I.Executive Summary
Sandisk makes NAND flash memory and the SSDs, cards and embedded storage built on it. Western Digital spun it off on 21 February 2025, so it has only about a year and a half of history as a standalone public company. Earlier figures are "carve-out" accounts of Western Digital's flash business. In fiscal 2026 (the 53 weeks to 3 July 2026) revenue rose 175% to $20.25 billion, GAAP operating margin was 61.2%, net income was $11.43 billion and diluted EPS was $73.76, after a net loss in each of the three prior years. Management guides first-quarter fiscal 2027 revenue to $10.3–10.8 billion, more than half of the whole of fiscal 2026.
The stock has risen about ten-fold in just over a year: from a $112.20 close on 30 September 2025 to $885.01 on 13 November 2026. On consensus fiscal 2027 earnings it trades at 3.9×, which looks cheap. On trailing earnings it trades at 12.0×, on book value at 8.2×, and against its average profit over the five years we can see (FY2022–FY2026, $1.6 billion a year) at about 105×. The history that exists is a full cycle in miniature. Operating margin went from 12% (FY2022) to −33% (FY2023) and back to 61% (FY2026), and gross margin was just 29.8% as recently as the September 2025 quarter.
Sandisk is a clean, debt-free NAND pure play riding an extraordinary shortage, and it is signing long-term contracts to lock the shortage in. Valuing a mid-cycle level of earnings, 30–50% of the fiscal 2028 consensus, at 9–13×, our projection implies a roughly flat annual return in the base case.
Our illustrative five-year projection takes the fiscal 2027–fiscal 2028 consensus as Years 1–2, then applies a mid-cycle EPS level (a stated share of the fiscal 2028 consensus) and an exit P/E of 9.0×–13.0×; the base case is the midpoint of the bear–bull range. That gives Year-5 values of $780 (bear), $1,234 (base) and $1,879 (bull), against today's $885.01: an implied return of about +6.9% a year in the base case (−2.5% to +16.2%). Discounted at 10% a year, the base case is worth $767 today, −13% versus the price. Today's price enters only the implied return; the Year-5 values change only when earnings estimates or the stated assumptions do. Evidence: Weak. 21–22 analysts, but the normalisation rests entirely on assumptions and the standalone history is short. *All values are Pathos's illustrative estimates for educational purposes, not price targets or recommendations.
II.Business Overview
What it sells. NAND flash wafers, components and finished products in three end markets (fiscal 2026 revenue): Edge (PC and mobile storage, $12.16B), Datacenter (enterprise SSDs, $5.15B, up 437%), and Consumer (Sandisk-branded (and, for a transition period, WD-branded) cards, USB drives and portable SSDs, $2.94B). Sequential Q4 growth came about one-third from volume and two-thirds from price, according to the results release.
How it makes it: the Kioxia joint ventures. This is the most important structural fact for investors. Per the 10-K, all of Sandisk's flash memory comes from its joint ventures with Kioxia ("Flash Ventures"). These operate eight fabs in Yokkaichi and Kitakami, Japan, which Kioxia owns. Sandisk buys about 50% of the output at cost plus a small markup, pays half of the ventures' fixed costs whatever volume it takes, and funds 49.9–50% of their capital investment when their own cash flow falls short. It may not make flash elsewhere or with third parties while the ventures operate. On 29 January 2026 the ventures were extended to 31 December 2034, and Sandisk agreed to pay Kioxia $1.2B over 2026–2029 for manufacturing services and continued supply. The consequence is that Sandisk's own capex ($177M in FY26) understates the true capital needs of its supply. Much of that spending sits inside the ventures and reaches Sandisk through wafer costs, JV loans and guarantees.
