MARKET WATCH
AXTI has had one of the most extraordinary runs of any semiconductor stock in recent memory. Let's put it in context.
Three overlapping catalysts drove the rally:
AXT, Inc. (founded 1986, Fremont CA) manufactures compound semiconductor substrate wafers — the base material on which chips are grown. This is not the chip itself, it's the foundation.
| Substrate | Application | AXT Role | Demand Driver |
|---|---|---|---|
| InP Indium Phosphide |
800G optical transceivers, LiDAR, 5G base stations | ⭐ Core growth engine | AI data centers, autonomous vehicles |
| GaAs Gallium Arsenide |
Power amplifiers (smartphones, WiFi), LEDs | Mature business | 5G device upgrades |
| Ge Germanium |
Solar cells (space, CPV), fiber optics | Niche, stable | Space/defense |
| Metric | FY2025 | FY2024 | FY2023 | FY2022 | Signal |
|---|---|---|---|---|---|
| Revenue | $88.3M | $99.4M | $75.8M | $141M | -11% YoY, recovering from trough |
| Gross Margin | 12.7% | 24.0% | 17.6% | 36.9% | Collapsed — key risk ❌ |
| Operating Loss | -$22M | -$14.8M | -$21.6M | +$12.6M | Still burning ❌ |
| Net Loss | -$21.3M | -$11.6M | -$17.9M | +$15.8M | 3 consecutive loss years ❌ |
| Free Cash Flow | -$18.8M | -$17.9M | -$7.1M | ~$0M | FCF negative 3 years ❌ |
| Cash | $128M | $33.8M | $50.1M | $41.4M | Raised capital — 3x jump ✅ |
| Book Value/sh | $6.22 | $4.47 | $4.78 | $5.19 | P/B = 10.3x (stock at $64) ❌ |
| Revenue FY26E | $127.8M (+44.7%) — consensus estimate | Strong growth expected ✅ | |||
| EPS FY27E | $0.27 — first projected profitable year | P/E ~240x at $64 ⚠️ | |||
At $64.22, AXT is valued at $3.57B market cap on $88M trailing revenue and -$21M net income. Here's how that compares to reality:
If we apply a generous 8x P/S (premium semi materials multiple), AXT would need $450M revenue to justify today's price. Their best year ever was $141M (FY2022). At +44%/year (consensus), they'd reach $450M around 2029–2030. That's a lot of "wait and see."
Or at the current $3.57B market cap, using a 30x P/E (reasonable for high-growth semi), they need $119M net income. At a 15% net margin (aspirational), that requires $793M in revenue. Never been close to that.
| Indicator | Value | Signal |
|---|---|---|
| RSI (14) | 71.47 | 🔴 Overbought — previous visit to 70+ led to -16% single day |
| vs SMA 20 | +43.78% | 🔴 Extremely extended — 44% above 20-day average |
| vs SMA 50 | +106.66% | 🔴 Double the 50-day average |
| vs SMA 200 | +415.35% | 🔴 5x the 200-day average — parabolic |
| ATR (14) | $6.68 | ⚡ Daily swings of ~$7 typical — very high volatility |
| Short Float | 11.91% | ⚠️ Elevated — could fuel short squeeze OR orderly decline if momentum fails |
| Rel. Volume | 1.77x | Above-average volume today — momentum buying |
| Beta | 1.54 | 50% more volatile than the market on average |
| Level | Price | Significance |
|---|---|---|
| 🔴 Current / Resistance | $64.22 | 52W High — new ATH territory |
| 🟡 Key Support 1 | $48–$52 | Post-earnings base (Feb 19) |
| 🟡 Key Support 2 | $32–$38 | Previous analysis zone, post-correction floor |
| 🟢 Strong Support | $20–$25 | Analyst target zone — fundamental floor |
| 🔴 Bull Invalidation | Below $18 | B. Riley lowest target — thesis fully broken |
The underlying business is legitimate and the InP/AI demand is real. But at 40x P/S with a -$21M net loss and gross margins at 12.7%, the stock is priced for a perfection scenario that may take 4+ years to materialize. The risk is not "the business fails" — it's "the multiple corrects."
At 40x P/S, AXTI is priced like a SaaS company with 80% gross margins and 30%+ growth. Their gross margin is 12.7% and revenue is declining YoY. When momentum fades, the multiple re-rating can be brutal and fast (see SMCI -80%, ANET corrections, etc.)
JQA produces in China. US chipmakers buying InP substrates from a Chinese entity creates policy risk. Both Needham (Jan 20) and B. Riley (Jan 9) cited this when downgrading. Any new semiconductor export control affecting compound semi materials hits AXTI directly.
12.7% gross margin in FY2025 is the main fundamental concern. The company argues this is temporary — as InP demand scales, fixed costs get absorbed and margins recover. Q4 2025 showed early signs of improvement. But revenue declined in FY2025 despite InP momentum — the GaAs/Ge drag is real.
AXTI is a real company with a real AI angle — InP substrates for 800G+ transceivers is genuinely a high-growth market and AXT is one of the few companies that can supply it at scale. The cash raised ($128M on balance sheet), the institutional buying, and the revenue growth trajectory are all real.
But at $3.57B market cap on $88M revenue and -$21M net income, the stock has priced in 5+ years of perfect execution. The two analysts covering it both downgraded before the biggest part of the rally — their combined target of $24.50 represents a -62% correction to "fair value."
Our rating drops from B+ to B — not because the thesis weakened, but because the price more than doubled. The story is the same. The risk/reward is not.