Sandisk Falls 22% as AI Memory Stocks Sell Off

Sandisk shares closed down 22% on Thursday, November 5, 2026, in a steep selloff across AI-linked stocks, according to MarketWatch. The same report put the decline in Nvidia (NVDA) at 9% for the session.

Micron Technology (MU) closed down at least 16% that day, according to a Morningstar syndication of the MarketWatch report. Reuters reported that Samsung Electronics and SK Hynix each fell more than 10%. The pattern is a sharper decline in memory names than in Nvidia, the graphics processor designer whose shares fell by less than a tenth.

For scale, a 22% fall requires a gain of about 28% to get back to even, and a 16% fall requires about 19%. Those are the break-even figures for anyone holding through the drop.

The run-up beforehand was large. A Yahoo Finance report dated August 28 said Sandisk was the best-performing stock in 2026 to that point, with a gain of nearly 600%, and that Micron was fourth-best, up nearly 250%.

In my view, that run-up is why a single-session drop of this size carries weight. Much of the year's gain sat in a small group of memory names, so the group is highly sensitive to shifts in sentiment.

Analyst targets were also in place before the selloff. In mid-October, 247wallst reported consensus price targets of $328.72 for Nvidia, implying about 45% upside; $1,555.13 for Micron, implying about 56% upside; and $2,173 for Sandisk, implying about 46% upside. A consensus target is the average of sell-side analysts' forecasts, meaning those employed by brokerages, and implied upside is measured against the share price at the time of the report. In September, Investor's Business Daily had reported that Micron, Sandisk and SK Hynix were in or near buy zones, a chart-based term for price areas where traders often look to enter.

One caveat on those upside figures: they were calibrated to mid-October prices. The November drops would have changed the arithmetic on each target without changing the targets themselves.

MarketWatch framed the decline around investor discomfort with uncertainty, and its summary noted that bargain hunters saw a buying opportunity in the drop. Reuters followed with a headline, "Wall Street stocks rally after heavy AI-related selloff," which points to a rebound in broader US equities after the decline.

In my view, the more useful question is which part of the November move reflects the sector and which reflects positioning, meaning how heavily investors had bet on these names. The data cited here cannot separate fundamental revisions from the unwinding of crowded trades. Memory pricing and AI capital spending are the variables to watch, and neither is quantified in the reporting above.

For ordinary savers, exposure is often indirect, through index funds and sector funds that hold these names at weights set by market capitalization, the total stock market value of each company. A selloff concentrated in a few large holdings moves those funds more than the headline index. That is how such products are built, and it applies whether the November decline turns out to be a pause or the start of a longer correction.

To be clear, nothing here is a view on whether the shares are cheap or expensive. What the record shows is a sharp, one-session repricing of the group that led the year, analyst targets set weeks earlier, and a rebound headline the following day.