To The Moon

What actually happened this week is three separate things colliding at once, and they need to be separated before you make any positioning decision.

What Actually Caused the Drop

The BiCS10 announcement had almost nothing to do with the selloff. The new 10th-generation 3D NAND chip launched on the same day the stock fell 14%. The market did not sell SanDisk because the product is bad. It sold because a stock up 858% year to date has no margin for error when sentiment shifts, regardless of what is on the press release.

Three things hit simultaneously.

The June jobs report printed 57,000, well below expectations, with prior months revised down 74,000 combined. Weak jobs data raises the question of whether AI capex cycles slow. That question, even when premature, is enough to trigger profit-taking in the most extended AI-adjacent names. SNDK is the most extended name in the entire S&P 500 by YTD return.

Meta Platforms announced it is building a cloud service to sell its own excess AI computing power externally. That single headline hit the entire AI infrastructure supply chain because it raises the possibility that hyperscalers may start competing with neocloud providers rather than just buying from them. If Meta sells spare compute, the demand signal from hyperscalers becomes less clean. Memory stocks traded the implication.

SK Hynix's upcoming US listing is creating genuine structural selling pressure. Traders on platforms like Stocktwits raised the specific question of whether the memory complex is being orchestrated lower to allow SK Hynix to price its US listing attractively. The practical effect is that existing holders of SNDK and MU have an incentive to rotate into the SK Hynix offering, which means selling pressure is supply-driven, not thesis-driven.

That combination, jobs data creating macro doubt, Meta creating demand doubt, SK Hynix creating supply pressure, is why the stock dropped 14% on the same day it announced a genuinely impressive product. The BiCS10 was irrelevant to all three triggers.

The Fundamental Case Is Still Intact

This is the uncomfortable part for bears, because the data genuinely does not support the thesis that the supercycle is over.

Bank of America raised its SNDK target to $2,500 on the same day the stock crashed 14%. That is not a noise trade. BofA explicitly stated it expects the NAND supply-demand imbalance to persist through calendar 2027 with pricing remaining strong through mid-2027. Citi made the identical call days earlier, also at $2,500. Bernstein is at $3,000. China Renaissance just raised to $3,169. These are not bullish holdovers from a prior regime. They are fresh calls made this week into the selloff.

Micron's own CEO said on the blowout earnings call that there is no line of sight to AI memory supply catching up with demand, with shortages persisting beyond 2027. That statement was made with full knowledge of Samsung and SK Hynix capacity ramp plans. The CEO is not blind to supply-side dynamics.

The structural case is unchanged: every AI training run requires memory. Every inference deployment requires storage. AI data center exabyte demand growth is running in the high 60s percent range. NAND supply growth runs at 15 to 17% annually. That gap does not close in one quarter.

At the current price around $1,762, SNDK trades at roughly 24 to 25x forward earnings, which given that the stock was at 60x just before the selloff represents a meaningful compression. For a company with 65% to 67% guided gross margins, 251% revenue growth, and a sold-out order book extending into 2027, that multiple is not self-evidently expensive in isolation. The problem is not the valuation in absolute terms. The problem is the beta.

The Three Risks That Are Actually New

Morningstar's Lorraine Tan said AI-linked stocks could fall 20 to 30% before becoming genuinely buyable. She pointed to two structural concerns worth taking seriously.

New supply from Samsung and SK Hynix is coming. Both companies have confirmed capacity expansions. Memory cycles turn when supply catches up faster than demand compounds. Management teams universally believe this cycle is different because of AI. Management teams in 2016 also believed that cycle was different. The structural concern is real even if the timing is uncertain.

An antitrust suit accusing Samsung, SK Hynix, and Micron of inflating DRAM prices has added regulatory risk that was not present six months ago. Pricing power that is partly structural and partly collusive has a different risk profile than pricing power that is purely structural.

The BiCS10 itself deserves a careful read. The product is impressive, offering significant advances in density and power efficiency. But it also shows that the supply side is innovating rapidly. If competitors match the BiCS10 architecture within two to three quarters, the capacity lead that underpins SanDisk's pricing power narrows. This is a long-term risk, not a 2026 risk, but it matters for anyone underwriting multi-year thesis sustainability.

The insider selling data is the most uncomfortable near-term signal. $8.9 million in insider sales in the last three months. A GuruFocus GF Score of 46 out of 100, suggesting the stock is not positioned for strong long-term performance on their proprietary metrics. Neither of these is dispositive, but both are directionally worth noting when the stock is up 858% in a year.

The SK Hynix Listing Is the Most Important Near-Term Catalyst Nobody Is Talking About

The SK Hynix US Nasdaq listing is happening within weeks. This is arguably the most important near-term variable for SNDK specifically.

If SK Hynix prices its US listing at a premium and the offering is oversubscribed, it confirms institutional appetite for memory names at current valuations and removes the overhang. Memory stocks could re-rate sharply higher as SK Hynix brings fresh institutional money into the sector.

If the SK Hynix offering struggles or prices below expectations, it signals that institutional buyers are losing conviction in the memory supercycle at current multiples. That outcome puts meaningful downward pressure on SNDK, MU, and the DRAM ETF simultaneously.

This is the specific event to watch in the next two weeks, and it is being largely underappreciated in the Tiger community commentary focused on entry points.

The Trade Framework

At $1,762 and down 14% in a single session after an 858% YTD run, SNDK is not obviously cheap and it is not obviously broken.

The glut thesis, which is the bear case, requires Samsung and SK Hynix to ramp capacity faster than AI demand compounds. That has not happened yet. The supercycle thesis, which is the bull case, requires the shortage to persist through 2027 as Micron's CEO and multiple major bank analysts predict. The data currently supports the bull case.

But here is the trade reality that the prompt is missing. A stock up 858% in six months does not need bad news to fall 30%. It just needs good news to be less than expected. SNDK is now at a stage where anything short of accelerating perfection becomes a sell-the-news event. That asymmetry is structural, not cyclical, and it is not resolved by the fundamental thesis being correct.

The SK Hynix listing in the next two weeks is the binary. Watch how it prices. Watch how SNDK and MU trade on listing day. That tells you more about institutional conviction in this trade than any analyst note written this week.

If you already hold SNDK from below $1,000, the thesis is intact and this is a shakeout, not an exit signal. If you are thinking about initiating a fresh position at $1,762, wait for the SK Hynix listing to resolve. There is no urgency to buy a stock down 14% in one session that was up 858% before the session. The shakeout may not be finished.

I am not a financial advisor. Trade wisely, Comrades.