For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for Sandisk Corp. (SNDK) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-10-07): Designation Low · Gem Score -17 (−100…+100 Quality+Value blend) · Quality -20 · Value -15 · Sentiment 67 (timing only, not weighted) · Composite fair value $3,083.02 vs $1,614.39 at analysis
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
Sandisk Corp.
SNDK NASDAQSandisk Corp. is a technology company that develops, manufactures, and sells data storage devices and solutions based on NAND flash memory. Its product portfolio includes solid state drives, memory cards, USB flash drives, embedded storage products, and related wafers and components. The company serves a wide range of end markets, including personal computing, gaming systems, mobile devices, automotive, industrial, connected home, and cloud infrastructure. Sandisk also supplies products to computer manufacturers, original equipment manufacturers, data centers, cloud service providers, distributors, resellers, and retailers. Headquartered in Milpitas, California, Sandisk focuses on flash and advanced memory technologies that support everyday consumer electronics as well as enterprise and industrial storage applications.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics TTM · through Jul 3, 2026
EPS (Diluted): 74.24
Total Equity: $15.74B
Shares: 154,000,000
Total Debt: $0.00
Cash: $4.76B
EBITDA: $12.54B
Total Debt: $0.00
Cash: $4.76B
Revenue: $20.25B
Revenue: $20.25B
Revenue: $20.25B
Total Equity: $15.74B
Tax Rate: 12.2%
Equity: $15.74B
Total Debt: $0.00
Cash: $4.76B
Current Liabilities: $5.58B
Long-Term Debt: $0.00
Total Debt: $0.00
Total Equity: $15.74B
Shares: 154,000,000
Shares: 154,000,000
CapEx: -$177.00M
Shares: 154,000,000
Stock Price: $1,614
Net Income: $11.43B
Industry Benchmarks
Income Statement (Annual)
Last updated: Sep 17, 2026 2:03am (20d ago)| Metric | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|
| Revenue | $6.1B | $6.7B | $7.4B | $20.2B |
| Cost of Revenue | $5.7B | $5.6B | $5.1B | $5.8B |
| Gross Profit | $430.0M | $1.1B | $2.2B | $14.5B |
| Operating Expenses | $2.5B | $1.5B | $3.6B | $2.1B |
| Operating Income | -$2.0B | -$468.0M | -$1.4B | $12.4B |
| Net Income | -$2.1B | -$672.0M | -$1.6B | $11.4B |
| EBITDA | -$1.6B | -$244.0M | -$1.2B | $12.5B |
| EPS | $-14.78 | $-4.63 | $-11.32 | $77.78 |
| EPS (Diluted) | $-14.78 | $-4.63 | $-11.32 | $73.76 |
Balance Sheet (Annual)
Last updated: Sep 17, 2026 5:08pm (20d ago)| Metric | 2022 | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|---|
| Cash & Equivalents | — | — | $328.0M | $1.5B | $4.8B |
| Total Current Assets | — | — | $3.5B | $5.1B | $12.8B |
| Total Assets | — | — | $13.5B | $13.0B | $22.5B |
| Current Liabilities | — | — | $2.1B | $1.4B | $5.6B |
| Long-Term Debt | — | — | $0 | $1.8B | $0 |
| Total Liabilities | — | — | $2.4B | $3.8B | $6.8B |
| Total Equity | $13.0B | $11.4B | $11.1B | $9.2B | $15.7B |
| Retained Earnings | — | — | $0 | -$1.8B | $9.6B |
Cash Flow (Annual)
Last updated: Sep 17, 2026 5:13pm (20d ago)| Metric | 2023 | 2024 | 2025 | 2026 |
|---|---|---|---|---|
| Operating Cash Flow | -$713.0M | -$309.0M | $84.0M | $11.7B |
| Capital Expenditure | -$219.0M | -$166.0M | -$204.0M | -$177.0M |
| Free Cash Flow | -$932.0M | -$475.0M | -$120.0M | $11.5B |
| Acquisitions (net) | — | — | — | — |
| Net Debt Issued / (Repaid) | $0 | $0 | $1.9B | -$1.9B |
| Dividends Paid | — | — | — | — |
| Stock Buybacks | — | $0 | $0 | -$4.5B |
| Net Change in Cash | -$43.0M | $36.0M | $1.2B | $3.3B |
Growth Trends (YoY %)
Last updated: Sep 17, 2026 2:03am (20d ago)| Metric | 2024 | 2025 | 2026 |
|---|---|---|---|
| Revenue Growth | +9.5% | +10.4% | +175.3% |
| Gross Profit Growth | +149.3% | +106.3% | +554.2% |
| Operating Income Growth | +77.0% | -194.2% | +999.7% |
| Net Income Growth | +68.6% | -144.2% | +796.7% |
| EBITDA Growth | +84.6% | -397.5% | +1,132.8% |
Deep Analysis
Risk : Reward — upside vs downside from this company's own quarters
Computed 2026-10-06 02:03A +1σ run of quarters pays -0%; a −1σ run costs 77%. Ratio 0.0:1 (μ 170.0%, σ 168.3% , 6 pairs).
