embedded memory, and select RF chips, all find demand along this chain.
In plain terms = the hotter AI gets, the more servers are built, and every server needs more "power-supply parts." Hua Hong Grace is the foundry for those parts. Edge AI devices — smart appliances, consumer electronics — also pack control, sensing, and local-inference chips that land squarely on mature nodes like 28 nm and 40 nm.
What is the Hua Li Microelectronics acquisition for?
Hua Hong Grace plans to acquire 97.50% of Shanghai Hua Li Microelectronics to add 12-inch logic process capability and broaden its customer mix — not simply to bulk up revenue.
The deal still awaits regulatory approval. Goldman explicitly states the valuation cannot treat Hua Li as already consolidated.
The company's HKEX stock short name changed in June 2026 from "HUA HONG SEMI" to "HUA HONG GRACE." This reflects ongoing asset consolidation and platform restructuring within the Hua Hong group.
Which signals should trigger a reassessment?
Goldman's falsification checklist: whether Q2 revenue guidance reaches $690–700 million, whether gross margin holds 14–16%, 12-inch revenue share trajectory, growth at 65 nm and below, capex intensity, and Hua Li approval timeline.
This means → if any one of these six falls short, the current valuation faces compression — Goldman issued a high target but simultaneously drew a clear "prove-it" line.
Content is for reference only, not financial advice.