SK Group Chairman Chey Tae-won, SK Hynix CEO Kwak Noh-jung, and other officials ring the opening bell at the Nasdaq MarketSite in New York to mark the commencement of trading for SK Hynix American Depositary Receipts (ADRs). SK Hynix ADRs rose by more than 13% from their offering price on the first day of trading. [Photo/Yonhap News] 

Last week, Stock Radar raised several questions: whether the sell-off in memory stocks like Micron and SanDisk was stabilizing; whether the correction in Nvidia would spread to Korean memory stocks; whether Meta's debate over excess computing power would solidify into a theory of slowing AI demand; and whether earnings expectations for Samsung Electronics and SK Hynix would hold up.

The answer this week was: "It is not a collapse in demand, but it is no longer a market that will accept any price."

Micron and SanDisk rebounded early in the week but fell 4.7% and 7.3%, respectively, on the 7th. This revealed once again the price burden on memory and storage stocks that had surged. Although they recovered some of their losses afterward, there is still insufficient evidence to conclude that the adjustment for memory stocks has completely ended.

The contagion from Nvidia was not straightforward either. On the 7th, the Philadelphia Semiconductor Index fell 4.65%, but Nvidia actually rose. This adjustment was less a case of Nvidia dragging down Korean memory stocks and more a case of the accumulated price burden across memory, storage, and equipment stocks—which had all rallied sharply—coming to light at once.

The debate over Meta’s excess computing power has not been settled as a theory of slowing AI demand, either.

While there were reports that Meta is considering a cloud business to sell its surplus AI computing capacity to third parties, the company also presented plans to double its computing capacity to 14GW by 2027 and enter into internal AI chip production. Excess computing capacity can be read as both a sign of potential over-investment and an attempt to turn existing infrastructure into a new revenue stream.

Samsung Electronics' earnings expectations were confirmed by the numbers. Samsung announced preliminary results for the second quarter with revenue of 171 trillion KRW and an operating profit of 89.4 trillion KRW. The operating profit was approximately 19 times that of the same period last year.

However, the stock price fell 6.9% on the day of the announcement. This was not because the results were poor, but because the record-breaking profit was already priced in, and the market began to question memory prices for the following quarter and the sustainability of those earnings.

SK Hynix received the exact opposite market reaction. The ADR offering price was set at $149, raising approximately $26.5 billion. The offering attracted orders more than seven times the amount issued. In its first day of trading on the Nasdaq on the 10th, it opened around $170 and closed at approximately $168, a price about 13% higher than the offering price.

The theme of this week's market flow is "Same profit, different price tags."

Looking at the same AI memory industry landscape, the Seoul market was more concerned with the highs of Samsung Electronics and SK Hynix and profit-taking after the significant rally. The New York market, however, evaluated SK Hynix as a new asset that allows direct investment in the AI memory supply chain, and paid a premium for that accessibility.

However, one should not immediately equate the U.S. premium on the first day of ADR trading with a long-term valuation increase. Exchange rates, conversion ratios, dilution from new share issuance, time differences, and arbitrage all influence the gap between the ADR and the Korean common stock. The scarcity of a new listing and initial offering demand can also inflate the price difference.

The question for this week is:

"Can the U.S. price tag for SK Hynix actually lower the discount rate applied to Korean semiconductors?"

Looking only at the first day's results, the potential has been confirmed. However, one must distinguish between initial offering demand and long-term valuation. We can only conclude that the U.S. listing has successfully lowered the discount rate if the difference between the ADR and the converted price of the Korean common stock is maintained for several weeks and foreign capital continues to flow into the Korean market.

The conclusion in one sentence is:

"While AI memory profits have been confirmed, Seoul worried about the peak, while New York paid a premium for accessibility."

One should not overstate the situation by translating the 13% rise on the first day of ADR trading into an immediate target growth rate for the Korean common stock. The success of the U.S. listing is an event that confirms demand from a new investor base, but it is not evidence that the discount rate on Korean stocks has already vanished.

The checkpoints for next week are fourfold.

First, it is necessary to check whether orders for AI accelerators and demand for advanced processes are maintained in the second-quarter earnings and capital expenditure plans to be released by TSMC of Taiwan on the 16th.

Second, we must monitor how quickly the price gap between the SK Hynix ADR and the Korean common stock narrows.

Third, we must verify if Micron and SanDisk can hold their lows from the 7th.

Fourth, we need to check if the earnings estimates for Samsung Electronics and SK Hynix remain intact following the stock price corrections.

※ Key data sources: Samsung Electronics, SK Hynix U.S. securities registration statements, Nasdaq, Philadelphia Stock Exchange, TSMC Investor Relations materials. Stock prices are based on the closing price as of July 10, 2026.

※ This material is an analysis for reference purposes summarizing market trends and does not constitute a recommendation to buy or sell any specific financial product.