| Case | Growth | Margin | Fair value | vs price ($1,614.39) |
|---|---|---|---|---|
| Bull — recovery | +180% | 35.0% | $19,289.37 | +1095% |
| Base — stabilizes | +120% | 35.0% | $7,081.42 | +339% |
| Bear — keeps slipping | +60% | 35.0% | $2,044.35 | +27% |
| Upside — a +1σ run of quarters (v2) | +50% | 35.0% | $1,610.70 | -0% |
| Stress — a −1σ run of quarters (v2) | +2% | 29.6% | $377.77 | -77% |
Narrative Economics
market-narrative step).
Growth Outlook
Analyzed 2026-09-17 17:16The question every valuation on this page silently assumes: is this company likely to grow? Judged forward — the business, its category, the world — against what's already printed.
Claude Reading
The quarterly trajectory is the entire story and it's screaming cyclical peak, not structural growth. Revenue went $1.88B → $1.70B → $1.90B → $2.31B → $3.03B → $5.95B → $8.97B across seven quarters, with net margin traversing -114% → 77% in the same window. That is textbook NAND supercycle behavior — think Micron 2017-2018 or Hynix 2021 — not the earnings algorithm of a durable compounder. The pre-flight AI's speculation that "this is NOT the historical flash memory manufacturer" and is secretly a GPU/AI accelerator play is simply wrong; SNDK is Sandisk, the NAND spinout from Western Digital, and the margin curve traces exactly the current enterprise SSD/HBM-adjacent shortage cycle. Any model output built on the "traditional / AI accelerator" premise should be discarded.
Once you accept this is cyclical NAND, the valuation math changes completely. TTM revenue of $20.25B is running at an annualized $35.9B pace based on the July quarter alone; NI is annualizing at ~$27.6B. A 21.7x TTM P/E on peak earnings for a memory manufacturer is not cheap — it's the classic trap where the P/E looks reasonable precisely because E is unsustainable. Mid-cycle revenue for Sandisk is probably $9-12B (vs FY2024's $6.66B trough and $7.36B FY2025), and mid-cycle net margin for standalone NAND historically sits at 5-15%, not 56%. Normalize to $10B revenue at 10% net margin = $1B mid-cycle earnings; even at a generous 20x that's a $20B fair value, roughly 8% of the current $236B market cap. The synthesis model's $2,757 "signal-adjusted" fair value is derived from a DCF that extrapolates peak margins — it should be ignored. The Thesis Evaluation's -30 score is directionally right but understates the severity.
The corroborating signals all point the same way: insider activity is exclusively sales and in-kind (tax) dispositions with zero buys, FCF quality is flagged as poor (worth investigating — NI of $11.43B vs OCF of $11.67B is actually reasonable conversion, so the "poor" flag likely reflects working capital swings from the price spike inflating receivables), and the Market Forces model correctly identifies this as peak-cycle. The narrative layer's observation that the discount is "rational caution, not storytelling" gets the vibe right but misreads the discount — the market isn't discounting Sandisk 41% below DCF; the DCF is wrong. A contrarian bull would argue that HBM-adjacent enterprise SSD demand for AI training clusters represents genuine structural NAND consumption growth that could extend this cycle 2-3 years longer than typical, and that Sandisk's zero-debt, $4.76B cash balance sheet lets it survive any downturn and consolidate. That's a real argument, but it argues for holding through the cycle, not for paying 12x sales at the peak.