Who buys it. The top ten customers accounted for 44% of fiscal 2026 revenue (40% in FY25). Starting in fiscal 2026 Sandisk has signed multi-year "New Business Model" (NBM) agreements with Datacenter and Edge customers. These commit volumes, use fixed plus variable pricing, and are backed by financial guarantees such as cash deposits. The Q4 release says ten had been signed. At 3 July 2026 remaining performance obligations under NBMs were $59.8B, about 19% of it expected within twelve months, and after year-end two more NBMs added $31.3B. For comparison, fiscal 2026 revenue was $20.2B.
Competition. The 10-K names Kioxia (also its partner), Micron, Samsung, SK hynix and YMTC, plus many smaller module makers.
III.Share Price & Valuation Multiples
Price as of Nasdaq close, 13 Nov 2026 · Fundamentals: FY2026 10-K| Metric | Value | Basis |
|---|---|---|
| Share price | $885.01 | Close, 13 Nov 2026 |
| 52-week range | $183.00 – $2,354.39 | Low 21 Nov 2025 · high week of 22 Jun 2026 (intraday) |
| Market capitalisation | $129.6B | 146.42M shares (10-K cover, 7 Aug 2026) × $885.01 |
| Enterprise value | $123.1B | Mkt cap − $4.76B cash − $1.78B Nanya stake; no debt |
| P/E, trailing (GAAP) | 12.0× | FY26 diluted EPS $73.76 (non-GAAP $70.88: 16.2×) |
| P/E, forward | 3.9× | FY27 consensus EPS $227.48 (21 analysts; likely non-GAAP) |
| P/E on Q1 FY27 guidance, annualized | 4.9× | Non-GAAP EPS guide $44–46; midpoint × 4 = $180 |
| Market cap / FY22–26 average net income | 105× | Five-year cumulative net income $8.04B (FY22–24 carve-out) |
| EV / EBITDA | 9.8× | FY26 op. income $12.39B + D&A $0.15B |
| Price / Sales | 4.9× | FY26 revenue $20.25B (EV/Sales 6.1×) |
| Price / Book | 8.2× | Equity $15.74B; $107.47 per share. Excluding $4.99B goodwill: 15.7× |
| Price / Free cash flow | 11.3× | FY26 FCF $11.49B; on company "adjusted FCF" of $8.74B (excludes NBM prepayments): 19.3× |
| Dividend yield | None | No dividend planned "in the foreseeable future" (10-K) |
Because fiscal 2026 was loss-making in its first quarter and hugely profitable by the fourth, trailing multiples blend two different businesses. The Q4 annualized run-rate (EPS $43.97 × 4 ≈ $176) and the Q1 FY27 guide point to high-single-digit P/Es on current earnings. The value question is how long those earnings last. For context only (not an input to our work), the mean analyst target compiled by Yahoo Finance was $2,105.29 (24 analysts). We exclude the Nanya stake from EV at its 3 July 2026 fair value. It is subject to a three-year lock-up.
IV.Financial Record (Limited History)
Source: Form 10 information statement (FY2022) and Forms 10-K (FY2023–FY2026). *Carve-out financials of Western Digital's flash business before the Feb 2025 spin-off. OM = GAAP operating margin, including goodwill impairments of $671M (FY23) and $1,830M (FY25).
