Samsung Posts Record Q3 Operating Profit on AI Chips, but Shares Fall 4% as Revenue and Net Profit Miss

Samsung Electronics posted operating profit of 107.44 trillion won, about $80.4 billion, for the July-September quarter, a company record and up 783.1% from a year earlier, per Korea JoongAng Daily. The semiconductor division accounted for nearly all of it at 106.5 trillion won, on revenue of 147.1 trillion won, according to Samsung's release. Total revenue more than doubled to a record 195.3 trillion won. Samsung said its HBM capacity for 2027 is fully booked, with supply tight through at least 2028 (JoongAng).

The market focused on what fell short. Operating profit beat the average forecast by only about 1.7%, per Seoul Economic Daily, while revenue missed the 200.3 trillion won forecast and net profit of 75.14 trillion won came in below FnGuide's 87.4 trillion won estimate (JoongAng). The chip unit's profit was also shy of expectations above 107 trillion won. Outside chips, the mobile and networks business swung to an operating loss of 600 billion won and TVs and appliances lost 200 billion won, as component costs rose.

Shares fell 4.13% to close at 232,000 won in Seoul on Thursday, per JoongAng, after briefly rising 2.89% early in the session (Seoul Economic Daily). Market watchers cited profit-taking after the KOSPI jumped more than 6% the day before on SK hynix's results, plus uncertainty over how Samsung will split planned shareholder returns. Samsung declared a 30 trillion won dividend, or 4,601 won per common share, per its release. Overnight, U.S. chip stocks slipped too: MU fell 3.4% and INTC 3.3%, with AVGO and AMD down about 2.8% each, as Samsung's net-profit miss landed alongside lighter-than-expected capex figures from Microsoft and Alphabet.

What to watch: Samsung's detail on the remaining shareholder-return program, HBM pricing into 2027, and whether the device businesses can absorb higher memory costs. The quarter confirms that AI memory now drives almost all of Samsung's profit, which also concentrates its exposure to the AI capex cycle.