In a quarter where revenue fell 7% and X-FAB posted a loss, what got confirmed was a 2028 story

X-FAB's second-quarter 2026 revenue came to $199.8 million, down 7% from a year earlier. EBITDA fell 35% to $33.6 million, and the quarter produced a net loss of $14.8 million. What this quarter confirmed was not what the company earns now but where it stands in 2028.

X-FAB second quarter 2026 results (X-FAB press release, 2026-07-30)

출처: finanznachrichten.de

What X-FAB actually makes

A foundry builds chips on behalf of whoever designed them. The best known ones make the phone and AI chips at the leading edge. X-FAB is not that. This company makes analog and power semiconductors. Chips that switch electricity on and off and change its direction, chips that turn light and sound into electrical signals. Its main customers are automotive and industrial equipment, and those markets are weak right now. That is where the revenue decline sits.

So what got confirmed

Separate from the results, the release carried three things. Three silicon carbide design wins, several data center design wins, and a line holding volume production of photonics chips at 2028. Silicon carbide is a material that withstands higher voltages than silicon. As data center power distribution moves up to 800 volts, conventional silicon parts struggle and this material gets used instead. The company naming that transition as a tailwind for its own business is what this release contained. Photonics sits in the same place. When the links between servers move from copper to light, someone has to build the chips that handle that light, and X-FAB puts its production at 2028.

The income statement and the thesis are in different years

The second-quarter numbers were made by automotive and industrial equipment. The design wins and the 2028 date are a data center story. A design win is not revenue. It is a customer committing to build a chip on your process, and it usually takes two to three years before that chip actually sells. So this quarter's figures and this quarter's news point at different years.

So what settles it

The question comes down to one thing. Whether the current weakness is the cost of crossing a transition, or a sign that this company's share of the new demand is small. 2027 settles it. If those three silicon carbide design wins start showing up as 2027 revenue, the crossing is under way, and if revenue is still tied only to automotive and industrial then, the data center story lived only in the slide deck. Third-quarter guidance is $195 million to $205 million with an EBITDA margin of 17% to 20%. That looks above this quarter's 16.8%, so whether a recovery has started shows up first in the next set of numbers.