I dissent from the synthesis "fair value" verdict. This is overvalued on any normalized earnings basis — the composite methodology is anchored to trailing peak numbers and produces nonsense outputs ($4,662 fair value on a $1,614 stock trading at peak-cycle earnings is a red flag that the model is broken for cyclicals). The right frame is EV/mid-cycle-sales or normalized P/E, both of which put fair value in the $400-700 range, implying 55-75% downside when the cycle turns. Timing is the only real question: NAND spot pricing, hyperscaler capex commentary in the next 1-2 quarters, and any signs of Samsung/Hynix capacity additions are the tells. Insiders are already selling into strength; that's the informed money's answer to the timing question. A disciplined short or put position sized for a 6-12 month cycle inflection is the trade; long-only investors should be trimming aggressively, not adding.
GPT Reading
The numbers do not describe an ordinary storage or memory name; they describe an earnings machine that has gone from distressed to absurdly profitable in four quarters. Revenue in the trailing twelve months is $20.25B versus $7.36B the year before, a 175% jump, but the more important point is what happened underneath: gross margin went from 30.0% to 71.5%, operating margin from -18.8% to 61.2%, and net margin from -22.3% to 56.5%. Quarterly progression is even more extreme: revenue rose from $2.31B in the October quarter to $3.03B, then $5.95B, then $8.97B in July, while net income exploded from $112M to $803M to $3.62B to $6.90B. A 77% net margin in the latest quarter is not merely strong; it is so far outside normal hardware economics that it demands skepticism about sustainability, but it also means the stock at $236B is only 20.7x this trailing earnings base and about 20.6x trailing free cash flow of $11.49B. For a business currently compounding revenue at this rate with zero debt and $4.76B cash, that is not an obviously expensive multiple if even part of the margin step-change is real.
What stands out most is the cash conversion. Operating cash flow of $11.67B against net income of $11.43B says the earnings are not obviously fake, and capex of just $177M leaves almost all of it dropping into free cash flow. That capex figure is bizarrely low for a real capital-intensive semiconductor manufacturing business, which reinforces the suspicion that the reported classification is mismatched or that this entity is economically far more fabless and IP-rich than the label suggests. Either way, on the raw numbers, returns are extraordinary: ROE 91.6%, ROIC 99.2%, no leverage, current ratio 2.29. If I ignore the name and look only at the financials, the market is pricing a company that just delivered one of the sharpest earnings inflections in large-cap tech at only 11.5x EV/revenue and 18.5x EV/EBITDA. Those are not giveaway multiples, but they are also not peak-bubble multiples for a business posting nearly $9B of quarterly sales and almost $7B of quarterly profit.
My read is that the market is still applying a cyclical discount to a business that may have structurally changed, but I cannot get to the very bullish fair-value outputs that suggest 70% upside without imposing heroic durability assumptions. The issue is simple: if the latest quarter annualized were the new run-rate, this stock is cheap; if the latest quarter is a spike, it is not. The full-year TTM operating income of $12.39B is already heavily back-end loaded, with the July quarter alone contributing more than half of TTM net income. That concentration matters. A stock at $1,614 looks undervalued if earnings can hold anywhere near $9B-$11B annually over the next couple of years, but overvalued if profits normalize back toward even half the current margin structure. My base case is that this is not a complete mirage, because the ramp was sequentially consistent across four quarters and cash tracked it, yet the latest-quarter profitability is too extreme to capitalize at face value. I land modestly bullish: the market cap implies a sharp comedown already, and the balance sheet removes the usual cyclical balance-sheet risk.
