SK Hynix is not trading at seven times earnings, whatever the screen says. The Korean memory maker closed at ₩1,678,000 on 25 August 2026, up 568.5% over twelve months and still 43.8% below its 52-week high of ₩2,987,000. On headline numbers that puts it on a price/earnings ratio of 7.4x, which for a company whose net income grew 461% looks like one of the great mispricings in global equities. It is not, and the reason is visible in a single line of the income statement: SK Hynix’s pre-tax margin is 109.3%. A company cannot earn more before tax than it collects in revenue from operations. The difference has to come from somewhere else.
That somewhere else is worth ₩78.05 trillion. Trailing operating income (EBIT) is ₩128.71tn on revenue of ₩189.17tn — a genuinely spectacular 68.0% operating margin. Reported pre-tax income is ₩206.76tn. The ₩78tn gap, equal to 41.3% of revenue, arrives below the operating line as non-operating income, and it is what carries reported net income to ₩161.97tn and the headline EPS to ₩227,588. Strip it out and tax operating profit at Korea’s roughly 24% corporate rate, and normalised earnings fall to about ₩134,000 a share. At ₩1,678,000 that is a P/E near 12.5x, not 7.4x. Still cheap for this growth rate. But 12.5x is an argument, whereas 7.4x is a screening error waiting to disappoint whoever buys it.
Key facts
- SK Hynix (KRX: 000660): ₩1,678,000, up 0.42% at the close of 25 August 2026 — StockAnalysis
- Up 568.5% over one year, but 43.8% below the 52-week high of ₩2,987,000; the 52-week low is ₩253,000 — StockAnalysis, 25 August 2026
- Pre-tax margin 109.3%, profit margin 85.6%, operating margin 68.0% — StockAnalysis trailing twelve months to 30 June 2026
- Reported net income ₩161.97tn exceeds EBIT of ₩128.71tn, implying roughly ₩78.05tn of non-operating income — derived from reported margins
- Market capitalisation ₩1,217.93tn on revenue of ₩189.17tn (+145.0%) — StockAnalysis, 25 August 2026
- ₩40 trillion (about $28.6bn) buyback announced 19 August 2026, cancelling roughly 24.07m shares, or 3.3% of the 730.5m issued — FinanceFeeds
- Net cash of about ₩69 trillion at the end of the second quarter — FinanceFeeds, 19 August 2026
- Analyst consensus: Strong Buy, ₩3,170,279 target, implying 88.9% upside; beta 2.41 — StockAnalysis, 25 August 2026
The earnings-quality problem in one chart
The clearest way to see what is happening is to lay the income statement out end to end and draw a line at revenue.
Everything down to operating income behaves exactly as you would expect from a memory manufacturer in the best pricing environment the industry has seen: 76.3% gross margin, 68.0% operating margin. Then the pre-tax bar crosses the revenue line, which is not something a normal manufacturing business does.
To be clear about what this is and is not: non-operating income of this size is not evidence of wrongdoing, and it is disclosed. In a Korean chaebol structure it typically reflects some combination of equity-method gains from affiliates, foreign-exchange effects on a large dollar-denominated receivables book, and mark-to-market movements on investments. Those are real gains. They are simply not the memory business, they do not recur on a schedule, and several of them reverse when the won or asset prices move the other way.
The practical consequence is that every valuation multiple built on reported EPS is flattered, including the 4.0x forward P/E the screens display. Anyone anchoring a bull case to “it trades at four times earnings” is anchoring to a number that includes ₩78tn of income the memory cycle did not generate. FinanceFeeds has seen this pattern distort headline numbers elsewhere in the sector — the same below-the-line effect is why some chipmakers’ reported losses and profits have swung far more violently than their operations.
The bull case: a real cycle, real cash, and a record buyback
None of the above makes the bull case weak. It makes it narrower and more specific.
The operating business is extraordinary on its own terms. A 68.0% operating margin on ₩189.17tn of revenue means SK Hynix is converting the high-bandwidth memory shortage into profit at a rate almost no manufacturer of physical goods ever achieves. Revenue grew 145.0%. Return on equity is 92.68%. This is what a genuine supply bottleneck looks like when it lands on a company with the capacity to serve it.