| US$ millions, FY to ~30 Jun | FY2022* | FY2023* | FY2024* | FY2025 | FY2026 |
|---|---|---|---|---|---|
| Revenue | 9,754 | 6,086 | 6,663 | 7,355 | 20,248 |
| Growth | — | −37.6% | 9.5% | 10.4% | 175.3% |
| Gross profit | 3,244 | 430 | 1,072 | 2,212 | 14,472 |
| Gross margin | 33.3% | 7.1% | 16.1% | 30.1% | 71.5% |
| Operating income | 1,200 | (2,035) | (468) | (1,377) | 12,389 |
| Operating margin | 12.3% | −33.4% | −7.0% | −18.7% | 61.2% |
| Goodwill impairment included | — | 671 | — | 1,830 | — |
| Net income | 1,064 | (2,143) | (672) | (1,641) | 11,433 |
| Diluted EPS | n/a | n/a | ($4.63) | ($11.32) | $73.76 |
| Operating cash flow | 1,151 | (713) | (309) | 84 | 11,671 |
| Capital expenditure | (410) | (219) | (166) | (204) | (177) |
| Free cash flow | 741 | (932) | (475) | (120) | 11,494 |
Fiscal 2026 by quarter: from breakeven to 85% gross margin
| US$ millions | Q1 (Oct 25) | Q2 (Jan 26) | Q3 (Apr 26) | Q4 (Jul 26) | Q1 FY27 guide |
|---|---|---|---|---|---|
| Revenue | 2,308 | 3,025 | 5,950 | 8,965 | 10,300–10,800 |
| Gross margin | 29.8% | 50.9% | 78.4% | 84.6% | 83–85% (non-GAAP) |
| Operating margin | 7.6% | 35.2% | 69.1% | 78.5% | n/a |
| Net income | 112 | 803 | 3,615 | 6,903 | n/a |
| Diluted EPS | $0.75 | $5.15 | $23.03 | $43.97 | $44–46 (non-GAAP) |
Within one fiscal year quarterly revenue almost quadrupled and gross margin rose 55 percentage points. That is the signature of a price-driven shortage, not of a gradual franchise build. The flip side is that the three years before it (FY2023–FY2025) lost a cumulative $4.5B, including $2.5B of goodwill impairments, and burned $1.5B of free cash flow.
Bars: quarterly revenue and net income, US$ billions (carve-out figures before the February 2025 spin-off). Line: trailing P/E = closing share price at quarter-end ÷ trailing-12-month diluted EPS (sum of the last four quarters; fiscal Q4 derived as full-year minus nine-month figures). Gaps in the line: no P/E where trailing EPS was negative (Q3 FY25–Q2 FY26); no P/E before the shares began trading (Q2 FY24–Q2 FY25). Both axes share zero; the P/E scale extends below 0× only to make room for net-loss bars (P/E is never plotted below zero). Sources: Form 10-Q/10-K data via SEC XBRL company facts; Yahoo Finance price history. Hover over a quarter for exact values.
Total tangible book value, US$ billions = shareholders’ equity minus goodwill and other intangible assets, at quarter-end. Before the February 2025 spin-off, equity is Western Digital’s net investment in the business (carve-out figures). Goodwill is not tagged for every carve-out quarter, so Q2–Q3 FY2024 and Q1 FY2025 are shown as n/a. Per-share values in the tooltip start in Q3 FY2025. Source: balance sheets in Forms 10-Q/10-K via SEC XBRL company facts. Hover over a quarter for the components and the value per share.
V.Cash Conversion & Earnings Quality
FY2026 cash conversion was strong. Operating cash flow of $11.67B slightly exceeded net income of $11.43B. Two items flatter it, and both should be adjusted for:
- Customer prepayments. $2.48B of the cash came from NBM prepayments and deposits, which will be delivered as product later. Sandisk's own "adjusted free cash flow", which excludes them and nets Flash Ventures loan activity, was $8.74B.
- A non-operating gain. Net income includes an $807M unrealized gain on the Nanya Technology shares bought in March 2026 for $970M. That gain is market-driven and could reverse.
- Capex is low by design. The ventures own the equipment, so Sandisk's reported capex was only 0.9% of revenue and D&A only 0.7%. Economically, its share of JV depreciation is inside cost of revenue, and JV funding runs through notes receivable (issued $462M, repaid $187M in FY26) and lease guarantees. The 10-K says Sandisk expects "increased capital investments" in fiscal 2027 as it moves to newer nodes. JV funding and the $1.2B of Kioxia payments are cash needs that EPS does not fully show, which is worth remembering when reading the P/E projection in Section VIII.
- Receivables rose sharply, to $4.71B from $1.07B (about 48 days of Q4 sales), and refund liabilities (customer rebates and credits) rose to $1.50B from $0.13B. Both are consistent with higher prices and volumes, but they are worth watching.