The best case against my view is strong and data-driven: almost everything good here could be peak-cycle. A jump from 4.9% net margin in the October quarter to 77% by July is more characteristic of a temporary shortage, accounting distortion, or one-off pricing environment than a durable competitive moat. The valuation optics also flatter the stock because the denominator is inflated by the most recent quarter; trailing P/E of 21.7 and EV/EBITDA of 18.5 may be cheap only if the E and EBITDA are sustainable. Price-to-book at 15.8 is rich for a hardware-linked business, and the tiny capex base relative to revenue makes the free-cash-flow figure hard to compare with normal semiconductor economics. If this really is tied to memory or other commodity-like silicon, history says supernormal margins collapse faster than models adjust. On that framing, $236B on $20.25B of sales and one extraordinary year of profits is absolutely too high, and the recent small insider sales do nothing to challenge that caution.
What would change my mind is straightforward. If the next two quarters show revenue holding above roughly $7B per quarter while gross margin stays above 60% and quarterly operating cash flow remains in line with earnings, then the current price is too low and I would move materially more bullish, because the “peak quarter” argument would weaken fast. Conversely, if revenue stalls below $6B or net margin snaps back below 35%-40% with cash flow lagging earnings, I would treat 2026 as a one-year profit spike and view $1,614 as too rich, with fair value closer to a mid-teens multiple on normalized earnings rather than on the latest run-rate.
Grok Reading
The raw numbers describe a company that did not merely recover—it inverted its entire economic identity in four quarters. Trailing revenue hit $20.25B versus $7.36B the prior year, with the quarterly sequence $2.31B → $3.03B → $5.95B → $8.97B still accelerating into the July 2026 print. Net income swung from a $1.64B loss to $11.43B profit; the latest quarter alone delivered $6.90B of net income on $8.97B of sales (77% margin). TTM gross margin sits at 71.5%, operating margin 61.2%, net margin 56.5%. Free cash flow of $11.49B converted almost dollar-for-dollar from $11.67B of operating cash against a trivial $177M of capex, on a fortress balance sheet carrying zero debt and $4.76B cash. At $1,614 and a $236B enterprise, the stock trades at 21.7× TTM earnings, 12.3× sales and roughly 4.9% FCF yield—multiples that would look ordinary for a software compounder and cheap for a hardware name printing mid-50s net margins with 92% ROE and 99% ROIC. The contradiction with the “NAND cycle peak” framing is the capex line itself: no traditional memory manufacturer generates $20B of revenue and $11.5B of FCF while spending $177M on capital. Either the cost structure has become radically asset-light, or the earnings power is being mis-categorized as a pure commodity ASP spike.
That said, the bear case writes itself from the same tape. Four quarters earlier the company was losing money at scale; six quarters earlier net margin was –114%. Memory pricing cycles have historically erased precisely this kind of margin expansion within 12–18 months once supply responds, and a 56% net margin is an invitation for both competitors (Samsung, SK Hynix, and any hyperscaler captive effort) and customer pushback. The valuation synthesis that spits out a signal-adjusted fair value near $2,760 embeds continuation of growth rates the momentum block already flags as extreme (145% revenue CAGR, 175% recent YoY). If even half of the latest-quarter run-rate proves transient and TTM earnings normalize toward $5–6B, the current 22× multiple expands into expensive territory fast. Insider activity is cleanly neutral (mostly tax-withholding F-prints), so it supplies no counterweight, and the narrative layer is correctly thin—there is no cult premium to cushion a down-cycle. The market’s 31% drawdown from the highs and the sub-22× multiple on peak reported earnings already embed substantial skepticism; treating that skepticism as pure mispricing requires believing AI storage demand permanently rewrites NAND (or NAND-adjacent) economics.
I am willing to own the stock as modestly undervalued relative to demonstrated cash generation, but only with the explicit assumption that mid-cycle net margins settle in the high-20s to mid-30s rather than collapsing back to mid-single digits, and that revenue does not give back more than ~30% of the trailing peak. What flips me is straightforward: two consecutive quarters of sequential revenue decline combined with gross margin compression below 45%, or a capital-spending guide that suddenly jumps into the multi-billion range (revealing the current FCF as a temporary under-investment artifact). Conversely, another quarter north of $8B revenue with net margins still above 50% would force a material upward revision to sustainable earnings power and make $1,614 look like a clear mispricing.
