Record Profit, Yet the Share Price Down 25 Percent: An Audit of Samsung's Quarterly Ledger

On the left side of the ledger, a number glows — 107.4 trillion won. On the right side, another number quietly slides downward — the share price is more than 25 percent below its June record high. One company, one quarter, yet two numbers giving contradictory testimony. Samsung Electronics' third-quarter operating-profit guidance is roughly nine times higher than the same period a year earlier — on paper, one of the best quarters in its history. Yet the equity market is not applauding. That mismatch is the subject of this piece.

I have spent many years logging matches into ledgers, turning every game into an auditable number-trail. That habit now places me on an unfamiliar pitch — not a stadium, but a corporate earnings statement. The method does not change. Ledger first, then phases, then caveats. I read Samsung's quarter with exactly that eye: what is written, what is not, and what hides in the gap between the two.

A methodological confession up front. This is a piece about a company's financial results, not a game. In any analysis I first separate the tiers of sourcing, and I do so here. Samsung's own guidance is the primary tier — an announcement coming directly from the company. The LSEG SmartEstimate is the second tier — a weighted analyst-consensus estimate, sitting at 106.1 trillion won. TrendForce is the third tier — a specialist market-research house that tracks chip-price paths. The fourth tier is individual analyst commentary — Kim Seok-hwan of Mirae Asset Securities, Douglas Kim of Douglas Research Advisory. Overall, these sources carry medium-to-high credibility. But remember: the loudest headline and the weakest forward view often sit side by side in the same release.

Some context is needed. Under this quarter's (July–September) guidance, operating profit is 107.4 trillion won, roughly equivalent to 80.17 billion dollars. Revenue is about 195 trillion won, up 127 percent year over year. This is the fourth straight quarter in which the company posts record operating profit. The story looks simple: the artificial-intelligence wave has inflated demand for memory chips, lifted prices, and the benefit of those prices has been deposited into Samsung's books. But when I read a ledger, I never look only at the total — I look at which segment deposited how much, and which segment ate away at it.

The core judgment is here: this record profit is a one-engine vehicle. The bulk of the profit came from the memory segment. Yet in the same quarter, the mobile segment lost more than 1 billion dollars, and the foundry segment is running at low capacity utilisation while loss-making. In other words, extraordinary profit in memory on one side, quiet erosion in two divisions on the other. The total is a record, but the structure is fragile.

That fragility can be put into numbers. The memory segment is the main driver of the earnings improvement — that fact is the most important one. Mobile and foundry sit hidden in the shadow of memory's profit. A simple question follows: if memory price growth stalls, whose shadow will cover the deficits of the other two segments? There will be no shadow. The very structure of the ledger says the current profit stands on a temporary engine, and questions about that engine's speed have already begun.

Proof of the deceleration is in the guidance itself. Fourth-quarter profit growth is forecast to slow to 8.2 percent quarter-over-quarter, a sharp cooling from the 20 percent QoQ growth of the third quarter. More specifically, DRAM contract-price growth is forecast to fall from about 60 percent in the second quarter to 10–15 percent in the fourth. That dramatic slowdown in commodity-price growth is the most honest signal — not the headline, but this number tells you where the story is turning.

There is an external cause for this slowdown too — currency. A stronger won reduces the won value of dollar-denominated overseas sales. Analysts are therefore trimming forecasts. This is a pull beyond the company's control, making the quality of the profit one notch more uncertain.

Now to the divergence that anchors this piece. The gap between record profit guidance and a share price down 25 percent is the single most important data point. The share price was flat-to-down through the middle of the third quarter, more than 25 percent below the June high. The market is not celebrating the current profit number; it is pricing future risk. The question has changed — the market no longer asks "how big is the profit?" but "how long does it last?"

The competitive field also makes the maths clear. In HBM (high-bandwidth memory), Samsung is pushing ahead — bit shipments rose about 50 percent QoQ, and the gap to leader SK Hynix is narrowing. In conventional DRAM and NAND, Samsung stands at rough parity with Micron and SK Hynix. But the foundry picture differs: TSMC is the clear leader, and Samsung is a trailing follower. The recent gleam of memory profit does not erase the structural gap to TSMC in foundry.

The rise of Chinese competition is a medium-term risk. If Chinese memory makers keep adding supply, the supply-demand imbalance forecast to run into 2028 could clear faster than expected and compress pricing power. On the ledger this point is still small, but the direction is worth watching.

Here I must open a caveat box I keep in every piece: "What the data cannot say." First, four straight record quarters are a robust backward sample, but a single up-cycle cannot prove forward durability. The sample is limited. Second, where memory prices will peak is not in this data. Third, the exact size of the mobile and foundry losses has not yet been broken out. I do not guess at data that has not arrived — I simply wait for it.

This is where the most tempting error hides, and it is exactly where my contrarian view sits. The natural instinct is to take record profit as proof of a company's health. But here, conflating two things is dangerous. Record profit is a backward number; a 25 percent share-price slide is a forward signal. Their coexistence is not a contradiction — it is a mature message. When the market marks prices down even in the face of record results, it is saying the peak is moving away. I do not worship models; I reconcile them with the muddy receipts of the season. And these receipts say profit growth is flattening and the price upcycle is cooling.

A second contrarian point concerns data discipline. This company's results may first sit under the shadow of a wrong category label — a reminder that labels cannot be trusted blindly. Similarly, the gap between the blazing language of headlines ("world first," "80 billion dollars") and the market's cold behaviour is the real signal. Enthusiasm is in the media; scepticism is in the price. When the gap between headline heat and price coldness widens, the market is usually the one that is right.

Three conclusions emerge from the core analysis. First, this is a cycle-peak result — a record, yet already carrying signals of deceleration. Second, profit is concentrated in one segment (memory), with two divisions in loss. Third, the divergence between the share-price slide and record profit is the most honest expression of the market's forward view.

Seen through the value chain, the transmission pushes from the top down. Memory price inflation benefits the upstream oligopoly (Samsung, SK Hynix, Micron) but raises input costs for the company's own device divisions. This is a kind of internal cannibalisation — its own profit becoming its own cost. AI-infrastructure demand is the upward driver, but that is a forecast, not a certainty.

And the most curious divergence: in this corporate ledger there is no player, team, or match I recognise — no formation, no pressing metric, no referee. So anyone forcing this result into a sporting frame would be committing ledger fraud. I do not commit fraud; I write what is there, and I say plainly what is not.

Five signals are worth watching ahead. First, the October 29 detailed release and shareholder-return disclosure — it could expose the size of the segment losses and add pressure to the share price. Second, DRAM/NAND contract prices — if fourth-quarter growth falls below 10–15 percent, the deceleration is confirmed. Third, the path of AI capital expenditure (capex) — slower hyperscaler guidance would raise memory-demand risk. Fourth, won/dollar currency moves. Fifth, shareholder-return policy — a buyback or dividend announcement could turn sentiment.

One final word in the language of the ledger. When record profit and a falling share price sit together, the market is usually proven right. So the question stands — if chip-price growth drops from 60 to 10–15 percent, and the mobile and foundry deficits step out of memory's shadow, will next quarter's headline stay as bright as this one? The ledger is open; the answer has not yet been posted.