- Stock compensation was $232M in FY26 (2.0% of net income). Consensus EPS, which drives our P/E projection, is likely non-GAAP and excludes it.
VI.Balance Sheet & Book Value
| US$ millions | 27 Jun 2025 | 3 Jul 2026 |
|---|---|---|
| Cash & equivalents | 1,481 | 4,762 |
| Marketable equity securities (Nanya, fair value) | — | 1,777 |
| Total debt (incl. current portion) | 1,849 | — |
| Net cash / (net debt), incl. Nanya | (368) | 6,539 |
| Accounts receivable | 1,068 | 4,708 |
| Inventories | 2,079 | 2,698 |
| Notes receivable & investments in Flash Ventures | 654 | 678 |
| Goodwill | $4,994M at 3 Jul 2026, after $2.5B of impairments in FY23 and FY25 | |
| Contract liabilities (NBM advances) | 25 | 1,242 |
| Refund liabilities | 126 | 1,500 |
| Total shareholders' equity | 9,216 | 15,736 |
| Per share (146.4M shares, Aug 2026) | n/m | $107.47 |
Sandisk repaid its remaining term debt (about $1.9B) during fiscal 2026 and is now debt-free. On 9 September 2026 it refinanced its revolving credit facility at $1.5B, with collateral released if it reaches investment-grade ratings. Book value is $107 per share, and nearly a third of it ($4.99B) is goodwill inherited from Western Digital's acquisitions of the business. The real productive assets, the fabs, sit on Kioxia's and the ventures' books. That makes book value a weaker anchor here than at Micron. Even so, 16× book is a very high price for a manufacturer whose five-year average return on that book is low.
VII.Capital Allocation & Ownership History
| Date | Event | Detail |
|---|---|---|
| 21 Feb 2025 | Spin-off from Western Digital: 80.1% distributed (1 SNDK share per 3 WDC shares) | WDC kept 19.9% |
| 9 Jun 2025 | WDC debt-for-equity exchange and secondary offering | 21.3M shares (14.6%) |
| 29 Jan 2026 | Flash Ventures extended to 31 Dec 2034; $1.2B payments to Kioxia (2026–29) | Supply secured |
| 18 Feb 2026 | Further WDC exchange and offering; WDC retains 1.69M shares | 5.82M shares |
| 25 Mar 2026 | Strategic investment in Nanya Technology (~3.9%) plus a DRAM supply arrangement | $970M |
| 30 Apr 2026 | $6.0B share repurchase program | $4.54B used by 3 Jul |
| May–Jul 2026 | Repurchased 2.84M shares | Avg. $1,600 |
| 5 Aug 2026 | Additional $14.0B program | $15.5B remaining |
| 9 Sep 2026 | Revolving credit facility amended | $1.5B |
- Buybacks at a high price. All of fiscal 2026's $4.5B of repurchases were made in May–July 2026 at an average of $1,600 a share, roughly 15× book. The remaining $15.5B authorization is about 9% of today's market value. Buying back stock near a cyclical peak creates value only if the peak lasts.
- Dilution. Diluted shares (155M in FY26, ~155M guided for Q1 FY27) are well above the 146.4M outstanding, reflecting employee equity awards at a much higher share price. Sandisk also withheld about 1.0M shares ($0.6B) for employee taxes on vesting RSUs.
- The overhang is gone. Western Digital's retained stake has been almost entirely disposed of (1.69M shares left), so there is no longer a large forced seller.
