$XFAB DD: the European chip foundry that just went viral, and the actual bull case sitting underneath the hype 🇪🇺🔌
TL;DR: X-FAB (XFAB.PA, OTC: XFABF) is a profitable specialty semiconductor foundry with genuine exposure to the two hottest niches in chips right now: power semis (silicon carbide and gallium nitride for EVs and 800V systems) and silicon photonics (the optical plumbing AI datacenters are moving toward). It has CHIPS Act money on both continents. Full honesty up front: it ripped \~76% in a single day on May 27 off a viral tweet with zero company news, it's already fading, and every fair value model says it's 2 to 3x too rich right here. The business is a real long-term bull case. The stock two days after a parabola is not a chase. Entry matters more than the thesis on this one. Positions at the bottom.
This is real DD, including the parts that hurt. If you came for the pump, this ain't it. 🫡
What it actually does
X-FAB Silicon Foundries. Founded 1992, \~4,300 employees, fabs across Germany, France, Malaysia and the US. They are a specialty/analog foundry, meaning they don't chase the bleeding-edge 3nm logic that TSMC fights over. They make the unsexy but mission-critical stuff: analog and mixed-signal chips, MEMS, sensors, and crucially silicon carbide (SiC), gallium nitride (GaN), and silicon photonics. Customers are automotive, industrial, and medical, where parts have to not die for 15 years.
Listed on Euronext Paris as XFAB.PA, currently around €10.70, market cap roughly €1.4B. It briefly tagged €15.88 and a €2.06B cap during the Tuesday spike before fading back.
The bull case 🐂 (this is a real one, long-term)
1. Power semiconductors are the EV/grid pick-and-shovel. SiC and GaN are how you move high voltage efficiently. Every EV, fast charger, solar inverter and increasingly every AI datacenter power stage wants them. X-FAB has been building this capability for years rather than bolting it on last quarter. As 800V vehicle architectures and high-density power delivery scale, specialty foundries with proven SiC/GaN lines get pulled along.
2. Silicon photonics is the AI angle that isn't priced into most names yet. AI clusters are hitting a wall on electrical interconnect: copper can't move data between thousands of GPUs fast enough without cooking. The industry answer is moving the optics onto the package (co-packaged optics). That runs on silicon photonics, which X-FAB actually fabs. This is the part of the AI buildout almost nobody on this sub is talking about, and it's the part of X-FAB's story with the longest runway.
3. Government is literally paying them. X-FAB has pulled in around €128M from the EU CHIPS Act and roughly $50M from the US CHIPS Act for capacity. Free-ish capital to expand the exact lines (power semis, photonics) that matter, plus political cover as both blocs onshore chip supply.
4. It actually makes money. This is not a cash-burning meme. X-FAB has posted positive earnings every year for the last five and positive operating cash flow each year. Trades around €0.14 EPS. For a small-cap chip name with secular tailwinds, real profitability is rare and it means the company isn't dependent on dilution to survive.
The bear case 🐻 (read this twice, it's the whole point)
I'm bullish on the company. I am not going to lie to you about the stock right now.
It just pumped on a tweet, not on news. On May 27 X-FAB spiked as much as 76% intraday and got halted multiple times for volatility. There was no earnings, no contract, no upgrade, no takeover. The entire move came from one viral X account ("Serenity") posting a long idea, whose follower count went from \~58k to 411k in the process. The CEO basically responded with "we've noticed our share price went up, no comment." When a stock triples YTD with the catalyst being a tweet, you are not early. You are the audience.
Valuation is detached from the business. Morningstar fair value is around €4.72 versus \~€10.70 now, a \~247% premium. Analyst consensus price target sits near €5.34, implying roughly 65% downside. Normalized P/E is around 75 to 80. The fundamentals did not change on May 27; the price did. That gap closes eventually, usually downward.
The actual business is soft right now. Management called the automotive market weak last month and is only guiding for a "steady recovery" in H2 2026. Q2 revenue guidance is $190 to $200M at a 17 to 20% EBITDA margin. Solid, not explosive, and nowhere near enough to justify a near-tripling of market cap in five months.
You probably can't trade it cleanly. It's a Euronext Paris listing. The US OTC line (XFABF) is thin and the spreads are wide. If you pile in and it reverses, getting out at a sane price is not guaranteed. Low liquidity rips both directions.
Viral attention fades faster than it arrives. It already gave back a big chunk from the €15.88 high. The same account was reportedly linked to a sharp move in Raspberry Pi earlier this year. Pump narratives mean-revert once retail stops buying, and you do not want to be the last one holding when it does.
The play 🎯
This is a long-term story, not a Tuesday YOLO. If you genuinely believe in SiC/GaN power semis and silicon photonics over the next 2 to 3 years, X-FAB is a legitimate way to own that theme, because the business is real and profitable and the tailwinds are real. But the move here is to buy the eventual pullback, not the spike. Let the viral froth bleed out, let it mean-revert toward a price where the fair-value gap isn't screaming at you, and build a position into weakness.
Chasing it here, two days after a 76% no-news candle, is buying someone else's exit. Don't be the exit.
I'm playing it as a watchlist-and-accumulate-on-dips name, not a market buy today. Size small, this thing is volatile as hell.
Positions: 159 shs
