Kornic Automation's 3.15 Billion Won Contract Is a Hardware Resale, Not a Moat

Kornic Automation, a Korean maker of semiconductor equipment software and smart-factory systems, disclosed on September 11 that it signed a definitive contract to supply "IT infrastructure goods" to AJ Networks for 3.15 billion won — roughly US$2.3 million. Measured against the company's 2025 revenue, it is 10.47%. That is the number a headline loves: a small company landing business equal to a tenth of its sales in a single order.

The fine print argues the opposite of celebration. The contract is for IT infrastructure goods, not the control software or factory automation that is Kornic's actual franchise. It runs from September 11 to December 30 — a three-month delivery window — with no advance payment and 100% of the amount due on delivery. Kornic states it has no special relationship with AJ Networks and no similar contract history with the buyer over the previous three years. This is a one-off purchase order, not an embedded, repeat customer.

Who the buyer is matters. AJ Networks is a B2B rental company that manages assets in the hundreds of billions of won and is pushing hard into AI infrastructure: its own job postings describe renting and distributing NVIDIA GPU servers and HPC gear, plus a growing robotics fleet. Read "IT infrastructure goods" through that lens — a software-maker reselling hardware to a firm that rents and distributes servers — and the deal looks like pass-through supply rather than the differentiated software that carries Kornic's story. That reading is inference, not disclosure; nothing in the filing contradicts it, and plenty hints at it.

That distinction is the crux. Kornic's business divides into control software — its semiconductor-rooted differentiator — plus smart factory, robot logistics, and IT infrastructure. Smart factory was 48.9% of revenue in the first quarter of 2026. The IT infrastructure line is the commodity end: boxes procured and resold with no pricing power, no recurring element, and a customer free to change suppliers on the next order. This is the opposite of a bottleneck that earns supernormal margin; it is where a company books revenue while building none of the durability its valuation presumes.

And the size barely moves the underlying problems. Kornic lost money in 2025 — an operating loss of 2.7 billion won on 30.1 billion won of revenue, and a net loss of 2.9 billion. Around the announcement it traded near 1,080 won with a market cap close to 46 billion won (roughly US$34 million) and negative earnings per share. It is a theme stock that ran toward 3,700 won on the smart-factory and robotics enthusiasm and has since fallen by about two-thirds. A single hardware order worth a couple of million dollars against a loss-making, $30-something-million company is noise in the valuation, not a change in it.

For anyone still learning how to read a "contract win," keep this as the template. First size the number against revenue — 10.47% is the flattering framing the company chose to lead with. Then check quality: is this your differentiated product or a resale? Does the customer come back, or is this the first order in three years? Do the terms show commitment — advances, multi-year ties — or just delivery? A headline percentage says only that the contract is large relative to the company. It says nothing about whether the company earns anything durable from it.

The genuinely defensible part of Kornic — the semiconductor control software and the smart-factory automation built on it — is real expertise with a longer qualification story. This contract sits at the other end of that business, a thin, one-off hardware supply to a new customer. Keep the two separate. A real company can still be an unattractive stock, and a real, disclosed order can still tell you nothing about the moat.