VIII.Five-Year P/E Projection (Illustrative)
Method. We take fiscal 2027 and fiscal 2028 consensus EPS as Years 1 and 2 in every scenario, then apply a mid-cycle Year-5 EPS, set as a stated share of the fiscal 2028 consensus (memory earnings are cyclical and near a record), and an exit P/E to Year-5 (fiscal 2031) EPS. The bear and bull cases are the low and high ends of our best-estimate assumptions; the base case is the midpoint. Year-5 value = exit P/E × Year-5 EPS, and present value = Year-5 value ÷ 1.105. Only then do we bring in today's price, to show the implied annual return = (Year-5 value ÷ price)1/5 − 1. A price move changes the implied return, not the value; the values change when consensus or the stated assumptions do.
| Input | Value | Type and basis |
|---|---|---|
| Share price | $885.01 | Source: Nasdaq close, 13 Nov 2026 |
| Year 0: FY2026 EPS | $70.88 non-GAAP $73.76 GAAP | Source: company results (Section XIV). Non-GAAP shown because consensus is on that basis |
| Trailing P/E | 12.5× 12.0× GAAP | Calculation: price ÷ FY2026 EPS. Shown for reference, not applied |
| Year 1: FY2027 consensus EPS | $227.48 | Source: Yahoo Finance, 21 analysts (range $219.30–$241.23), 15 Nov 2026 |
| Year 2: FY2028 consensus EPS | $289.05 | Source: Yahoo Finance, 20 analysts (range $243.95–$356.85); +27.1% on Year 1 |
| Today's forward P/E (context only) | 3.9× | Calculation: price ÷ Year-1 consensus EPS (the fiscal year already under way). Not used for the Year-5 value |
| Analysts' 5-year EPS growth (context) | 63.5% a year | Source: Finviz "EPS next 5Y", 31 Oct 2026. Not used directly |
| Standalone EPS history | Too short | Source: 10-K. Net losses in FY2023–FY2025; spun off Feb 2025, so no through-cycle growth rate exists |
| Mid-cycle Year-5 EPS, share of Year 2 | Bear 30% · Base 39% · Bull 50% | Assumption: Mid-cycle EPS of 30–50% of the FY2028 consensus: NAND is more commoditised than DRAM, with no HBM-like product, and current margins are without precedent. |
| Exit P/E on Year-5 EPS | Bear 9.0× · Base 10.8× · Bull 13.0× | Assumption: 9–13× on normalised earnings. Base = midpoint |
| Required return (present value) | 10% a year | Assumption |
How the range is built. Earnings. Memory earnings are cyclical and fiscal 2028 consensus is near a record, so we do not grow it. Instead Year-5 EPS is a mid-cycle level. Mid-cycle EPS of 30–50% of the FY2028 consensus: NAND is more commoditised than DRAM, with no HBM-like product, and current margins are without precedent. That puts Year-5 EPS at $86.71 (bear) to $144.53 (bull), against $289.05 in fiscal 2028. Exit multiple. 9–13× on normalised earnings. The base case is the midpoint of that range: its implied annual return is exactly halfway between the bear and bull returns, and its exit multiple is the midpoint of the two multiples (base growth is solved to match). None of these values uses today's price: the price enters only the implied return, so a price move changes the return, not the value. For context, the shares trade at 3.9× the current-year consensus today.
Year-by-year EPS
| US$ per share | Y0 FY2026A | Y1 FY2027E | Y2 FY2028E | Y3 FY2029E | Y4 FY2030E | Y5 FY2031E |
|---|---|---|---|---|---|---|
| Bear EPS | $70.88 | $227.48 | $289.05 | $193.50 | $129.54 | $86.71 |
| Base EPS | $70.88 | $227.48 | $289.05 | $212.05 | $155.57 | $114.13 |
| Bull EPS | $70.88 | $227.48 | $289.05 | $229.42 | $182.09 | $144.53 |
| Growth, Years 1–2 (consensus) | +221% | +27% | ||||
| Year-5 EPS as share of Year 2 | Bear 30% · Base 39% · Bull 50% | |||||
Year-5 value and implied return
| US$ | Bear | Base | Bull |
|---|---|---|---|
| Year-5 EPS, share of Year 2 | 30% | 39% | 50% |
| Year-5 EPS (FY2031) | $86.71 | $114.13 | $144.53 |
| Exit P/E | 9.0× | 10.8× | 13.0× |
| Year-5 value | $780 | $1,234 | $1,879 |
| vs. today's price of $885.01 | −12% | +39% | +112% |
| Implied annual return (5-yr CAGR) | −2.5% | +6.9% | +16.2% |
| Present value at 10% a year | $485 | $767 | $1,167 |
| Present value vs. today's price | −45% | −13% | +32% |
IX.Return vs. Today's Price
Pathos illustrative Year-5 (fiscal 2031) value range (bear–bull ends of our assumptions, base = midpoint), with each scenario's implied annual return from today's price. Red line: today's price. Dashed green line: the base-case Year-5 value discounted at 10% a year.
