
Buy, target $1,940: the price assumes a bust that our base case does not
Buy, $1,940 target, +141% total return (no dividend). Sandisk makes NAND flash for data-center, edge and consumer storage. At $806 it trades on 3.5x FY27E non-GAAP EPS; the price implies $7.7bn of mid-cycle earnings against our probability-weighted $23.1bn.
- The price discounts a trough: 3.7x implied mid-cycle P/E against our 11.0x. Implied FY29E–FY33E NOPAT is $7.7bn against our probability-weighted $23.1bn (Debate 1).
- Contract floors are worth $410 a share. Eight customers signed $93.9bn of minimum contracts; if renegotiated, base value falls to $1,262 (Debate 2).
- Peak cash is banked first: $66bn of FY27E–FY28E free cash flow, 53% of market value. All excess cash is returned; shares fall from 154m to 122m by FY28E (Debates 3, 4).
What the market is missing. The price needs a 24% mid-cycle operating margin against our 70%. We are 13.5% below consensus on FY28E revenue.
Catalysts. FQ2 results in late Jan 2027 (we expect $12.39bn, guide $12.2–12.6bn); hyperscaler capex guides, Jan–Feb; 8-for-1 split, 20 Nov.
Valuation and target. 11.0x mid-cycle P/E plus FY27E–FY28E cash: bear $784 (35%), base $1,672 (45%), bull $3,271 (20%); weighted $1,681, rolled to $1,940 at 14.0%. DCF: $1,427.
Risks. NBM floors fail (−$410), a deeper price reset (−$200), slower data-center builds (−$158). Conviction is medium: the bear case equals the price.
| Year to June, US$m | FY25A | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|---|
| Revenue | 7,355 | 20,248 | 48,832 | 52,392 | 41,505 |
| Growth y/y | +10.4% | +175% | +141% | +7.3% | −20.8% |
| EBITDA | −1,214 | 12,538 | 39,309 | 42,175 | 30,986 |
| Margin | −17% | 62% | 80% | 80% | 75% |
| Net income | −1,641 | 11,433 | 34,205 | 35,803 | 26,195 |
| EPS, non-GAAP ($) | – | 70.9 | 230.1 | 277.5 | 230.1 |
| EPS growth | – | n.m. | +225% | +20.6% | −17.1% |
| Free cash flow | −120 | 11,494 | 31,570 | 34,805 | 27,123 |
| At $806.43 | FY25A | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|---|
| P/E, non-GAAP | n.m. | 11.4x | 3.5x | 2.9x | 3.5x |
| EV/EBITDA | n.m. | 9.5x | 3.0x | 2.8x | 3.8x |
| P/B | 13.5x | 7.9x | 6.5x | 6.1x | 6.2x |
| FCF yield | −0.1% | 9.3% | 25% | 28% | 22% |
| Dividend yield | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% |
| ROE | −16% | 92% | 197% | 182% | 130% |
| Net debt/EBITDA | n.m. | net cash | net cash | net cash | net cash |
| P/E at our target | n.m. | 27.4x | 8.4x | 7.0x | 8.4x |
Source: Company filings; Penmark Research model (per share pre-split; 8-for-1 split effective 20 Nov 2026, divide by 8). Right: multiples at the 5 Nov 2026 price; only the last row is at our target. ROE on average equity. Shaded columns are estimates.
Base value of $1,672 is 107% above the price; the bear case of $784 is 3% below it
Source: Nasdaq closes via yfinance; Penmark Research model. Scenario values rolled forward at the 14.0% cost of equity to 5 Nov 2027.
Company FY28–30 targets met: prices hold near FY27 to FY30, NBM share 75%; 11.0x on $50.2bn NOPAT.
Floors largely hold; non-NBM price −50%/−35% in FY29–30; 11.0x on $22.8bn NOPAT.
Floors renegotiated, 2023-style bust, slower builds; 11.0x on $8.1bn NOPAT.
