Goldman Sachs: Korean AI Stocks Oversold, Maintains KOSPI Target at 12,000
Miles Bennett
Goldman Sachs reaffirmed its 12,000 twelve-month target for the KOSPI, implying roughly 80% upside, arguing July's 22% single-month plunge overshot fundamentals and that the memory-chip super-cycle is far from priced in.
80% upside — what is Goldman's case?
Goldman maintains a KOSPI target of 12,000, implying about 80% upside from current levels.
Core thesis: accelerating AI compute demand + severe memory supply shortage (potentially lasting to 2030) → chip prices and earnings will hold up far longer than the market expects.
This means → Goldman believes the market is pricing this cycle like an ordinary one, when its scale and duration could be historically unprecedented.
A 22% crash in July — what happened?
The KOSPI more than doubled in the first half of the year, then plunged 22% in July alone — one of the steepest monthly drops in recent history.
Goldman compared the sell-off to major corrections over the past two decades: the 2021 tech pullback, the Covid crash, and the dot-com bust.
In plain terms = the rally ran too hot on too much leverage; once sentiment turned, the unwind overshot fundamentals by a wide margin.
How much leverage has been flushed out?
Korean leveraged-ETF assets fell from a June peak of $53 billion to $25 billion — a 54% drop.
Retail margin balances declined from $25 billion to $19 billion; regulators simultaneously tightened leverage-product rules.
This means → Goldman judges positioning is now "much cleaner" — speculative capital from the prior leg has largely exited, giving any rebound a firmer base.
Big Tech capex worries — how does Goldman respond?
The market worries that Big Tech capex is too large, funding is stretched, and competition is intensifying — Goldman concedes these concerns have merit.
But it argues they are not enough to overturn the core call: this AI hardware cycle is "unusually large and unusually long."
This reflects Goldman's stance: near-term noise does not change the medium-term trend — the current moment is an expectations-gap repair window, not the start of a cycle reversal.
Beyond chips — what else does Goldman like in Korea?
Outside memory chips, Goldman favors Korean power infrastructure, industrial automation, defense, shipbuilding, robotics, and energy.
Samsung Electronics and SK Hynix together account for roughly half of the KOSPI's weighting; both rose sharply on Wednesday — 4% and 6% respectively.
In plain terms = Goldman's bull case is not a single-sector bet — it sees the entire Korean market as undervalued, with chips as the main thread but not the only one.
Content is for reference only, not financial advice.