SemiAnalysis Bullish on SK Hynix, Forecasts Q2 DRAM Profit of KRW 55 Trillion
Alina Collins
SemiAnalysis has pushed back against Korean broker KIS's downgrade, forecasting SK Hynix's Q2 DRAM ASP will rise ~45% quarter-on-quarter with project-level operating profit hitting KRW 55 trillion — framing the sell-off as a buying opportunity.
What triggered this sell-off?
On July 13, Korean broker KIS projected SK Hynix's Q2 operating profit at KRW 60.4 trillion — roughly 8% below the consensus estimate of KRW 65 trillion.
SK Hynix shares fell more than 10%, breaking below KRW 2 million and pulling back 33% from their June 25 all-time high.
This means → a single broker's downgrade wiped out a third of the stock's value — a sign of extreme sentiment sensitivity around memory names.
Why did KIS cut its forecast — is something wrong with fundamentals?
KIS explained that SK Hynix earns a large share of revenue from HBM — high-bandwidth memory, the premium DRAM built for AI chips — and HBM prices are locked in by long-term agreements (LTAs) that cannot ratchet up with the spot market short-term.
KIS forecast Q2 DRAM ASP up ~30% QoQ and NAND up ~50%, but the blended gain was dragged down by fixed HBM contract prices.
In plain terms = KIS is not bearish on SK Hynix. It simply priced in the constraints of contracts already signed, which naturally produces a lower profit number than a pure spot-price model. KIS kept its target price at KRW 3.8 million and its overweight rating.
What is SemiAnalysis's counter-argument?
Analyst Ray Wang forecast Q2 blended DRAM ASP up ~45% QoQ, with DRAM project-level operating profit at ~KRW 55 trillion.
His core logic: DRAM spot prices have surged ~60% QoQ — enough to lift overall earnings sharply. HBM contract prices moved only low-single-digits quarter-on-quarter, a "stable range" that is not a material drag.
This means → the real disagreement is about methodology — contract-price-weighted vs. spot-price-weighted. SemiAnalysis believes the spot market's upside is being underestimated.
Are the two firms truly at odds?
Not entirely. KIS itself projects Q2 operating margin at 74.6%, a record high — that alone is a powerfully bullish fundamental signal.
Both firms converge on the medium-term view: the memory industry is shifting from "how much did ASP rise this quarter" to "how long can margins stay this high," as three-to-five-year LTA structures reshape valuation logic.
This reflects a deeper question: the real suspense is not whether Q2 is strong, but how many quarters this profit level can last. SK Hynix's actual Q2 results will be the decisive test of where the two forecasts diverge.
Content is for reference only, not financial advice.