SanDisk stock prediction: Friday closed at $1,299.30, bull…

SanDisk’s slide is not a NAND price break. The shares closed Friday, 16 October 2026, at $1,299.30, down $95.53, or 6.85 percent, from Thursday’s $1,394.83. Seven straight lower closes, starting 8 October, leave the stock $393.12, or 23.23 percent, under the 7 October close of $1,692.42. The last 90 daily log-returns, carried across the 52 New York sessions from 19 October through 31 December, give a bull of $2,532.60, a base of $1,133.12 and a bear of $662.86.

The missing fact is the product under the headline. CXMT’s year-end claim, as reported from its president Cao Kanyu, is a denser DRAM cell in a server DDR5 module, not a NAND contract cut. Rich Smith wrote on 15 October that SanDisk does not make DRAM at all. Friday’s log-return was about 0.95 of one sample standard deviation on the last 90 sessions: a large day inside the recent range, not a break in it.

What just happened, and why the obvious reading is wrong

The obvious reading is that a Chinese memory rival cut the NAND price and the stock admitted it. From the 7 October close of $1,692.42 the next seven simple declines were 4.90, 1.72, 5.97, 0.34, 5.33, 0.61 and 6.85 percent.

The sample standard deviation of the last 90 daily log-returns is 0.07435, about 7.4 percent. Friday’s log-return, the natural log of $1,299.30 over $1,394.83, is -0.07095, or 0.95 of that deviation. The unusual part is the streak: 7 October, at $1,692.42, had itself been an up day from $1,660.46.

The technology item that travelled with the selling is the wrong product. Tom’s Hardware reported that CXMT president Cao Kanyu said the first 4F-squared product, a server DDR5 module, is due by year-end. The Yicai account says the cell is not yet commercialized, that vertical transistors cut the footprint by about one-third on paper, and that no speeds or capacities were given.

Rich Smith, a contributing Motley Fool analyst, wrote on 15 October: “Sandisk doesn’t actually make DRAM at all.” The same paragraph says SanDisk manufactures NAND flash instead. He cited TrendForce DRAM shares of about 9.5 percent for CXMT, 39.4 percent for Samsung, 24.9 percent for SK Hynix and 23.3 percent for Micron. SanDisk is not on that list.

His Friday piece still tied the selling to fear of CXMT and put the week at a 15 percent loss while the session was open. The saved closes, measured from 7 October, are down 23.23 percent.

Who is exposed

Equity holders are first. Nasdaq’s saved summary puts market cap at $188,824,257,223. Divided by Friday’s $1,299.30, that is about 145.3 million shares, against 149 million outstanding on 3 July and about 155 million diluted shares in the Q1 guide. No October filing fetched here closes that gap, and the September insider sales were plan filings, not this tape.

Cash was $4.762 billion at 3 July and long-term debt was zero, down from $1.829 billion a year earlier. Fourth-quarter repurchases were $4.524 billion, and $15.5 billion of authorization remained on 5 August. If that authorization were still unused, Friday’s price would turn it into about 11.9 million shares, roughly 8 percent of the July count. No mid-October release updates it.

July-quarter customers were three books. Edge revenue was $5.432 billion, or 60.6 percent of the quarter. Datacenter was $2.977 billion, or 33.2 percent. Consumer was $556 million, or 6.2 percent, and it fell while the other two rose.

Micron holders are a read-across, not the same position. Micron sells DRAM and NAND, while SanDisk, separated from Western Digital on 21 February 2025, describes itself as the NAND company. Micron’s last reported quarter is the industry print already on the desk. The NAND piece of that quarter is a comparison, not a SanDisk forecast.

What the tape and the August release show

No NAND price break was in the pages fetched for this close. The tape and the 5 August release answer a different question than the headline.

The 5 August release is still the last full operating picture. Revenue of $8.965 billion was up 51 percent from $5.950 billion, and about two-thirds of that $3.015 billion sequential gain came from higher pricing.

The Q1 guide is not a collapse: the $10.55 billion midpoint is 17.68 percent above the quarter just printed. The 13 August investor day repeated the tight-supply claim, and the night of those results already showed a strong quarter the stock could sell.