| Income, US$m | FY25A | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|---|
| Revenue | 7,355 | 20,248 | 48,832 | 52,392 | 41,505 |
| Gross profit | 2,212 | 14,472 | 41,700 | 44,715 | 33,579 |
| EBITDA | −1,214 | 12,538 | 39,309 | 42,175 | 30,986 |
| EBIT | −1,377 | 12,389 | 39,161 | 41,921 | 30,618 |
| Net financial items | −102 | 628 | 689 | 200 | 200 |
| Tax | −162 | −1,584 | −5,645 | −6,318 | −4,623 |
| Stock compensation | 182 | 232 | 382 | 419 | 444 |
| Net income | −1,641 | 11,433 | 34,205 | 35,803 | 26,195 |
| Cash flow, US$m | FY25A | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|---|
| Net income | −1,641 | 11,433 | 34,205 | 35,803 | 26,195 |
| Operating cash flow | 84 | 11,671 | 32,668 | 36,102 | 28,151 |
| Cash capex | −204 | −177 | −1,099 | −1,297 | −1,027 |
| Free cash flow | −120 | 11,494 | 31,570 | 34,805 | 27,123 |
| FCF margin | −2% | 57% | 65% | 66% | 65% |
| Share repurchases | 0 | −4,524 | −31,332 | −34,805 | −27,123 |
| Debt repayment | −176 | −1,900 | 0 | 0 | 0 |
| Buybacks, % of FCF | n.m. | 39% | 99% | 100% | 100% |
| Balance, US$m | FY25A | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|---|
| Cash | 1,481 | 4,762 | 5,000 | 5,000 | 5,000 |
| Debt | 1,849 | 0 | 0 | 0 | 0 |
| Net cash / (debt) | −368 | 4,762 | 5,000 | 5,000 | 5,000 |
| Equity | 9,216 | 15,736 | 18,991 | 20,409 | 19,925 |
| End diluted shares (m) | 145 | 157 | 139 | 122 | 109 |
| Book value/share ($) | 64 | 100 | 137 | 168 | 182 |
| ROE | −16% | 92% | 197% | 182% | 130% |
| Net cash / equity | −4% | 30% | 26% | 24% | 25% |
| Per share, KPIs | FY25A | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|---|
| Avg diluted shares (m) | 145 | 155 | 148 | 130 | 115 |
| EPS, non-GAAP ($) | – | 70.9 | 230.1 | 277.5 | 230.1 |
| EPS, reported ($) | −11.3 | 73.8 | 230.9 | 274.8 | 226.8 |
| Dividend per share ($) | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 |
| FCF per share ($) | −0.8 | 74.2 | 213.1 | 267.1 | 234.8 |
| Data center, % sales | 13% | 25% | 46% | 52% | 55% |
| Gross margin | 30% | 71% | 85% | 85% | 81% |
| Opex, % of sales | 49% | 10% | 5% | 5% | 7% |
Source: Company filings (FY25A–FY26A); Penmark Research estimates (E). Per share pre-split; 8-for-1 split effective 20 Nov 2026. FY27E non-GAAP EPS excludes stock compensation and equity-security gains. Balance sheet lines are condensed in the forecast. Bold = totals and thesis lines; italic = ratios.
Consensus is close on FY27E but we sit 13.5% below it on FY28E revenue, where we model the price decline
| Growth, margins | FY25A | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|---|
| Growth y/y | |||||
| Sales | +10.4% | +175% | +141% | +7.3% | −20.8% |
| EBITDA | n.m. | n.m. | +214% | +7.3% | −26.5% |
| EBIT | n.m. | n.m. | +216% | +7.0% | −27.0% |
| EPS, non-GAAP | n.m. | n.m. | +225% | +20.6% | −17.1% |
| Margins | |||||
| Gross margin | 30.1% | 71.5% | 85.4% | 85.3% | 80.9% |
| EBITDA margin | −16.5% | 61.9% | 80.5% | 80.5% | 74.7% |
| EBIT margin | −18.7% | 61.2% | 80.2% | 80.0% | 73.8% |
| FCF margin | −1.6% | 56.8% | 64.6% | 66.4% | 65.4% |
| Capex / sales | 2.8% | 0.9% | 2.2% | 2.5% | 2.5% |
| Ratios and valuation | FY25A | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|---|
| Returns and leverage | |||||
| ROE | −16% | 92% | 197% | 182% | 130% |
| ROIC | – | – | 321% | 269% | 178% |
| Net cash / equity | −4% | 30% | 26% | 24% | 25% |
| Valuation at $806.43 | |||||
| P/E, non-GAAP | n.m. | 11.4x | 3.5x | 2.9x | 3.5x |
| EV/EBITDA | n.m. | 9.5x | 3.0x | 2.8x | 3.8x |
| P/B | 13.5x | 7.9x | 6.5x | 6.1x | 6.2x |
| FCF yield | −0.1% | 9.3% | 25% | 28% | 22% |
| Dividend yield | 0.0% | 0.0% | 0.0% | 0.0% | 0.0% |
| P/E at our target | n.m. | 27.4x | 8.4x | 7.0x | 8.4x |
Source: Company filings; Penmark Research estimates. ROE on average equity; ROIC is NOPAT on opening invested capital. Net cash excludes the Nanya stake. Multiples at the 5 Nov 2026 share price, except the last row (at our target).