The balance sheet backs it up. SK Hynix reported roughly ₩69 trillion of net cash at the end of the second quarter, and on 19 August it announced a ₩40 trillion buyback — about $28.6bn — that the company described as the largest treasury-share cancellation ever undertaken by a listed South Korean company. That programme retires around 3.3% of shares outstanding. Critically, it came after the stock fell 9.8% in a session, which is management buying its own equity into weakness rather than at the highs.
Korean capital returns are also broadening. FinanceFeeds reported this week that Samsung paid out 110 trillion won while SK Hynix bought back its own stock — two of the country’s largest companies returning capital simultaneously, which is a meaningful shift for a market long discounted for poor shareholder treatment.
Access has improved too. The company now has a US listing route: FinanceFeeds covered the arrival of 2x long and 1x short SK Hynix ETFs as its ADRs began trading. That widens the buyer base beyond Seoul, and it is one reason the shares now move on US session news as well as Korean.
The bear case: cyclicality, China, and a 2.41 beta
Memory is the most violently cyclical business in semiconductors, and the numbers above are what a peak looks like, not a plateau.
The mechanism is well understood: extraordinary margins attract capacity, capacity arrives with a two-to-three year lag, and pricing collapses when it does. A 68% operating margin is not an equilibrium — it is a signal to every competitor and every government with an industrial policy. The relevant question is not whether margins normalise but when, and from what level.
China is the specific vector. FinanceFeeds reported that CXMT’s 466% Shanghai debut hit Micron and SK Hynix, though its prospectus contains no HBM. That detail cuts both ways and deserves precision rather than a headline. CXMT arriving with capital and no high-bandwidth memory product is not an immediate threat to SK Hynix’s HBM franchise — but it is a direct threat in commodity DRAM, which still underpins a large share of volume. Competition does not have to reach the crown jewels to compress the average.
The market has already shown how fast this repricing happens. In late July the stock fell 14.65% in the AI chip selloff that also hit AMD. With a beta of 2.41, SK Hynix moves roughly two and a half times the market, and the drawdown from ₩2,987,000 to today’s ₩1,678,000 happened without any deterioration in reported earnings at all. A stock that can lose 44% while its profits are still rising can lose considerably more when they stop.
The concentration risk this creates for Korea itself is not theoretical. In late July, KOSPI trading was halted after an 8% plunge as the SK Hynix ADR fell below $140. A single stock large enough to trip a national index circuit breaker is a stock whose own drawdowns become self-reinforcing: index funds sell the market, the market sells the largest constituent, and the largest constituent is the company you own. At a ₩1,218tn market capitalisation, SK Hynix is not a position that can be exited quietly when sentiment turns, and that illiquidity premium is part of what the current multiple is compensating for.
There is a second-order risk in the non-operating income itself. If a meaningful portion is foreign-exchange or mark-to-market driven, it is not merely non-recurring — it is potentially reversing. A stronger won or a drawdown in affiliate valuations would turn a tailwind into a headwind below the operating line, and reported EPS would fall faster than operating profit. That is the scenario in which a “4x forward earnings” stock re-rates upward on the multiple while the price goes down.
The two cases, priced
Here is what each level implies, using both the headline EPS of ₩227,588 and the normalised figure of roughly ₩134,000 derived from taxing EBIT at 24%.
| Scenario | Price (KRW) | Move from ₩1,678,000 | Headline P/E | Normalised P/E |
|---|---|---|---|---|
| 52-week high | ₩2,987,000 | +78.0% | 13.1x | 22.3x |
| Analyst consensus target | ₩3,170,279 | +88.9% | 13.9x | 23.6x |
| Bull case | ₩3,000,000 | +78.8% | 13.2x | 22.4x |
| Today | ₩1,678,000 | — | 7.4x | 12.5x |
| Bear case | ₩1,000,000 | −40.4% | 4.4x | 7.5x |
| 52-week low | ₩253,000 | −84.9% | 1.1x | 1.9x |
The normalised column is the one that should govern decisions. On it, today’s 12.5x is a reasonable but unremarkable multiple for a peak-cycle semiconductor business. The bull case at ₩3,000,000 requires paying 22.4x normalised earnings at what may prove to be the top of the memory cycle — which is precisely the multiple investors were paying in January before the shares halved. The bear case at ₩1,000,000 would put the stock on 7.5x normalised, a level that historically appears when the market has concluded the cycle has turned.
