Sign in
Shared read

Ask Tessara

A research answer shared from Ask Tessara. Every answer carries its receipts.
The Read
Asked Jul 15, 2026· on the record
what's going on with SK Hynix. insane volatility. and is this a dip to buy?
CONFIDENCEMODERATEevidence 7/104 constraint reads
SK Hynix's thesis is intact: every node it sells into tightened again today, and at 4x forward it prices a cyclical peak the disclosed supply structure does not support.
  • HBM3e holds at 80/100, Korea customs HBM-tier export unit price well above its 12-month trend; SK Hynix total memory revenue running well above trend, the dominant pull.HBM3e
  • Korea customs DRAM export unit price runs well above trend and DRAM lead-times continue tightening, consistent with a supply-constrained market [DDR5 reads 86/100, tightening]; DRAM revenue is the strongest single pull on the bundle.DRAM DDR5
  • SK Hynix is down 9% over 30 days, up 161% over 180 days, and holds at about 4x forward earnings, net-cash balance sheet with interest coverage near 93x.NAND Flash

The volatility is a positioning unwind, not a thesis break. Every constraint SK Hynix sells into tightened again on today's read: holds at 80, DDR5 at 86 and still tightening, at 80 and tightening. The Korea customs HBM-tier export unit price runs well above its 12-month trend and SK Hynix's own memory revenue reads well above trend, the dominant supply-side confirmation. Nothing in the data has moved against the bull case.

The entry math is the interesting part. The stock is down 9% over 30 days but up 161% over 180 days, sitting at about 68% of its 52-week range. Forward P/E reads about 4x on the vendor estimate, which is cyclical-peak pricing: the market is treating today's HBM margins as mean-reverting. The balance sheet is net-cash with interest coverage near 93x, so financial fragility is not the source of volatility.

The Memory Desk (held 8 days, not re-affirmed this week, medium confidence, 2 confirmed of 6 resolved) leans SK Hynix as sole Tier-1 HBM supplier to the leading AI accelerator customer through the foreseeable horizon. That track record is weak empirically, so treat it as directional context, not a graded signal. The physical data does the heavy lifting here.

The dip case rests on one question: is 4x forward cheap because the HBM earnings cycle is structurally longer than the multiple implies, or is it fair value because ASPs roll over in the coming years as Samsung qualifies? The data today says the gate is intact and pricing is holding; the forward ASP trajectory is what the multiple has not settled.

generated 3d ago

Tessara is independent research and analysis, not investment advice. We are not a registered investment adviser. Past performance is not indicative of future results, and you are responsible for your own investment decisions.