| US$m / $ | Q2 FY27E | FY27E | FY28E |
|---|---|---|---|
| Revenue, US$m | |||
| Penmark estimate | 12,394 | 48,832 | 52,392 |
| Consensus | 12,510 | 49,560 | 60,590 |
| Prior-year period | 3,025 | 20,248 | 48,832 |
| Diff. vs consensus | −0.9% | −1.5% | −13.5% |
| Growth y/y (Penmark) | +310% | +141% | +7.3% |
| Non-GAAP EPS, $ | |||
| Penmark estimate | 56.95 | 230.1 | 277.5 |
| Consensus | 57.12 | 227.8 | 287.3 |
| Prior-year period | 6.20 | 70.88 | 230.1 |
| Diff. vs consensus | −0.3% | +1.0% | −3.4% |
| Growth y/y (Penmark) | +819% | +225% | +20.6% |
Source: Company filings; Penmark Research estimates; consensus from Yahoo Finance analysis pages, 6 Nov 2026 (16–20 estimates). Q2 FY27E guide: revenue $12.2–12.6bn, EPS $55–57.
Source: Company filings (A); Penmark Research estimates (E); consensus from Yahoo Finance, 6 Nov 2026.
The stock is cheap against its peers’ margins only if prices do not collapse: April 2027 is the test
| Current price | SNDK | Peers | Diff. |
|---|---|---|---|
| Valuation | |||
| P/E, FY27E / CY27E (peers) | 3.5x | 3.1x | +16% |
| P/E, forward | 2.9x | 2.4x | +18% |
| P/E, trailing | 6.7x | 9.8x | −32% |
| Growth and returns | |||
| Revenue growth y/y | 378% | 354% | +24pp |
| Operating margin | 79.6% | 80.0% | −0.5pp |
| ROE | 141% | 90% | +51pp |
| Net debt / EBITDA | net cash | n.m. | – |
| Beta (peer-relevered) | 1.8 | 1.8 | 0.0 |
Source: Penmark Research model Peers tab (Bloomberg CY27E via broker note, 4 Nov 2026); yfinance, 6 Nov 2026. Peers: Samsung, SK hynix, Micron, Kioxia; median of available.
- Nov 208-for-1 split effective; no change to value, target $242 post-split.
- Late JanFQ2 results: we expect $12.39bn revenue and $56.95 EPS, inside guide; Q3 guide is the test.
- Jan–FebHyperscaler CY27 capex guides decide data-center bit growth (Debate 3).
- Late AprFQ3 results and CQ1’27–CQ2’27 contract prints show the price reset depth (Debate 1).
Source: Company filings and calls; Penmark Research.
- NBM floors fail. Customers renegotiate in a downturn. Likelihood 25%; −$163/share (−$410 on the full market-view path).
- Deeper price reset. Non-NBM prices fall 60%/40% in FY29–30. Likelihood 45%; −$143/share (−$200 on the full path).
- Data-center builds slip. Bit growth 12% a year after a policy shock. Likelihood 25%; −$117/share (−$158 on the full path).
- Upside: pricing holds to FY30. Company FY28–30 targets are met. Bull case, 20%; +$1,599 vs base.
Source: Penmark Research model, scenario and single-driver variants. “As an event” values the risk alone against the base case; “on the full path” values the market’s whole assumption set for that driver (Exhibits 81 and 82).
Source: Signals (Micron 10-K key metrics), fiscal year-end price; FY23 (loss) and FY24 (137.8x on trough earnings) not plotted; median FY18–FY26 excluding FY23 is 13.6x. Sandisk has no multiple history (listed Feb 2025).
Source: yfinance, 6 Nov 2026 (forward P/E, operating margin). Includes uncovered Samsung and SK hynix.
The stock traded with revenue until October; the price now discounts earnings of $7.7bn against our probability-weighted $23.1bn
Source: Company filings (quarterly revenue, US$m); Nasdaq closes via yfinance at fiscal quarter-end. FQ3’25 to FQ1’27.
Source: Penmark Research model; consensus from Yahoo Finance, 6 Nov 2026 (19–20 estimates). Revenue, US$bn.
Source: Company filings; Penmark Research model. Bear 35%, base 45%, bull 20%.
Source: Company filings; Penmark Research model. Value per share, pre-split.
