Executive Summary:
SanDisk NASDAQ:SNDK closed at $1,015.89 on July 29, down 54% from its June peak near $2,354, after touching an all-time-high RSI reading of 99.22
The stock is now the most oversold it's been in 15 months, with 14-day RSI sitting near 29 to 32
Despite the collapse, SNDK remains up roughly 362% year to date, one of the best-performing stocks in the S&P 500
The trigger came from Asia: South Korea's SK Hynix missed quarterly forecasts and the KOSPI fell nearly 11% on July 28, its eighth circuit breaker of 2026
SanDisk reports fiscal Q4 earnings on August 5, with the stock currently testing the psychologically important $1,000 support level
Few stocks in 2026 have offered a cleaner lesson in how fast a crowded trade can reverse than SanDisk. This is a company that traded near $100 at the start of the year and touched $2,354 in June on AI-driven NAND flash demand, then lost more than half its value in a single month while the story behind it, structurally tight memory supply, barely changed at all.
This isn't the first scare this stock has had this year, either. SNDK has been repeatedly whipsawed since its February spinoff from Western Digital, with earlier selloffs in early and mid-July already testing the $1,500 and $1,300 support zones before this latest leg down broke through $1,300 entirely and now has price testing $1,000. Each prior support break has come faster and sharper than the one before it, a pattern technicians read as accelerating distribution rather than a series of unrelated dips.
SNDK daily chart: price collapsing from the $2,354 June peak to $1,015.89, RSI near 29-32 after touching 99.22 at the top.
The Mechanics: From the Most Overbought Reading in History to Deeply Oversold
The scale of this reversal is what makes it remarkable. SanDisk's RSI hit 99.22 at its June peak, described as the most overbought reading in history for a large-cap stock. That kind of extreme almost never resolves gently, and it hasn't here: the 14-day RSI has now collapsed to the 29 to 32 range, deeply oversold territory, in the span of about a month.
Price action has broken every meaningful support level along the way. The stock lost the $1,500 zone in mid-July, then $1,300 shortly after, then fell another 12% to test $1,000, the next major technical area, with the 100-day moving average, which had acted as support throughout the rally, breaking cleanly on the way down. That's not a single bad session, it's a structural breakdown across multiple support levels in rapid succession.
Market Impact: A Made-in-Asia Selloff Hitting a Made-in-America Rally
The proximate trigger for the latest leg down wasn't anything SanDisk itself announced. SK Hynix, the South Korean memory giant, posted quarterly results that missed forecasts, reviving fears that AI-driven memory demand is cooling faster than the market had priced in. The KOSPI's nearly 11% single-day drop on July 28, its eighth circuit breaker of the year, sent that fear rippling through every memory name globally, Micron and Western Digital included, with SanDisk absorbing some of the sharpest damage given how far and fast it had run.
Layered on top of the sector-specific fear is a broader rotation. Federal Reserve Chair Kevin Warsh's recent comments flagging stretched asset valuations, combined with rising Chinese competition from CXMT, now the world's fourth-largest DRAM producer following its blockbuster Shanghai IPO, have given investors two separate reasons to question whether SanDisk's valuation ever reflected sustainable pricing power or simply momentum chasing a hot theme.
Competing Interpretations: Healthy Reset or Bubble Deflating
The bull case treats this as exactly what extreme moves like this eventually do: mean-revert hard, without the underlying business actually deteriorating. SanDisk's fundamentals remain genuinely strong, record NAND pricing, a multi-decade manufacturing agreement with Kioxia, and multi-year customer commitments from hyperscalers that haven't been canceled or renegotiated. A stock still up 362% for the year, even after this crash, has given back speculative excess rather than real value, and RSI this deeply oversold has historically preceded sharp relief rallies in high-beta names.
The bear case is that a 99.22 RSI reading was never sustainable in the first place, and the speed of the unwind, four separate support breaks in under a month, suggests the market is repricing NAND economics more broadly rather than simply blowing off froth. SK Hynix's miss and CXMT's emergence as a serious low-cost competitor both point to the same underlying risk: memory pricing power, the entire basis for SanDisk's 2026 re-rating, may be less durable than the rally assumed, and August 5 earnings will be the first real test of whether customer commitments are holding at the prices that justified this year's numbers.
Risk Implications
Buying SanDisk into this level of oversold conditions ahead of an earnings report carries obvious binary risk: a beat that reaffirms NAND pricing strength could spark a violent short-term reversal given how stretched the selling has become, while a miss or cautious guidance amid the SK Hynix-driven fear could open a clean path toward the next support levels below $1,000. Waiting for the report removes that binary risk but also means potentially missing the sharpest part of any relief bounce, which tend to happen fast in stocks that fell this hard this quickly.
For anyone still holding from the June highs, the risk calculus is different: a stock still up 362% year to date, even after a 54% drawdown, has a very different break-even psychology than one bought near the top, and that gap in cost basis across the shareholder base is part of why the selling has been this intense.
Conclusion
SanDisk's chart currently holds two true numbers that point in opposite directions: down 54% from its peak, and up 362% for the year. Which one matters more depends entirely on when an investor got in, and August 5 earnings will decide whether the 54% figure was the market correcting an unsustainable extreme, or the beginning of a real repricing of the entire NAND flash story that took this stock to $2,354 in the first place.
Does 362% year-to-date change how you read this crash, or is a 54% drawdown a red flag regardless of where the stock started the year? Let me know how you're thinking about this one.