At $885.01 the base case implies about +6.9% a year over five years (−2.5% bear, +16.2% bull), before any dividend. Discounted at a 10% required return, the base Year-5 value is worth $767 today, −13% versus the price. Put the other way: at the base exit P/E of 10.8×, a 10% annual return needs Year-5 EPS of about $131.77; EPS could fall about 23.0% a year in Years 3–5 and still deliver it. Evidence grade: Weak.
X.Superinvestor Holders (Dataroma)
Which of the portfolio managers that Dataroma tracks own SNDK, according to their latest 13F filings? The figures below come only from Dataroma’s SNDK ownership page and activity page.
Source: Dataroma · retrieved 15 Nov 2026 · 13F period Q2 2026 (holdings at 30 Jun 2026)Dataroma shows 1 tracked superinvestor with a position at 30 Jun 2026. The largest weight belongs to Lee Ainslie - Maverick Capital, at 0.01% of the manager’s reported portfolio. Dataroma also shows full exits in Q2 2026: David Tepper - Appaloosa Management (Sell 100%).
XI.Key Risks for Value Investors
NAND pricing reverses
Two-thirds of Q4's sequential growth came from price. In the only downturn in the data, gross margin fell to 7.1% (FY2023), with three straight years of losses. Gross margin was 29.8% as recently as the September 2025 quarter. Industry capacity is growing (Micron, for example, is building new fabs), and Sandisk itself expects higher capital investment in fiscal 2027.
Short, carve-out history
Only fiscal 2026 and part of fiscal 2025 reflect Sandisk as an independent company. Carve-out years include allocated Western Digital costs. $2.5B of goodwill was written off in FY23 and FY25, the second time shortly after the spin-off, when the market valued the company at a fraction of today's level. Any "normalized" estimate rests on thin evidence.
Dependence on Kioxia
All of Sandisk's flash comes from the ventures with Kioxia, in Kioxia-owned fabs in Japan. Sandisk must pay half of their fixed costs even when it cuts purchases (it booked $249M of under-utilization charges in FY24). It guarantees half of their equipment leases ($923M), and the 10-K puts its maximum estimable loss exposure at $2.9B. It cannot build or source flash elsewhere while the ventures run. Kioxia is also a direct competitor.
Long-term agreements cut both ways
NBMs give visibility ($59.8B of remaining obligations plus $31.3B signed after year-end), but they also commit Sandisk to deliver volumes it depends on Kioxia to make. The 10-K warns that Sandisk could face damages or termination if it fails to deliver, that financial guarantees may not fully offset a customer's failure to buy, and that the agreements limit its flexibility to pursue more favourable pricing. Pricing terms and customer names are not disclosed.
Extreme volatility
The stock rose from $112 to a $2,354 intraday high within nine months, fell below $1,000 intraday in late July 2026, and then recovered to about $1,740. Moves of that size can happen again in either direction on a single pricing data point.
Capital allocation near the peak
$4.5B of buybacks at an average $1,600 (~15× book), with $15.5B more authorized. If earnings normalize, repurchases at these prices will have destroyed value. Management states buybacks will be funded from operating cash flow.
Customer concentration and competition
The top ten customers are 44% of revenue. Rivals such as Samsung, SK hynix, Micron, YMTC and Kioxia are larger or state-supported.
