Modulight Finally Looks Like a Company Again — Just Not the One Its Data-Center Fans Want Yet

Modulight Finally Looks Like a Company Again — Just Not the One Its Data-Center Fans Want Yet

Modulight Oyj is a €47.5 million market-cap Finnish laser maker that sells into cancer surgery and a handful of high-value niches. There is no price tape in the context to lean on, so the tape is silent and the transcript has to do all the work. It does. This was, by the company's own framing, the quarter where the operating story stopped being a promise and started being a P&L — and simultaneously the quarter where the market's favourite narrative (data-center lasers) got more airtime than it has ever had, with almost nothing quantitative attached to it.

The quarter that finally worked

Revenue rose 7% to EUR 1.9 million, which sounds thin until you set it beside the first half: for the nine months, revenue was still 7% below the prior year. So Q3 was not merely growth, it was the inflection that dragged the year back toward zero. The engine was the PPT business — photodynamic therapy — which hit EUR 500,000 in the quarter, up 110% year over year, and EUR 1.5 million for nine months, up 127%. CFO Anca Guina was explicit that this is not a one-quarter artefact: “The PPT business continued to grow strongly and reached EUR 500,000 in Q3, which is basically an increase of 110% from same quarter last year.” — Anca Guina, Chief Financial Officer · 2026-10-26

The profitability line is the genuine surprise. EBITDA of EUR 1 million is a 53% margin — the level, as the CEO put it, that the company "used to have before the IPO." For the nine months the margin is 28% against negative 15% a year earlier. Free cash flow from operations turned to plus EUR 240,000 from minus EUR 1.1 million. Nine-month free cash flow improved to roughly minus EUR 1 million from almost minus EUR 4 million. A company that had been burning cash to build a vertically integrated fab is now, at the margin, funding itself. That is a materially different company from the one that has been apologising for cash consumption for several years.

And then the number that actually matters most: order intake of EUR 3.5 million, up 157%. The composition shift is the real story inside the story, and the CEO said it twice — once in prepared remarks about the order book being recurring orders "rather than one-offs," and again under questioning: “most of them, probably more than ever, were related to just customers doing pilot orders or production orders and pushing forward with their operating business.” — Seppo Orsila, Founder and Chief Executive Officer · 2026-10-26 For a company whose revenue line has historically been hostage to lumpy non-recurring engineering work, a product-order-dominated intake is a structural change, not a seasonal one.

The indium phosphide question, and why it is being asked at all

Here is the tension. Modulight reported a competent industrial turnaround and spent its Q&A being interrogated about data centers. The reason is a piece of corporate archaeology: the company was founded to commercialise indium-phosphide laser technology, and those devices now happen to resemble what the optical-communications supply chain wants.

more than 26 years ago, Modulight was founded to commercialize indium phosphide short-ridge laser technology… many of those lasers that were invented and developed at that time are closest matched to what today is called the data center inside and outside the data center communication.

Note what is there and what is not. There is spare capacity — a by-product of buying high-volume manufacturing tools for repeatability — and there is lead time that management says could be "even a short one" for products already made for specialty applications. What there is not is a contract, a customer name, a capacity number, or a revenue implication. The company is in discussions with multiple partners and will announce when material. That is a reasonable, disciplined answer. It is also, for anyone buying the AI-optics angle, precisely nothing.

The keyword record makes the novelty explicit. data center entered this company's top-five themes for the first time this quarter, alongside semiconductor technology — where management volunteered that demand from major semiconductor-equipment manufacturers has probably exceeded that from flow cytometry or defense. That is a real, if small, revenue fact, and it sits inside a global wave where the tickers attached to AI data center were up strongly over 360 days (43 advancers against 11 decliners) but have decisively rolled over in the last 90 (6 up, 48 down), with the optics complex — co-packaged optics, high-bandwidth memory, gigabit-per-lane names — among the worst of the recent decliners. Modulight is arriving at the data-center story with optionality just as the tape has stopped paying for optionality. Whether that is early or late is the entire investment question, and the call gives you no help.

What management would not say

The deflections are as informative as the disclosures. On VISUDYNE supply — the pharmaceutical bottleneck that, if eased in 2027, would lift treatment volumes on Modulight's device — the CEO simply refused the frame: “I'm sorry that I have to deflect a bit that Modulight is not issuing the guidance on revenue or specific revenue line.” — Seppo Orsila, Founder and Chief Executive Officer · 2026-10-26 The actual pharmaceutical supply chain was ruled out of scope: "not our business to comment." Fair, but it means the single largest identifiable tailwind for the franchise is one the company will not underwrite.

Then there is the slip. Guina, explaining that costs cannot stay flat forever, offered an illustration — “If next year I come and tell you that our revenue is more than double, most likely you will not see the same operating expenses.” — Anca Guina, Chief Financial Officer · 2026-10-26 — and her CEO immediately walked it back: “Anca's intention was not to anyhow guide the future revenue.” — Seppo Orsila, Founder and Chief Executive Officer · 2026-10-26 That is the sound of a company that has been burned by forward numbers. The retained outlook is qualitative only: grow this year, grow next year, return to strong profitability. Meanwhile the cost decrease discipline that produced the 53% margin is explicitly not a permanent state — Guina conceded opex must rise with scale.

The read

Strip out the data-center romance and you have a micro-cap that just posted its second consecutive quarter of positive EBITDA trend, a product-order-dominated intake, halved cash burn, and a therapy franchise compounding at triple digits off a small base. Strip out the turnaround and you have a €47.5 million company whose most exciting asset is a capability it is not yet selling. Both are true. What changed this quarter is that the first sentence no longer needs the second to be investable — and what did not change is that management still won't tell you which of the two you are buying.