Qwen Reading local experiment · off-panel
The data presented here is a textbook case of a cyclical commodity super-cycle masquerading as a structural growth story, and the quantitative models are dangerously overfitting to the peak. The most glaring anomaly is the revenue trajectory: from $1.70B in Q1 2025 to $8.97B in Q3 2026, a 5x expansion in just two years. This is not the organic growth profile of a technology innovator; it is the violent price elasticity of NAND flash memory during a severe supply shortage. The "Pre-Flight Intelligence" model suggests this might be a GPU/AI accelerator business due to the 71.5% gross margins, but that is a fundamental misclassification. SanDisk is a memory maker, and while AI data centers are driving unprecedented demand for high-density SSDs, the underlying product remains a commodity. The 77% net margin in the most recent quarter is a statistical outlier that has never been sustained in the history of the memory industry. When you look at the annual data, the 2025 fiscal year showed a net loss of $1.64B on $7.36B in revenue. The swing to $11.43B in net income in 2026 is entirely driven by pricing power, not volume growth alone. The market is currently paying $236.38B for a company that is effectively printing money, but it is doing so because the supply of NAND wafers is constrained, not because SanDisk has a moat.
The valuation metrics are screaming a warning that the "high_growth_profitable" archetype is a trap. A P/E of 21.7x looks reasonable on the surface, but it is calculated against a TTM net margin of 56.5%, which is historically impossible for this sector. If we normalize the net margin to a historical average of 10-15%, the earnings power drops by 75-80%, making the current price exorbitant. The "Valuation Synthesis" model suggests a fair value of $2,757, which is a 70% upside, but this relies on the assumption that the current cash flow of $11.49B is sustainable. It is not. The "Market Forces" signal correctly identifies "poor cash flow quality," which is a cryptic way of saying that the operating cash flow is inflated by working capital benefits from rising prices, not by structural efficiency. The balance sheet is clean—zero debt, $4.76B in cash—but that is a byproduct of the boom, not a sign of long-term health. The insider activity is also telling: while the volume is small, the consistent "InKind" distributions and sales in August and September 2026 suggest insiders are taking profits while the public is still buying the "AI storage" narrative.
The strongest case against my bearish read is the genuine structural shift in demand. The "Market Narrative" layer correctly notes that every LLM training run requires massive storage, and SanDisk is a key supplier. If the AI buildout continues at the pace suggested by the 175% YoY revenue growth, the supply shortage could persist longer than historical cycles. The "Thesis Evaluation" gives a score of -30, which is actually too bullish for a commodity at peak pricing. The bull case argues that this is a "quiet-quality" asset with pricing power, but memory manufacturers have never had pricing power; they have only had temporary pricing leverage. The contradiction here is that the models are treating SanDisk like a software company with high margins and low capex, but it is a hardware company with high capex and high volatility. The "Sector Intelligence" says "Above Sector Benchmarks," but that is because the entire sector is in a boom. When the cycle turns, as it inevitably does, the benchmarks will collapse.
What would change my mind is a clear signal that the supply shortage is structural rather than cyclical. If SanDisk announces a long-term contract with a hyperscaler that locks in pricing for three years, that would validate the "structural growth" thesis. Alternatively, if the next two quarters show revenue growth of 20-30% (not 175%) with margins holding above 40%, that would suggest a new equilibrium. But until then, the data points to a peak-cycle company being valued as a growth stock. The $1,614 price is not a discount; it is a premium paid for a temporary anomaly. The "Verdict" from the synthesis model is "fair_value," but that is only true if you believe the current earnings are permanent. I do not.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
Sandisk is a pure-play NAND flash business that lost money for three consecutive years (2023 net loss $2.14B, 2024 $-672M, 2025 $-1.64B, all with negative FCF) before erupting in the TTM to $20.25B revenue, 71.5% gross margin, 61.2% operating margin, $11.43B net income and $11.49B FCF. Balance sheet is now strong: $4.76B net cash, Altman Z 24.65, and the business is decisively self-funding at the top of cycle. This is textbook memory-cycle behavior - massive operating leverage on a commodity product - not evidence of a newly durable franchise.