Source: Nasdaq closes via yfinance; Penmark Research model. Derived: 3.1x FY26A EPS ($70.88) to 3.5x FY27E ($230.1).
Four debates explain $847 of the $875 gap between price and value; three resolve by April 2027
| Debate | Market view | Penmark view | Evidence | Would change our view | Resolves |
|---|---|---|---|---|---|
| 1 Non-contract price reset | 2023-style bust: FY29–30 −60% / −40% | −50% / −35%, recovery from FY32; worth $1,672 vs $1,472 (+$200) | TrendForce sees supply relief from 2H CY27; October wafer contracts flat; ASP/GB was +248% y/y in 3Q26 | FY28 price guidance below −15% at the FQ3 FY27 results | Q3 FY27 results, late Apr 2027 |
| 2 NBM floors hold | Renegotiated: −40% / −35% / −15% in FY29–31 | Drift to floors: −25% / −25% / −10%; $1,672 vs $1,262 (+$410) | $93.9bn minimum contract value, eight customers; $16.5bn guarantees (18%) | A renegotiation or customer default | Q4 FY27 results, Aug 2027 |
| 3 Data-center build pace | Bits +12% a year after a policy shock | Mid-to-high teens per fab plan; $1,672 vs $1,515 (+$158) | Data center 43.5% of Q1 FY27 revenue, +61% q/q; RPO $88.9bn | Hyperscaler capex cuts | CY27 capex guides, Jan–Feb 2027 |
| 4 Cost per GB | DRAM content inflation stalls cost-down | BiCS8/9 migration, −8% to −11% a year; $1,672 vs $1,594 (+$79) | Cost per GB +5.5% q/q in Q1 FY27 (derived); BiCS9 entering production | Gross margin below 83% | Q2 FY27 gross margin, late Jan 2027 |
Source: Company filings and calls; TrendForce; Penmark Research model Market Check tab. Values per share, pre-split.
Source: Company filings; Penmark Research model. Value per share with one driver at the market’s (bear-case) view versus our base case.
Source: Company calendar; Penmark Research.
| Debate | Section 9 driver | Thesis chapter | Next number to watch |
|---|---|---|---|
| 1 | Non-NBM price change | 10 Floors, not spot, set the earnings base | FY28 non-NBM price −15% (base) |
| 2 | NBM price change | 10 Floors, not spot, set the earnings base | NBM floor near 45% of FY27 average price |
| 3 | Bit shipment growth | 11 Data center broadens the earnings base | Bits +17% in FY28E |
| 4 | Cost per GB (watch item) | 12 Peak cash is banked first | Q2 FY27 gross margin 85.0% |
Source: Penmark Research model.
Debates 1 and 2 carry 72% of the value at stake ($610 of $847). Debate 1 resolves first, on contract prints in 1H CY27 and the April results; Debate 2 resolves with the August results. Debate 4 is the smallest and the likeliest to resolve in our favour in January.
Sandisk turns 80 cents of every revenue dollar into operating profit at today’s NAND prices; the joint ventures carry the fabs
Source: Penmark Research model, FY27E (year to 2 Jul 2027); end-market split from company filings and model. US$m.
Source: Company filings, Q1 FY27 call (29 Oct 2026); Penmark Research model. Q1 FY27 is the latest reported quarter.
What it sells, to whom
NAND flash storage: SSDs, embedded devices, cards, USB drives and wafers, sold into data center, edge (PCs, phones) and consumer markets. Data center is 43.5% of Q1 FY27 revenue, up from 13% in FY25.
How it is priced
Two price books. About half of FY27E bits are sold outside NBM contracts at prevailing contract prices; the other half are NBM contracts with floor prices and prepayments or guarantees. Revenue = bits shipped × price per GB.
Cost and cash
Wafers come from Kioxia joint ventures; Sandisk owns 49.9% of each Flash Ventures entity and its depreciation sits in cost of revenue. Cash capex is 2% of revenue; the rest is funded through JV leases and wafer purchases.
Source: Company filings (10-Q, 10-K), US$m, FQ3’25 to FQ1’27.
Source: Company earnings releases (non-GAAP gross margin); Q2 FY27E from Penmark Research model (guide 84.5%–85.5%).
Two price books drive revenue: the NBM book falls 5% in FY28E while the non-contract book falls 15%
Source: Penmark Research model, FY28E base case. Highlighted boxes are the section 9 drivers.
Source: Company filings; Penmark Research model. Revenue by end market, US$m.