What SanDisk and the named analysts have said

David Goeckeler, chairman and chief executive, last spoke in the 5 August release, not this week. He said: “We closed fiscal 2026 with a leading technology portfolio, established datacenter as a key growth pillar, and deepened our customer partnerships.” He added: “Our technology and products are well positioned to create value for our customers and generate growing and durable free cash flow.” The lines do not mention CXMT or the share price.

On the same call, as transcribed by The Motley Fool, he said: “I think mid-80s gross margin, I would characterize as a fair return.” He also said, “In ’23, we are having very different earnings calls than we’re having now.” No October document fetched here updates that call.

The Motley Fool’s 12 October piece quotes chief financial officer Luis Visoso calling the guided sequential gain more storage shipped and “modest price increases.” That is the Fool’s quotation of the call. The same piece says TrendForce expects NAND contract prices to climb another 15 percent to 20 percent in the October-to-December quarter. A TrendForce sheet dated 16 October was not saved here, so that range stays the Fool’s citation.

Susan Park, vice president of product management for consumer products, is the company voice from this week. In the 14 October release she said: “We are thrilled to be the exclusive storage provider to help Meta power the future of virtual reality.” The cards “integrate seamlessly with new Meta VR Glasses offering immediate access to additional storage capacity.” The glasses launch in Spring 2027. That day’s close fell 5.33 percent, from $1,482.33 to $1,403.34, and the release does not mention it.

Jordan Klein, a TMT specialist at Mizuho, wrote on 13 October in a note carried by Yahoo Finance: “I would not be selling or in a panic if I owned MU, SNDK, SKHY or Samsung over comments yesterday by CXMT sr mgmt. about their new advanced DDR5 design roadmap. To me this changed nothing near term.” He added: “A year-end product introduction can’t be directly compared with Samsung’s reported 2028 mass-production target. A launch doesn’t establish manufacturing maturity, yields, or cost competitiveness.” He was talking about DRAM.

John Tinsman, founder of AOTInvest and portfolio manager of the AOT Growth and Innovation ETF, posted on 12 October: “I don’t buy into this hype but as a result today memory makers are trading down.” He had SanDisk down about 6 percent. Monday’s official close was $1,487.33, down 5.97 percent from $1,581.82.

A BazaarBrief post that afternoon said SanDisk is a NAND company that fell with the group anyway. A Ticker Report post on Friday printed $1,313.24, which is not the $1,299.30 close, and called the CXMT concern indirect.

John Ballard, at The Motley Fool on 16 October, sketched $4,128 over the next year or so if storage demand keeps outrunning supply. That figure is his path, not the bull below, and Nasdaq’s $2,200.00 target is a separate street mark.

The macro tension is real, and it is still not a NAND print

Smith’s Friday piece adds inflation, not a wafer price. He wrote that September consumer prices rose 3.7 percent year over year, 10 basis points faster than expected, and that higher inflation can push the Federal Reserve toward higher rates. That is his account of the Labor Department release, not a table saved here. SanDisk’s long-term debt was zero at 3 July, so a rate story is about the customer’s financing, not SanDisk’s own debt.

Pricing did about two-thirds of the last sequential revenue jump while consumer revenue was already falling. The Meta card does not replace the $8.409 billion of July edge and datacenter revenue. The 14 October risk factors name trade policy, tariffs, pricing, competition and Kioxia, which is boilerplate, and the dated test is still the 29 October call.

Bull, base, and bear into 31 December

The title bull, base and bear are distribution maths on SanDisk’s own closes. They are not Ballard’s $4,128 and not Nasdaq’s $2,200.00. The 22 June intraday high of $2,354.39 is a traded price, not this bull of $2,532.60 or this bear of $662.86.

The inputs are the last 90 daily log-returns, ending on the move from $1,394.83 to $1,299.30 and starting after the 9 June close of $1,646.54. The mean, mu, is -0.00263167. Sigma, the sample standard deviation using 89 in the divisor, is 0.07435442.

N is 52, from Monday 19 October through Thursday 31 December 2026, with weekends out. The full closures skipped are Thanksgiving on 26 November and Christmas on 25 December. Early closes on 27 November and 24 December still count, and the month split is 10, 20 and 22.

The square root of 52 is 7.21110, so 52 times mu is -0.136847, 1.5 sigma times that root is 0.804266, and one sigma times it is 0.536177. Drift is smaller than the upside shock, so the bull can clear a spot whose recent average day was down. This is not financial advice.

