SK Hynix's Stock Trades at Around 5 Times Next Year's Earnings Even After Revenue Grew 257%. Here's What a Multiple That Low Usually Signals About a Cyclical Business.

SK Hynix trades at ~6x forward earnings, which the article frames as a valuation discount tied to memory-chip cyclicality rather than current fundamentals. While HBM demand for AI has driven outsized growth (SK Hynix Q2 2026 revenue +257% YoY and net income +1,242% to ~94 trillion won / $66B, supported by >50% HBM share), the piece warns that as supply catches up, growth likely slows and could sharply pressure earnings (hypothetically a 90% net income drop would push the effective forward P/E from 6x to ~60x). Net: strong current AI-driven demand, but cautious investors should not assume the cycle has disappeared.

Analysis

The key market mistake is treating HBM scarcity as a permanent moat rather than a temporary pricing regime. Memory is still a supply discipline story: when capex, packaging, and yield issues normalize, the first-order effect is not just lower margins at SKHY/MU, but a much larger multiple compression because peak earnings are being capitalized as if they were durable. The current low P/E is less a bargain signal than an indication the market is already discounting a future reset.

Second-order, the real bottleneck is not demand alone but the coordination problem across DRAM, advanced packaging, and customer qualification. If Samsung and Micron continue to add capacity, the unwind can be abrupt: ASPs fall before unit volumes do, so reported revenue can look fine while gross profit collapses. That makes MU the cleaner U.S.-listed expression of the risk, while NVDA is more insulated unless memory tightness starts constraining system shipments or forcing higher BOM costs.

Consensus is likely underestimating how fast sentiment can flip once buyers sense supply is catching up. The bearish case is not a near-term break in AI demand; it is a 6-18 month earnings cliff if HBM becomes a more normal semiconductor product. The thesis is falsified if HBM lead times stay extended through the next two earnings seasons and SKHY/MU raise capex without any sign of ASP erosion.

Key Decisions for Investors

• Watchlist, not immediate action: avoid chasing MU or SKHY into strength until the next HBM pricing read-through; the better entry for a bearish view is on any post-earnings rally when management commentary turns even slightly more expansionary.
• Pair trade idea, 3-6 month horizon: short MU against long NVDA or SOXX as a relative value hedge if HBM supply signals improve; the trade works if memory margins revert faster than AI accelerator demand slows.
• Options idea, 6-12 month horizon: buy MU put spreads on the first sign of HBM ASP deceleration or capex normalization; define risk tightly because the current shortage can extend the trade against you for several quarters.
• Sector alert: if Samsung/Micron/SKHY all continue announcing HBM capacity additions while pricing remains elevated, treat that as a bearish forward indicator for memory names and a signal to trim any long exposure before the cycle inflects.
• Falsifier trigger: if next two quarters show sustained >20% sequential HBM revenue growth with no ASP pressure, the downcycle thesis is premature and bearish positioning should be cut.