Source: Company filings; Penmark Research model. NBM = contract bits with floor prices; non-NBM = spot-linked bits. US$m.
Source: Penmark Research model, base case. Indices, FY27E = 1.00.
Why the mix matters: contract revenue is 69% of FY28E revenue. Cash profit depends less on the fall in spot prices than on whether the NBM floor holds near 45% of the FY27 average price.
NAND prices rose 248% in the year to FQ3 FY26 (Jan–Mar 2026); the supply response from 2H CY27 sets the price reset
Source: Company 10-K FY26 (Item 1); Q1 FY27 call; Penmark Research. Sandisk owns 49.9% of each Flash Ventures entity; the venture co-terminates 31 Dec 2034.
Source: TrendForce, 18 Aug 2026 (revenue of the five listed NAND suppliers, CQ2’26, US$bn). Five suppliers sum to $68.87bn; YMTC not included.
Source: Signals company KPIs (ASP per GB, y/y growth, %). Latest served quarter is FQ3’26; FQ4’25 not reported.
Source: Company calls (price/volume split, derived by the model for FQ4’26 and FQ1’27); Penmark Research model estimates for FQ2’27E–FQ4’27E. Quarter on quarter, %.
FQ2 guide: revenue $12.2–12.6bn; we model $12.39bn, the midpoint.
FQ3: normally the weakest sequential quarter; we model a dip, then a recovery in FQ4.
CY27 demand: bits up about 40%, data center above 55% of the market.
Price path: quarterly price gains slow from +6% to +2% by FQ4’27E.
Bottom-up size
Five suppliers booked $68.9bn in CQ2’26. Management expects data center above 55% of the market in CY2027 and bits up about 40% after more than doubling in 2026.
Cycle position
TrendForce sees CQ4’26 contract prices up 15–20%, but October wafer contracts were flat and new capacity should ease supply in 2027.
Threats
Supply, not substitution, is the threat. Capex steered to DRAM and HBM keeps NAND tight into 1H CY27; relief from 2H CY27 drives our FY29–30 reset.
Sandisk competes on supply certainty and enterprise SSD, not breadth: its disadvantage is no DRAM or HBM
Source: yfinance, 6 Nov 2026 (latest-quarter operating margin, %). Samsung is a diversified group.
Source: yfinance, 6 Nov 2026 (forward P/E). Kioxia has no forward estimate in yfinance; its trailing P/E is 10.2x.
| Choice criterion | What customers weigh | Sandisk position | Evidence |
|---|---|---|---|
| Cost per bit | Price per GB at scale | Competitive, shared fabs | Cost/GB −6.3% q/q in Q4 FY26, +5.5% in Q1 FY27 on DRAM content (derived) |
| Technology | Layers, density, endurance | Leading edge, BiCS8 ramping | BiCS8 now the majority of bit output (Aug 2026); HBF samples planned CY27 |
| Supply certainty | Committed bits for AI builds | Strong: contracts to 2030 | $93.9bn minimum contract value, eight customers; pipeline exceeds supply |
| Breadth | DRAM plus NAND, HBM | Weak: NAND only | No DRAM or HBM; Samsung, SK hynix and Micron sell both |
| Balance sheet | Funding of capacity | Strong: no debt | $5.1bn cash, no debt at 2 Oct 2026 |
Source: Company 10-K FY26 and Q1 FY27 call; Penmark Research assessment. Competitors named in the 10-K: Samsung, SK hynix, Kioxia, Micron, YMTC and smaller companies.
Source: yfinance, 6 Nov 2026 (quarterly revenue growth, y/y %).
Source: yfinance, 6 Nov 2026 (trailing return on equity, %). HDD makers’ ROE is inflated by small equity bases.
Advantage that lasts: the Flash Ventures partnership and eight-customer contract book are hard to copy before 2030, but both are tied to Kioxia: the venture co-terminates on 31 Dec 2034. Sandisk guarantees part of the venture’s equipment leases; the amount is not in our equity bridge.
| Substitute or threat | Position | Effect on our view |
|---|---|---|
| DRAM and HBM breadth | Samsung, SK hynix and Micron sell both; Sandisk sells NAND only | Capex steered to DRAM and HBM keeps NAND tight into 1H CY27 (TrendForce) |
| New technology | HBF samples planned CY27; SLC may add to NAND demand | Excluded from our model; early (Q1 FY27 call) |
Source: Company 10-K FY26; Q1 FY27 call; TrendForce (Jul–Oct 2026).