Verify before trusting this (6)
- Reconcile the $11.43B net income vs OCF gap - working capital build, inventory, receivables composition in the 10-K
- NAND pricing environment and customer concentration disclosures - how much of the margin surge is spot pricing vs contracted
- Capex plans and fab investment cycle - is retained FCF being redeployed into the next capacity cycle
- Convertible or debt structure and any covenants tied to the recent spinoff from Western Digital
- Segment / product mix detail to gauge whether any portion is less cyclical than raw NAND
- Share issuance breakdown - what drove the ~9M share jump in diluted count this year
The e2e composite pins fair value at $4,663 and the signal-adjusted number at $2,758, versus a $1,614 price. Taken at face value that is a 40-70% discount, but the method spread is enormous: DCF says $5,955, anchored P/E says $3,893, and the EPV floor collapses to $250. That spread is the whole story - the DCF and P/E anchors are extrapolating a peak-cycle NAND print (net income $11.4B, but operating cash less than half of that and a flagged Beneish score) as if it were normalized. The EPV floor, which strips growth and capitalizes normalized earnings, is telling you what this commodity business is worth if the cycle mean-reverts. Truth is somewhere in between. Haircutting the composite for cyclicality and earnings quality (Mixed, quality score -20), a defensible deserved value is roughly $2,000-$2,500 - modestly above today's $1,614. That is a real but not glaring gap. The market is not asleep here; it is discounting the peak print because memory cycles always turn. I would call this modestly cheap on a through-cycle basis, not a fat-pitch dislocation. Deep Value would require either the EPV floor to be wrong (structural NAND demand from AI holds pricing) or the price to fall meaningfully further.
Verify before trusting this (5)
- NAND pricing and bit-shipment guidance in the next print - is the up-cycle extending or rolling
- reconciliation of net income vs operating cash flow; working capital and inventory build
- capex plan and free cash flow through the cycle
- any one-time gains, tax benefits, or accounting changes inflating the headline EPS
- management commentary on hyperscaler SSD demand durability
The dominant force on this name is narrative, not macro. SNDK is up roughly 1,700% in a year, just got promoted to the S&P 100, and the tape is full of stock-split chatter, 'memory shortage is not over' calls, and price-target ladders pointing to $2,125-$3,180. That is textbook mania pressure - the stock trades on flow, momentum, and the AI-infrastructure story rather than on the quiet-quality fundamentals the brief flags. Even the recent 35% drawdown from the peak is being framed by the financial press as a dip-buying opportunity, which itself is a tailwind signal. The risk-off tape (VIX 17.7, S&P -3% off highs) is a real crosswind for a high-beta, parabolic memory name - the September 14 AI-slowdown selloff and repeated profit-taking days show this stock amplifies any risk-off pulse 3-5x. But the offsetting narrative pressure is enormous: index inclusion just landed, analyst tone is aggressively bullish, peers (Micron) are being talked up in tandem, and there is no crack in the memory-cycle story yet. Net, the non-fundamental push is decisively up, with the caveat that a name this stretched can swing violently on any narrative wobble.
Verify before trusting this (5)
- NAND spot pricing trend into Q4 - any softening cracks the shortage narrative
- Post-index-inclusion flow exhaustion after September 21
- Whether Micron earnings/guide confirms or breaks the memory-cycle story
- Any concrete split announcement (would extend retail leg) or absence of one (deflates buzz)
- VIX trajectory - a spike above 22 would force de-risking of high-beta winners like SNDK
The world changed in Sandisk's favor: AI data center buildout turned NAND from a consumer-device commodity into a constrained infrastructure input, with QLC nearline SSDs structurally displacing HDD capacity. That is a real, durable demand-level shift. What is NOT durable is the price level clearing that demand — it exists only because supply was starved by the 2023-24 capex famine. So the honest read splits: demand base higher forever, pricing higher temporarily. Sandisk captures the first as volume and the second as a windfall. Anyone extrapolating the windfall is extrapolating the wrong variable.
When we made this prediction on Sep 18, 2026, SNDK was $1,738.43. We expect it to be $1,980.00 by Mar 2027, and we consider it great value under $1,350.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Sep 18, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.
Post-Report Due Diligence NOTES
Evidence for a closer look, not a verdict — no score or designation on this page has been changed by it. Items marked material are ones where a conclusion above moves to the other side of the price.