KOSPI Drops 7.6% in One Week, Samsung Valuation Hits 24-Year Low
Taylor Wilson
KOSPI fell 7.6% this week, pushing its forward P/E to 6.17x — the lowest since 2004 — yet forward earnings were revised *up* 1.4%. Record profits and record-low valuations now coexist, creating the most extreme dislocation in the Korean market.
Profits are rising, prices are falling — who's wrong?
KOSPI's 12-month forward P/E dropped to 6.17x, below the 2008 financial-crisis trough and 3.1 standard deviations beneath the long-term average.
Yet forward EPS was revised up 1.4% over the same period. This means → the market is not pricing in an earnings collapse; it is pricing in "even if profits are good, I don't want to hold."
In plain terms = companies are earning more and more, but investors are running faster and faster — the divergence has been pushed to an extreme.
Samsung at 4.81x — bargain or trap?
Samsung Electronics' forward P/E fell to 4.81x, the lowest since 2000. At the same time, Goldman Sachs raised its 2027 and 2028 EPS estimates by 3.7% and 4.4%, driven by stronger DRAM pricing.
This means → record profits and an all-time-low valuation coexist in a single stock. Samsung alone encapsulates the Korean market's central contradiction.
The July 30 earnings call is the key verification point. Investors need to hear long-term contract customer counts, pricing mechanisms, and HBM4 — the next-generation high-bandwidth memory chip — revenue, yields, and customer progress. If high prices lock into 2027 cash flow, 4.81x looks like a mis-pricing; if capacity ramps too fast or downstream demand weakens, the low multiple persists.
Are foreign sellers slowing down?
Foreign investors net-sold ₩4.12 trillion this week, down sharply from ₩19.84 trillion the prior week. The won appreciated 1.6% in tandem, from around 1,531 to 1,505 per dollar.
But tech-sector foreign net selling hit ₩5.06 trillion — exceeding total index outflows. This means → foreign buying in banks, telecoms, and chemicals offset part of the tech selling. Systemic withdrawal pressure is easing, but tech-position liquidation is not finished.
Year-to-date, foreigners have net-sold ₩161.58 trillion; domestic retail has net-bought ₩109.45 trillion. Retail is absorbing the foreign selling alone, making the market more sensitive to external shocks.
How much pessimism does 6.17x imply?
Goldman's stress test: cut consensus forward EPS of 1,175 by 33% to 787, then apply the historical trough P/E median of 11.4x — the result is roughly 8,936 for KOSPI, still above the current level.
In plain terms = even if profits really fell by a third and you used the most bearish multiple in history, the index would still be higher than today. This offers an odds reference, not a timing signal for buying.
Strip out Samsung and SK Hynix, and Korea's forward P/E is 10.1x. This reflects that the extreme index-level undervaluation comes mainly from the collision of semiconductor heavyweights' profit surge with a cyclical discount — not a market-wide cheapness.
Where is money rotating?
Banks and telecoms rose against the tide this week; tech and shipbuilding fell nearly 10%. Securities-sector earnings upgrades led, chemicals and insurance stayed positive, machinery and leisure turned negative.
This reflects a shift: growth is broadening from memory-chip-only to financials, materials, and parts of the industrial chain. Index-level earnings have not broken down, but the structure is changing.
On technicals, margin-loan ratios remain elevated, the VKOSPI volatility index rose to roughly 90, and the share of KOSPI constituents above their 200-day moving average fell to about 20%. In plain terms = prices are in oversold territory, but leverage and volatility have not returned to healthy ranges.
What to watch next?
Six signals will define the next move: KOSPI EPS trajectory, tech-sector foreign flows, Samsung's July 30 call, DRAM pricing, the won's direction, and the share of constituents above the 200-day MA.
Earnings keep rising + foreign tech selling converges → 6.17x looks more like a mis-pricing.
Earnings turn negative + prices keep falling → the low P/E will be read as a cyclical-peak discount (the market believes profits have peaked), not a bargain.
Content is for reference only, not financial advice.