Management has committed to return all excess cash: $5.0bn in Q1 FY27 bought back at $1,562, twice today’s price
Source: Company filings (cash flow statements); Penmark Research model (E). US$m. FY27E includes Q1 FY27 actual repurchases of $5,000m.
Source: Q4 FY26 call (5 Aug 2026) guidance; Q1 FY27 reported (29 Oct 2026). Revenue US$bn; margin %; EPS $ pre-split.
Management and record
David Goeckeler is chairman and CEO; Luis Felipe Visoso is CFO. The company separated from Western Digital on 21 Feb 2025 and has beaten its own guidance in Q1 FY27 and six consecutive quarters on EPS to Q4 FY26. Management committed to a financial model to 2030 at its 13 Aug 2026 investor day.
Governance
Sandisk is a standalone Nasdaq issuer. Institutional holders reported 109.4m of 146.4m shares outstanding at 30 Jun 2026 (75% coverage), led by Vanguard (14.8m) and BlackRock (10.3m). The latest insider sales were made under Rule 10b5-1 plans.
Capital allocation
Buybacks have been executed at much higher prices than today. Q1 FY27 repurchases of 3.2m shares averaged about $1,562 (derived), against $806 at the price date. The board added $14bn of authorisation on 5 Aug 2026 and the CFO said the company will stay debt-free and return all excess cash.
What it means for value
Our model buys back at our value, not the market price ($1,765 in FY27E), so the share count falls from 154m to 122m by FY28E. At today’s price the same cash retires more shares: an upside the model does not count.
Source: Company filings; Penmark Research model (period-end diluted shares, pre-split, m).
Source: Company 10-Q Q1 FY27 (average price, derived); Nasdaq close 5 Nov 2026; Penmark Research model repurchase prices (value rolled at cost of equity). Pre-split $.
Sandisk trades on memory pricing and Micron and peer share prices: on results days it moves with Micron, 2.3x as far
The stock trades on the memory price cycle, not on its own earnings beats. Three drivers pass two or more of four tests; six candidates fail. Our model values the two operating drivers at $410 + $200 + $158 = $768 of the $875 gap between value and price; driver C explains why the price fell despite a beat-and-raise on 29 Oct.
| Driver | What moves the stock | Tests | Market reaction vs our model | Debate | Next number |
|---|---|---|---|---|---|
| A Contract price path | Floors on contracts and the non-contract price reset | 3 of 4 | Price −53% since 2 Oct; model value $1,262 at the market’s floors vs $1,672 base | 1, 2 | FQ2 results, Late Jan 2027 |
| B Data-center bits | Hyperscaler build pace; data center 43.5% of sales | 2 of 4 | 26% of analyst questions; $158 of value between market and our path | 3 | Hyperscaler CY27 capex, Jan–Feb 2027 |
| C Peer share prices | Micron and peer share prices, not Sandisk’s own prints | 2 of 4 | Beat-and-raise on 29 Oct fell 13.5% as Micron fell 5.3%; no model value | 1 | Micron forward P/E 3.3x vs SNDK 2.9x |
Source: Penmark Research model (Valuation and Market Check tabs); yfinance; Q1 FY27 call (29 Oct 2026). Tests: event-day, return decomposition, factor sensitivity, call Q&A.
Source: Signals daily prices and yfinance (next-session close-to-close return after each result, %). Q1 FY27 reported 29 Oct 2026; Q1 FY26 (Nov 2025) not in the sample.
Source: yfinance daily returns, 13 Feb 2025 to 6 Nov 2026 (R-squared of SNDK daily return on each factor; weekly sample of 90 gives 0.63 for Micron and 0.18 for the S&P 500).
Source: Nasdaq closes via yfinance, weekly to 5 Nov 2026. Micron rebased to the SNDK price on 3 Nov 2025.
Analysts ask about contracts and data-center demand; they do not ask about consumer or short interest
Source: Signals transcript, Q1 FY27 call, 29 Oct 2026 (19 analyst questions classified by Penmark; one inaudible question excluded). Other calls not tested.
| Candidate dropped | Test result |
|---|---|
| Nasdaq / S&P 500 | R² 0.26 / 0.23 daily; adds 0.001 to Micron alone |
| HDD pricing (WDC, STX) | R² 0.52, but adds 0.06 once Micron is in the model |
| Gross-margin surprise | Correlation +0.56 with the next-day move on six prints; too few to confirm |
| Consumer revenue | 3.6% of Q1 FY27 sales ($397m); 1 of 19 questions |
| Short interest | Days to cover 1.0; 4.7m shares at 15 Oct |
| Insider sales | Rule 10b5-1 plans; no information on timing |
Source: yfinance; Signals; Penmark Research.