Non-operating exposures
The Nanya stake ($1.78B, TWSE-listed, three-year lock-up) adds equity-market and currency risk to earnings. The 10-K estimates a $178M hit from a 10% price decline. Yen movements affect JV costs.
XII.Advice to Management
Pathos's long-term owner principles judge management by the per-share value it builds over many years, not by quarterly results. Applied to Sandisk's figures in this note, we would offer the following recommendations.
Pause repurchases at today's price
All of fiscal 2026's $4.5 billion of buybacks were made in May–July 2026 at an average of about $1,600 a share, roughly 15× book, and $15.5 billion of authorization remains. Average profit over the five years we can see was $1.6 billion a year, against a market value now about 158 times that. Buybacks near a cyclical peak only create value if the peak lasts; we would stop until the price is clearly below a conservative through-cycle value.
Build a cycle reserve for the joint-venture obligations
Sandisk is debt-free, but it guarantees half of Flash Ventures' equipment leases ($923 million), estimates its maximum exposure to the ventures at $2.9 billion and owes Kioxia $1.2 billion over 2026–29. Operating margin was −33% in fiscal 2023. We would hold enough cash to meet these commitments through a downturn before returning more capital.
Keep separating one-offs and prepayments from earnings
Net income includes an $807 million unrealised gain on Nanya shares, and $2.48 billion of operating cash came from customer prepayments; the company's own adjusted free cash flow of $8.74 billion already strips these out. That candour is welcome. We would keep presenting the adjusted figure consistently and explain plainly how prepayments will unwind if prices fall.
Count equity awards at full cost
Diluted shares of about 155 million are well above the 146.4 million outstanding, and the company withheld about 1.0 million shares ($0.6 billion) for employee taxes on vesting awards. We would report the net change in the share count after all awards and repurchases each year, and treat share-based pay as an ordinary expense in headline earnings.
Disclose the terms behind the long-term contracts
Sandisk is signing long-term contracts to lock in today's shortage pricing. Contracts are only as good as their terms in a downturn. We would disclose, in aggregate, the duration, pricing mechanisms and prepayment obligations, so owners can judge how much of today's profit is protected.
Opinion, not advice. These recommendations are Pathos's opinion, based on long-term owner principles and the figures in this note; they are not investment advice.
XIII.Conclusion
The case for. Sandisk is a debt-free NAND leader in an acute AI-driven shortage. Its supply is secured to 2034, about $91B of long-term customer commitments are in place ($59.8B of remaining obligations at year-end plus $31.3B signed since), and it generated $11.5B of free cash flow in its first full standalone year, with quarterly earnings still rising. On forward estimates the shares trade at 8–10× earnings.
The case against. The business lost money in three of the last five years, and its profits now depend on NAND prices that have risen sharply within a year. The standalone record is too short to prove the new contracts change the cycle. Its manufacturing is controlled jointly with a competitor. The stock trades at 11× book and about 105× five-year average earnings, and management has been buying back shares at those levels.
Pathos view (illustrative). On our best-estimate assumptions the base case implies about +6.9% a year (Year-5 range $780–$1,879) and is worth about $767 today at a 10% required return. Evidence: Weak. 21–22 analysts, but the normalisation rests entirely on assumptions and the standalone history is short. Points worth watching: (a) a NAND price collapse when the cycle turns, (b) dependence on the Kioxia joint venture and competitors' capacity, (c) a short standalone record. This is an educational illustration, not a recommendation to buy, sell or hold.
XIV.Sources & Notes
As-of dates. Share price, market capitalisation and all price-based multiples use the Nasdaq closing price of $885.01 on 13 November 2026. The 52-week range, weekly price history and analyst consensus are as retrieved from Yahoo Finance on 15 November 2026. Fundamentals are from the FY2026 Form 10-K (period ended 3 July 2026, filed 17 August 2026) unless stated. Pre-spin-off figures are carve-out financials. The analysts' long-term growth estimate is from Finviz, retrieved 30 September 2026. Ratios, free cash flow and all projection outputs are Pathos calculations.