Base is $1,299.30 times e to the power of -0.136847, or times 0.872104, which is $1,133.12. That is $166.18 under the close, or 12.79 percent. Bull is $1,299.30 times e to the power of -0.136847 plus 0.804266, or times 1.949200, which is $2,532.60, up $1,233.30, or 94.92 percent.

Bear is $1,299.30 times e to the power of -0.136847 minus 0.536177, or times 0.510163, which is $662.86. That is $636.44 under the close, or 48.98 percent. The bull stayed above the spot, so the drift was kept. Nasdaq’s $2,200.00 sits $332.60 under this bull, and Ballard’s $4,128 sits $1,595.40 above it.

What happens next

Monday 19 October is the first dated test. One sample deviation, about 7.4 percent or roughly $97 on Friday’s price, is an ordinary day on this tape, not a forecast that Monday moves $97. A repeated CXMT headline with no new NAND price can still extend the streak. That would be volatility, not evidence of a cut contract.

Thursday 29 October 2026 is the earnings date on the saved Nasdaq file. August guided the quarter to $10.30 billion to $10.80 billion and a non-GAAP gross margin of 83 to 85 percent. A print inside that range matches a pre-selloff guide. It does not prove NAND prices broke between 8 and 16 October.

The causal test that day is the December-quarter language. If management still describes pricing as up, the way Visoso’s “modest price increases” described the guided quarter, the NAND-break reading fails on the company’s words. A cut to the margin range would give the bear a fundamental input the Friday file does not have.

Thursday 31 December 2026 is the horizon. Drift alone lands at $1,133.12. Drift minus one standard deviation lands at $662.86. Drift plus 1.5 standard deviations lands at $2,532.60, above the June close and the June intraday high, and none of the three requires a new broker target.

Frequently asked questions

Is SanDisk’s slide a NAND price break, or the stock’s own volatility, and where do the year-end bull and bear sit?

It is the stock’s own volatility, not a documented NAND price break. The bull is $2,532.60 and the bear is $662.86, with a base of $1,133.12, off the $1,299.30 close on 16 October 2026. The path uses 52 sessions through 31 December and the last 90 daily log-returns. Friday’s log-return was 0.95 of one sample deviation, and CXMT’s reported product is a DRAM module.

What did SanDisk close at on Friday 16 October 2026?

The official close was $1,299.30, down $95.53, or 6.85 percent, from Thursday’s $1,394.83. The Nasdaq historical file saved 17 October shows an open of $1,348.82, a high of $1,362.35, a low of $1,296.19 and volume of 8,954,661. An intraday post at $1,313.24 is not this close, and it is not a bar on the chart.

Does CXMT’s denser cell change SanDisk’s NAND price?

Not in any document fetched for this piece. Cao Kanyu’s reported product is a server DDR5 module on a 4F-squared DRAM cell, with no speeds or capacities in the Tom’s Hardware account. Rich Smith wrote on 15 October that SanDisk does not make DRAM and manufactures NAND instead. The 14 October release was a Meta storage deal, not a wafer-price update.

What is the base case into 31 December?

The base is $1,133.12, which is $166.18, or 12.79 percent, under Friday’s close. It is $1,299.30 times e to the power of 52 times the mean daily log-return of -0.00263167, and nothing else. NAND prices do not have to fall for that path. The bull adds 1.5 standard deviations, the bear subtracts one, and the base is not Nasdaq’s $2,200.00 target.

Are the $2,200 target and the $4,128 figure the bull case?

No. Nasdaq’s saved summary lists a one-year target of $2,200.00, which is $332.60 below this bull of $2,532.60. John Ballard wrote on 16 October that the shares could be worth $4,128 if storage demand keeps outrunning supply. That sketch is $1,595.40 above this bull, both figures sit outside this distribution, and the bear here is $662.86.

What is the next dated test for this SanDisk stock prediction?

Nasdaq’s quote file lists an earnings date of 29 October 2026. The August guide was $10.30 billion to $10.80 billion of revenue and a non-GAAP gross margin of 83 percent to 85 percent. A result inside that range would match a guide set before the seven-session slide. The comment that would change the NAND-break reading is the December-quarter price language.