A Contract prices set the earnings base
Floors on contracts decide how much of FY27–28 profit survives. The market took the stock from $1,720 to $806 after 2 Oct despite a beat-and-raise on 29 Oct. Our model values the market’s floors at $1,262 against $1,672 (−$410) and the deeper non-NBM reset at $1,472 (−$200). Debates 1 and 2. Next: FQ2 results in late Jan 2027: we expect $12.39bn revenue and $56.95 EPS against consensus $12.51bn and $57.12; the last six prints moved the stock 8.9% on average, either way.
B Data-center bits broaden the base
Data center grew from 13% to 43.5% of sales in two years. The market asks about it more than anything else (5 of 19 questions) and fell as its share rose, so growth is priced as cyclical. Model: bits +17% in FY28E; at +12% a year value is $1,515 (−$158). Debate 3. Next: hyperscaler CY27 capex guides, Jan–Feb 2027; our FY28E bits +17%; consensus not published.
C Peer share prices set the day
Micron explains 58% of daily variance and moved the same way on all six reactions. Sandisk fell 51% in a month while the Nasdaq fell 12%; Micron fell 37%. No model value attaches to a sentiment driver, but it is why the stock sits at 2.9x FY28E forward EPS against Micron at 3.3x. Debate 1. Next: Micron’s forward P/E, 3.3x on 6 Nov; a re-rating of Sandisk to Micron’s multiple is worth about 15% on our estimates.
| Driver | Next number to watch | Date | Penmark | Consensus | Likely stock move |
|---|---|---|---|---|---|
| A Contract prices | Q2 FY27 revenue and non-GAAP EPS | Late Jan 2027 | $12.39bn; $56.95 | $12.51bn; $57.12 | ±9% (mean of last six prints); a Q3 guide below flat is the downside trigger |
| B Data-center bits | Hyperscaler CY27 capex guides; data-center share of sales | Jan–Feb 2027 | FQ2 share above 46%; FY28E bits +17% | Not published | Capex cuts would test the $158 per share between the market’s path and ours |
| C Peer share prices | Micron forward P/E against Sandisk’s | Ongoing | Sandisk 2.9x FY28E forward EPS | Micron 3.3x | Re-rating to Micron’s multiple is about +15% |
Source: Penmark Research model; Yahoo Finance consensus, 6 Nov 2026; yfinance. Limits: six results since the Feb 2025 listing, 21 months of returns, one call tested; prices pre-split. Stock-move ranges are Penmark estimates.
Contract floors are worth $410 a share and the value stays above $1,391 even at the lowest floor setting
Source: Company filings; Penmark Research model (NBM price index, FY27E = 1.00; bear, base and bull cases).
Source: Company filings; Penmark Research model (Valuation tab, contract-floor grid). Probability-weighted value per share, pre-split, with the NBM price index at its FY31E trough from 0.24 to 0.72.
| NBM price index, FY31E trough | 0.24 | 0.32 | 0.40 | 0.48 (base) | 0.56 | 0.64 | 0.72 |
|---|---|---|---|---|---|---|---|
| FY30E gross margin | 60.6% | 68.0% | 73.0% | 76.7% | 79.5% | 81.7% | 83.5% |
| Mid-cycle NOPAT, $m | 11,400 | 15,190 | 18,970 | 22,760 | 26,550 | 30,330 | 34,120 |
| Base-case value, $ | 1,027 | 1,242 | 1,457 | 1,672 | 1,888 | 2,103 | 2,318 |
| Price target, $ | 1,605 | 1,716 | 1,828 | 1,940 | 2,052 | 2,163 | 2,275 |
Source: Penmark Research model (Valuation tab, contract-floor grid); NBM price index with FY27E = 1.00, base case, other drivers unchanged.
1 The claim
Contract floors, not spot prices, set the FY28–FY31 earnings base, and they hold in our base case. Rests on driver A (Debates 1 and 2).
2 What the market believes
The price implies mid-cycle NOPAT of $7.7bn at our 11.0x multiple, below our bear case of $8.1bn. The bear case assumes floors are renegotiated by −40%, −35% and −15% in FY29–31.
3 The evidence
- Contract book: $93.9bn minimum value across eight customers, at floor pricing; $16.5bn of financial guarantees, with $2.5bn held in cash (Q4 FY26 call).
- Pipeline: RPO of $88.9bn at Q1 FY27 and management says the NBM pipeline exceeds supply through the end of the decade.
