This tech company is set to overtake Nvidia as the world’s most profitable

Samsung Electronics is projected to overtake Nvidia as the world’s most profitable company in 2027, with operating profit forecast at KRW488 trillion versus Nvidia’s KRW485 trillion. SK hynix is expected to rank third globally at KRW358 trillion, while combined corporate tax payments from the two Korean chipmakers could rise to KRW203 trillion in 2027. The earnings surge is expected to lift KOSPI operating profit to KRW1,044 trillion in 2027 and support FX stabilization through large dollar inflows, while KB Securities set a 2026 KOSPI target of 7,500.

Analysis

The market is pricing a simple memory-cycle story, but the more important second-order effect is balance-sheet velocity: if Samsung and SK hynix really convert earnings into taxes and capex at the implied scale, Korea gets a quasi-sovereign FX backstop from private-sector dollar generation. That should compress KRW volatility and lower the cost of capital for domestically levered cyclicals, which matters more for index re-rating than the headline profit rank.

The real winners are not just the chipmakers; they are the upstream equipment, power, industrial land, and construction ecosystems tied to Pyeongtaek/Yongin, plus Korean banks and brokers if deposit growth and equity turnover follow the earnings windfall. A strong memory cycle also tends to pull in boardroom governance reform because management teams under pressure to deploy cash and defend valuations are more willing to authorize buybacks and higher payout ratios, creating a self-reinforcing multiple expansion.

The consensus risk is assuming the cycle survives intact through 2027. Memory is notorious for margin overshoot: if supply additions from rivals or accelerated capacity restarts hit even one quarter earlier than expected, 2026–27 earnings estimates can de-rate sharply, and the market will discount that well before profits roll over. Separately, the foreign investor selling trend suggests positioning is still skeptical; if the won stabilizes and KOSPI breaks prior highs, there is room for a violent short-covering squeeze, but that also means crowded long exposure can unwind fast on any macro shock.

From a timing perspective, this is a 6–18 month theme, not a days-long trade: the entry point matters less than whether the market starts to believe the earnings durability. The cleanest tell will be whether non-memory sectors begin to participate in index leadership; if that broadening fails, the index rally is more fragile than the profit forecasts imply.

Key Decisions for Investors

• Long SK Hynix vs. short a U.S. semiconductor beta basket for 6–12 months: express the view that memory margin expansion is underappreciated while AI-led U.S. names are more exposed to multiple compression if rates stay sticky.
• Buy KOSPI exposure on pullbacks via EWY or KOSPI futures for a 3–9 month trade, but size modestly and trail stops tightly; the setup offers upside to a re-rating toward higher P/B, with downside if foreign flows remain negative.
• Long Korean financials/market infrastructure against the broader KOSPI industrial basket over 6–12 months: the FX stabilization and equity participation should support banks, brokers, and exchanges even if semis remain volatile.
• Use call spreads rather than outright longs in Samsung/SK Hynix over 12–18 months; implied upside from the earnings cycle is large, but memory cyclicality argues for defined-risk participation.
• Watch for a rotation trigger: if non-memory sectors fail to confirm earnings breadth by the next two quarters, take profits on index-level Korea longs and pivot to pair trades within Korea rather than directional exposure.