NAVER Posted Record Revenue — So Why Did the Stock Crash 7%? Inside NAVER's $10B AI Factory

 

🧠 NAVER Posted Record Revenue — So Why Did the Stock Plunge –7%?

Take the numbers alone and this report card was flawless. In Q2 2026, NAVER's consolidated revenue came in at roughly 3.3888 trillion won, breaking its all-time quarterly high. Its established core businesses — commerce, pay, and advertising — kept up double-digit growth. Yet on the day results were announced, the stock instead fell nearly 7%. To understand this scene, which at first glance seems to make no sense — record revenue paired with a plunging share price — you first have to pin down where the market is looking at NAVER right now. To state the conclusion up front: the market's eyes are not on the smoothly running core business, but on the "AI Factory" whose game has only just begun.

📉 Revenue Rose, So Why Did Profit Slip Backward?

The part that truly deserves attention in these results isn't revenue — it's operating profit. Even as revenue grew at a double-digit pace, operating profit fell about 0.2% versus the same period a year earlier. It may look like a trivial decline at a glance, but the very fact that profit slid backward in a quarter of record revenue is an important signal. The cause is clear. Q2 operating costs swelled to 2.8684 trillion won, up 19.8% from a year ago, and a substantial portion of that increase came from AI infrastructure investment.

Let me pause to unpack one concept here. Operating profit is the figure left after subtracting the costs of running the business from the money a company earns through its core operations. No matter how much revenue grows, if costs grow even faster, profit shrinks. NAVER has entered exactly that phase. It is ramping up spending first for the sake of the future, but the point at which that spending brings money back in has not yet arrived. The very spot the market grew anxious about is this time lag.

🏭 So What on Earth Is This "AI Factory"?

The name AI Factory momentarily brings a factory to mind, but it isn't a place that stamps out products. It's a business that builds up GPUs and data centers on a massive scale and rents out the enormous computing power that comes from them to other companies or countries. Put simply, it's closer to building a "computing power plant" needed to run AI, and selling the electricity from it.

Let me unpack a few unfamiliar terms here. A GPU (Graphics Processing Unit) is a semiconductor originally made to handle game graphics, but because it specializes in processing countless calculations simultaneously, it has become the core component for training and running AI models. This is why today's AI competition is essentially called a war to secure GPUs. A data center is the giant building that gathers thousands to tens of thousands of these GPUs in one place and runs them 24 hours a day. On top of that, it must be equipped with cooling systems, a power grid, and backup power — which is why the initial investment scale leaps into the trillions of won.

The market NAVER is aiming at doesn't stop at the domestic one. It is targeting so-called Sovereign AI demand as well. Sovereign AI refers to AI that a nation trains and operates on its own, using its own data, language, and culture on infrastructure within its own territory. As more countries grow reluctant to entrust sensitive national data to the overseas servers of U.S. Big Tech, the trend of each nation wanting to build its own AI infrastructure is growing. NAVER is drawing a big-picture plan to grow beyond a platform company centered on search, commerce, and advertising into an infrastructure provider that supplies this computing resource itself to outside parties.

💰 A 14-Trillion-Won Table — Who Puts Up the Money?

The stakes of this business are known to be roughly 10 billion dollars, about 14.6 trillion won in our currency. It's an amount too heavy for a single company to shoulder alone — and here is where the deal structure gets interesting. NVIDIA comes in as an AI computing partner by making an equity investment of about 1 billion dollars (roughly 1.4 trillion won) into NAVER to secure a stake, while Brookfield, a Canadian alternative-asset manager, invests up to 9 billion dollars to set up a special purpose vehicle — an SPV.

The term SPV (Special Purpose Vehicle) may be unfamiliar. It refers to a company established separately for just one specific business or project. This SPV buys or owns the GPU and data-center equipment, and directly raises the enormous funds needed for it. The crux of this structure is that NAVER does not have to shoulder that heavy equipment-investment burden entirely on its own balance sheet. This is also why the stock jumped and the market cheered when the deal structure was disclosed — because it was a clever design that shed a substantial part of the financial burden.

🔍 But Here's a Question We Rarely Ask

Usually, when we encounter big investment news like this, our attention gets stolen by "how big are they building it?" and "how many GPUs did they secure?" But the truly more important question lies elsewhere. Once this enormous computing power plant is fully built, who exactly will pay money to buy and use that electricity?

No matter how grandly you build a factory, without orders it's just an expensive empty building. The AI Factory is exactly the same. No matter how many GPUs you pack in, if no customer appears to actually lease that computing power and pay for it, the trillions poured in so far will merely gnaw away at each quarter's results as a cost called depreciation. Depreciation is the accounting treatment of reflecting expensive equipment as an expense spread across its usage period — and when revenue doesn't back it up, this cost acts like a pure loss. The root cause of the market's current anxiety over NAVER is precisely this "uncertainty of demand."

Digging one step deeper, Sovereign AI demand is itself a future market whose substance hasn't fully taken shape yet. The direction — that each government wants its own national AI — is clear, but when, with how much budget, and through which operator it will materialize remains in the fog. There is also no guarantee yet that NAVER will establish itself as the main supplier of that demand. In short, a cool-headed recognition is needed: at this point in time, the AI Factory is closer to "a bet based on expectations" than to "confirmed revenue."

🧭 So What Should We Watch Going Forward?

From an investor's perspective, the real points to watch boil down to three.

First, whether the deal with Brookfield actually reaches final confirmation. The structure known so far is closer to a preferred-negotiation stage, and procedures remain to be cleared before the 9-billion-dollar fund is actually deployed. This financing must be wrapped up as planned for the business to fasten its first button properly.

Second, whether GPU supply proceeds on schedule. With the whole world lined up over NVIDIA GPUs, securing the planned volume on time is itself a form of capability. A supply delay leads directly to a delay in when revenue occurs.

Third, and most decisive of all, whether customers who will actually fill this AI Factory emerge. The company forecasts that AI Factory revenue will become visible from the first half of 2027, but only when that forecast is backed by concrete customer contracts will the market's doubt turn into conviction.

To sum up: NAVER's next re-rating hinges not on "how big it invested in AI," but on "the very moment AI connects to actual revenue and profit." Re-rating means the market's eye-level for viewing a company — the valuation standard itself — stepping up a notch. The current share-price decline is less a signal that NAVER has failed, and closer to the market demanding, "We've seen the investment scale, now show us the results." Rather than being swept up by dazzling blueprints and trillion-won figures, this is a time that calls for a gaze that calmly waits and watches for the moment the lights come on in that factory and the first customer's invoice is issued.

This article was written for the purpose of conveying information; it is not investment advice recommending the purchase or sale of any specific stock. Judgments regarding investment and the responsibility that follows rest entirely with the investor, so before actually proceeding with an investment, please be sure to check disclosure materials and expert opinions together.