Note how differently this reads from the AI infrastructure names on the other side of the trade. FinanceFeeds’ Oracle bull-versus-bear analysis covers a company spending 174% of operating cash flow on capital expenditure to serve AI demand. SK Hynix is the counterparty receiving a large share of that spending. The two stocks are opposite ends of the same capital flow, which is why they should never be sized as independent positions in a portfolio.
What happens next
One: the disclosure of what sits in non-operating income is the most important thing to watch. If the next filing shows the ₩78tn is largely equity-method gains from operating affiliates, the earnings are more durable than this article assumes and the bull case strengthens materially. If it is predominantly foreign exchange and mark-to-market, expect reported EPS to be far more volatile than operating profit in both directions. Read the notes, not the headline.
Two: the buyback provides a floor, but a known one. A ₩40tn programme against a ₩1,218tn market capitalisation is roughly 3.3% of the company, and management has shown it will step in after sharp falls. That supports the price near current levels. It does not stop a cycle turn — buybacks slow declines, they do not reverse them, and a company retiring stock at 22x normalised earnings on the way up destroys value as efficiently as it creates it on the way down.
Three: the bear case arrives through DRAM pricing, not HBM. The HBM franchise is defensible for now, and CXMT has no HBM product in its prospectus. Commodity DRAM is where new capacity lands first. Watch contract DRAM pricing and inventory days rather than HBM headlines, because that is where a 68% operating margin starts compressing, and margin compression from this altitude does not need to go far to halve the earnings.
The framing for anyone sizing a position: SK Hynix is a genuinely exceptional operating business at what is plausibly a cyclical peak, trading on a multiple that looks half as expensive as it actually is. That combination rewards discipline about the entry price and punishes anyone who bought because a screen said seven times earnings. This article is analysis, not investment advice.
Frequently asked questions
What is the SK Hynix stock prediction for 2026?
The bull case is ₩3,000,000, roughly the 52-week high and close to the analyst consensus target of ₩3,170,279, implying 78.8% upside. The bear case is ₩1,000,000, a 40.4% decline. SK Hynix closed at ₩1,678,000 on 25 August 2026, up 568.5% over a year but still 43.8% below its high.
Why is SK Hynix’s P/E ratio so low?
Because reported earnings include roughly ₩78tn of non-operating income. SK Hynix’s pre-tax margin is 109.3% — higher than its revenue — and reported net income of ₩161.97tn exceeds operating income of ₩128.71tn. Taxing operating profit at about 24% gives normalised earnings near ₩134,000 a share, putting the stock closer to 12.5x than the headline 7.4x.
How big is the SK Hynix buyback?
₩40 trillion, about $28.6bn, announced on 19 August 2026 after the shares fell 9.8%. It covers roughly 24.07m shares, about 3.3% of the 730.5m issued, and the company described it as the largest treasury-share cancellation ever undertaken by a listed South Korean company. SK Hynix also held about ₩69tn of net cash at the end of the second quarter.
Is CXMT a threat to SK Hynix?
In commodity DRAM, yes over time; in high-bandwidth memory, not immediately. CXMT’s Shanghai listing prospectus contains no HBM product, so SK Hynix’s most profitable franchise is not directly contested yet. But new Chinese capacity lands in commodity DRAM first, and that still represents a large share of industry volume, so it compresses the average selling price rather than the crown jewel.
Can I buy SK Hynix shares outside Korea?
Yes. SK Hynix ADRs began trading in the US in 2026, and leveraged products followed, including 2x long (SKHX) and 1x short (SKHZ) ETFs. Note that the shares carry a beta of 2.41, so leveraged wrappers on an already high-volatility stock compound risk substantially.
What would invalidate the bull case for SK Hynix?
Falling contract DRAM prices and rising inventory days would signal the cycle turning, which matters more than any HBM headline. A reversal in the non-operating income line — through a stronger won or falling affiliate valuations — would also cut reported EPS faster than operating profit falls, making the stock look more expensive precisely as the price declines.