Sivers Semiconductors: Redeye's Wide Bull-Bear Spread Hinges on Whether Glasgow's Laser Lines Find Buyers

Redeye sees a wide valuation spread for Sivers Semiconductors as its $30M indium phosphide expansion hinges on demand that has yet to become firm orders.

Redeye's latest take on Sivers Semiconductors lays bare a valuation gap that few analyst notes manage to contain within a single page. The spread between the firm's optimistic and pessimistic scenarios is unusually wide, a reflection of just how unsettled the outlook remains for the photonics supplier.

The shares currently change hands at EUR 2.81. Even after a recent advance, they sit roughly 6.9 percent below their 50-day moving average — a sign that the rebound has yet to feed through into the medium-term trend line.

A $30 Million Bet on Indium Phosphide

At the center of the debate is the company's decision, announced a little over a week ago, to expand its indium phosphide fabrication in Glasgow. The roughly USD 30 million project is designed to lift capacity beyond 100 million continuously operating DFB lasers per year — components that serve as key building blocks for optical networks and the data traffic flowing through AI data centers.

Production is slated to come online in the fourth quarter of 2027. Until then, Sivers is leaning on a demand surge that exists as expectation rather than as hard order volume. The stock has added 6.9 percent since the announcement, a muted response that suggests the market is extending only cautious credit to the story.

The Second-Quarter Report as the Valuation Anchor

Redeye's assessment rests on figures Sivers published about two weeks ago in its interim report for the second quarter of 2026. Revenue came in at SEK 53.8 million, or SEK 55.6 million adjusted for currency effects. Adjusted EBITDA remained deeply negative at minus SEK 35.5 million.

Management nonetheless pointed to product growth, a record-level order pipeline and customer projects now ramping — positioning the company, in its own words, for an acceleration in growth.

That claim frames the analysts' divergent scenarios. Should the acceleration management sketches materialize, the bulls see price targets well above the base case. If losses persist and the announced customer ramps slip, the bears foresee markedly lower valuation levels.

From Development Revenue to Serial Volume

In the same interim report, Sivers described its own transition from project-based development revenue toward scalable product sales, driven by production start-ups at customer sites. That phrasing sits at the heart of the investment case.

For years, Sivers earned its money on development contracts — predictable, but thin-margined and non-recurring. The real value driver is whether customers actually move from the development phase into serial orders large enough to keep the new Glasgow capacity busy. Without that shift, the expansion would amount to an outlay with no takers. With it, the fab could become the foundation of a new earnings base.

AI Infrastructure as the Demand Engine

Should the build-out of optical networks for AI data centers prove as strong as the industry expects, Sivers would be positioned early with its Glasgow expansion. The company justifies the capacity step with added process capabilities, greater automation and more manufacturing flexibility — all aimed at responding to growing demand from precisely that segment.

If forecasts that laser-based optics become a bottleneck for data centers prove correct, Sivers could benefit as a specialized supplier from the economies of scale of the new plant once it starts up in 2027.

The Case Against: Capex Without Committed Volume

A serious counterargument stands in the way. The investment decision rests on anticipated demand, not on booked serial orders of comparable magnitude. More than a year will pass before the planned fourth-quarter 2027 start-up, a window in which market conditions, the competitive landscape and technology paths can all shift. The interim report itself shows that the move from development to product revenue is still underway and far from complete.

Adding to the pressure: the stock trades roughly 73 percent below its 52-week high, evidence of how sharply expectations have already been revised this year. If the hoped-for demand wave fails to arrive or the customer ramp is delayed, high fixed costs from the new plant could weigh without matching utilization.

What to Watch Before 2027

For the months ahead, the valuation of Sivers Semiconductors hangs on a single variable: the pace at which development customers turn into paying serial buyers. If that trend holds and future interim reports confirm the shift described in the second quarter, the Glasgow capacity should be ready at the right moment.

Should the momentum reverse — because AI infrastructure spending grows more slowly than hoped, or because competitors scale faster — the investment would remain, for now, an expensive wager on a market that has yet to materialize.

The next concrete test is the Glasgow plant's announced fourth-quarter 2027 start-up. Until then, further quarterly reports should show whether the revenue mix is genuinely tilting toward serial production. Volatility in the shares is likely to stay elevated in the meantime.