Not Micron. Not Sandisk. This Memory Stock Is Quietly Gaining Share.

The article argues SK Hynix is undervalued at about 5x next year's earnings despite its leading HBM position: it controls more than half of the HBM market and produces over half of current-generation HBM4 chips. Analysts expect EPS to grow at a 107% CAGR from 2025 to 2028, while the company expands enterprise SSD production and bundles NAND products with HBM. This is a bullish investment case based on continued AI demand, not a reported earnings result or market-moving corporate announcement.

Analysis

The key question is whether HBM scarcity converts into durable pricing power or simply pulls forward capacity and earnings. A low forward multiple is not independently evidence of undervaluation in memory: if estimates capitalize peak pricing, even rapid forecast growth can leave the stock exposed to a sharp denominator reset. SK hynix’s HBM position could support better mix and customer stickiness, while bundling enterprise SSDs may help win broader infrastructure orders; neither guarantees that NAND economics improve or that customers cannot dual-source.

Near term (days to weeks), post-listing flows and any ADS-versus-home-market price normalization may dominate fundamentals. Over 1–3 months, verify reported HBM shipments, realized pricing, yields, customer qualification, inventory, and capex against the growth narrative. Over 6–18 months, additional competitor capacity or slower AI infrastructure spending could erode scarcity and revive memory cyclicality. Samsung and Micron are the direct HBM supply-side checks; a spending slowdown would also pressure NVIDIA’s deployment ecosystem. The contrarian risk is that investors treat HBM as structurally scarce while overlooking memory’s historically volatile supply response. The thesis weakens if SK hynix’s HBM mix or pricing disappoints, inventories rise, or forward EPS estimates are cut.

Key Decisions for Investors

• Prefer a staged SKHY position over chasing an IPO-related move: initiate only after checking ADS/home-market valuation and liquidity, then add on verified earnings and HBM shipment evidence. The upside case is sustained premium mix; the principal downside is estimate compression as supply catches up.
• Consider a small relative-value long SKHY / short SNDK only if current valuation and borrow costs support it. This expresses a preference for HBM exposure over a more NAND-concentrated thesis, but is not a clean hedge: both remain exposed to memory pricing and AI-capex expectations.
• Treat the cited forward multiples and EPS growth forecasts as scenario inputs, not established value. Before sizing, compare current consensus revisions with company-reported HBM pricing, output, yields, inventory, and capex; missing evidence argues for a watchlist rather than a full position.
• Falsify or reduce the thesis on downward guidance or consensus EPS revisions, rising memory inventories, weaker HBM qualification/shipments, or evidence that customer spending is slowing. Reassess the relative trade if SNDK’s earnings outlook improves materially or SKHY’s HBM advantage narrows.