Barclays believes SK Hynix's U.S. shares can double. Here's why

Barclays initiated coverage on SK Hynix with an overweight rating and a $330 price target versus ~$167, implying potential upside of roughly 97% over the next year. The key catalyst is ongoing memory shortages that support price increases, with upside to gross margins and a larger gap to consensus from materially higher 2027 revenues driven by HBM pricing. Barclays also highlighted capital-return optionality, estimating cash equivalents could exceed 40% of market cap by year-end 2027, enabling share buybacks.

Analysis

The market is starting to price SKHY less like a cyclical DRAM name and more like an AI bottleneck royalty stream. If HBM remains the constraint, incremental unit demand is not the main driver; pricing is, which means upside can outpace volume even if end-demand cools modestly. That also argues for a multiple re-rate versus generic memory exposure: the scarce asset is not bits, it is qualified HBM supply tied to a small set of customers.

The second-order beneficiaries are the adjacent capex and packaging chain, not just memory peers. If SKHY keeps taking pricing, competition will respond by expanding HBM capacity, which should pull through advanced packaging, test, and deposition tools over 6-18 months; that favors equipment names more than broad memory baskets. By contrast, a sector fund like DRAM is vulnerable to mean reversion because it carries the whole commodity complex, including names that may not convert pricing power into buybacks as efficiently.

The key risk is that the current rerating already embeds a long runway of tight supply, while memory is still one of the fastest sectors for supply response. A 1-3 month reversal would likely come from a flat-to-down HBM pricing datapoint, a capex surprise from Samsung/Micron, or weaker AI server order commentary that signals inventory digestion. Over 6-18 months, the thesis breaks if competitors close the qualification gap and the shortage normalizes faster than 2027 earnings estimates imply.

Key Decisions for Investors

• Buy SKHY on any 5-10% post-coverage pullback; use a 6-12 month horizon and underwrite to continued HBM price strength. Falsify if HBM pricing/lead times stop tightening or if management guides capex materially above expected maintenance levels.
• Pair trade: long SKHY / short DRAM ETF for 1-3 months to isolate idiosyncratic HBM share gain versus broad memory beta. Best if sector sentiment stays hot but dispersion widens; exit if DRAM regains momentum on a memory spot-price spike.
• For higher convexity, use bullish call spreads in SKHY rather than outright stock into the next earnings cycle. This captures rerating potential while limiting drawdown if the market begins to discount future supply normalization.
• Watch Micron MU and Samsung memory commentary as the earliest falsifier. If either signals faster-than-expected HBM capacity expansion, reduce SKHY exposure and rotate into equipment names (AMAT, LRCX, KLAC) that benefit from the capex response.