- Sandisk Corporation, Form 10-K for fiscal year ended 3 July 2026 (filed 17 Aug 2026): financial statements, Flash Ventures (Notes 10–11), NBMs and remaining performance obligations, Nanya investment, WDC share disposals, repurchases (Item 5), subsequent events (Note 17), customer concentration, competitors.
- "Sandisk Reports Fiscal Fourth Quarter and Fiscal Year 2026 Financial Results", Exhibit 99.1 to Form 8-K (5 Aug 2026): Q4 results, end-market revenue, NBM count, adjusted free cash flow, $14B repurchase program, Q1 FY2027 guidance.
- Form 10-K for fiscal year ended 27 Jun 2025 (filed 21 Aug 2025): FY2023–FY2025 figures and FY2025 balance sheet.
- Forms 10-Q for the quarters ended 3 Oct 2025, 2 Jan 2026 and 3 Apr 2026: quarterly results.
- Form 10 information statement, Exhibit 99.1 to Form 10-12B/A (27 Jan 2025): FY2022 carve-out financials and separation background.
- SEC EDGAR XBRL company facts for CIK 2023554: cross-check of annual and quarterly values.
- Form 8-K, 11 Sep 2026: Amendment No. 1 to the loan agreement ($1.5B revolving facility). Form 8-K, 16 Sep 2026: executive compensation adjustments.
- Yahoo Finance, SNDK price history and analyst estimates (retrieved 15 Nov 2026): 13 Nov 2026 close, 52-week range, weekly highs and lows, FY2027 consensus EPS $227.48 (14 analysts) and FY2028 $289.05 (20 analysts), quarterly non-GAAP EPS history (Q1 FY2026 $1.22), FY2027 revenue consensus, mean target $2,105.29 (24 analysts).
- Finviz, SNDK quote and statistics (retrieved 31 Oct 2026): analysts' five-year EPS growth estimate ("EPS next 5Y", 63.55% a year).
- StockAnalysis, SNDK forecast (retrieved 15 Oct 2026; data from S&P Global Market Intelligence): corroborated the close and FY2027 consensus EPS at that date.
- TipRanks, SNDK earnings calendar (retrieved 15 Oct 2026): expected Q1 FY2027 report date (~29 Oct 2026).
- SEC XBRL company facts (SEC EDGAR XBRL API, retrieved 15 Nov 2026): quarterly revenue, net income and diluted EPS from Forms 10-Q and 10-K (fiscal Q4 derived as full year minus nine months), and quarter-end equity, goodwill, intangible assets, preferred stock and shares outstanding. Used for the quarterly charts in Section IV.
- Yahoo Finance, SNDK daily price history (retrieved 15 Nov 2026): split-adjusted quarter-end closing prices for the trailing P/E line.
- Dataroma, SNDK superinvestor ownership page and activity page (retrieved 15 Nov 2026): ownership count, rank, % of all portfolios, hold price, holders, % of portfolio, shares, value and recent activity for 13F period Q2 2026 (holdings at 30 Jun 2026). 13F data are filed up to 45 days after quarter-end and cover long US-listed positions only.
Not verified or not used: pre-FY2022 history of the flash business (not reconstructed); Sandisk's share of Flash Ventures' capital spending in dollars (not disclosed as a single figure); NBM pricing, durations and customer identities (not disclosed); the current market value of the Nanya stake (we use the 3 Jul 2026 fair value); share repurchases since 3 Jul 2026 (the Q1 FY2027 10-Q is not yet filed); the exact date of the 52-week low. The consensus EPS figures are likely non-GAAP. Analysts' five-year growth estimates come from a single aggregator (Finviz) and are not audited. The P/E projection's calculations are kept with the site files (model/pe_projection.py) for audit.