- Mix: NBMs are about half of FY27 bits and two-thirds in FY28 (Investor Day, 13 Aug 2026).
4 What it is worth
Floors are worth $410 a share. On the market’s floors the base case is $1,262 against $1,672. At the lowest grid point the weighted value is $1,391 and the rating stays Buy.
5 Signposts
- Late Apr 2027: FQ3 results; confirmed if FY28 non-NBM price falls no more than 15%.
- Aug 2027: FQ4 results and NBM renewals; broken if floors are cut.
- Counter-evidence: floor prices are undisclosed; the 45% of average FY27 price we use is implied by the minimum contract value.
Data center is 43.5% of sales and carries bit growth of 17% in FY28E; a build-pace shock costs $158 a share
Source: Company filings; Penmark Research model (end-market revenue, US$m; FY27E includes Q1 actual of $4,802m).
Source: Company filings; Penmark Research model (bit index, FY27E = 1.00; no exabyte volumes are disclosed).
| US$m | FY26A | FY27E | FY28E | FY29E | FY30E |
|---|---|---|---|---|---|
| Data center | 5,153 | 22,463 | 27,244 | 22,828 | 19,672 |
| Edge | 12,160 | 24,416 | 23,576 | 17,432 | 14,403 |
| Consumer | 2,935 | 1,953 | 1,572 | 1,245 | 1,054 |
| Total revenue | 20,248 | 48,832 | 52,392 | 41,505 | 35,129 |
| Data center share of sales | 25.4% | 46.0% | 52.0% | 55.0% | 56.0% |
Source: Company filings (A); Penmark Research model (E), base case. FY27E includes Q1 FY27 actuals.
1 The claim
Data center broadens the earnings base from edge devices to hyperscaler SSDs, so bits grow 14–19% a year even as prices fall. Rests on driver B (Debate 3).
2 What the market believes
Bear-case bit growth is 12% a year after a policy shock to data-center builds. Consensus FY28E revenue of $60.6bn is 16% above our $52.4bn because it assumes no price fall.
3 The evidence
- Reported: data center was $4.8bn in Q1 FY27, up 61% q/q; 43.5% of sales (company filings).
- Industry: TrendForce sees enterprise SSD bits up over 80% in 2026; management sees data center over 55% of the market in 2027.
- Guidance: FY28 bit growth of mid-to-high teens (Q1 FY27 call).
4 What it is worth
Build pace is worth $158 a share. At bear-case growth the base case falls from $1,672 to $1,515. FY28E revenue is $52.4bn in our model and $60.6bn at consensus.
5 Signposts
- Jan–Feb 2027: hyperscaler CY27 capex guides; confirmed by growth above 15%, broken by cuts.
- Late Jan 2027: data-center share of sales above 46% in FQ2.
- Counter-evidence: data-center revenue grew 61% in Q1 yet the stock fell 13.5% on the result: the market treats the growth as a peak.
Explicit FY27–36 cash flows alone are worth $946 a share, 17% above the price, with no terminal value
Source: Company filings; Penmark Research model (unlevered free cash flow, US$m; FY27E is the Q2–Q4 stub).
Source: Company filings; Penmark Research model (gross margin, %; bear, base and bull).
| Base case, US$m | FY27E | FY28E | FY29E | FY30E |
|---|---|---|---|---|
| Operating income | 39,161 | 41,921 | 30,618 | 23,867 |
| Free cash flow (CFO less cash capex) | 31,570 | 34,805 | 27,123 | 21,159 |
| Diluted shares, period end, m | 139.1 | 121.5 | 109.5 | 101.3 |
| Non-GAAP EPS, $ | 230.14 | 277.52 | 230.08 | 197.86 |
Source: Penmark Research model, base case (pre-split per-share data). Cash is held at $5.0bn; surplus buys back shares at our probability-weighted value rolled at the cost of equity.
1 The claim
Peak cash is banked first: $59bn of base-case free cash flow lands in Q2 FY27 to FY28 before prices reset, and all excess cash returns to holders. Rests on driver A and Debate 4 (cost per GB, a watch item).
2 What the market believes
The price implies a 25.6% discount rate on our probability-weighted cash flows, against our 14.0% cost of equity, and a 3.7x mid-cycle P/E against our 11.0x.
3 The evidence
- Reported: Q1 FY27 free cash flow of $6.3bn on $11.0bn of revenue; $5.0bn returned through repurchases; no debt.
- Call: 100% of excess cash returned to shareholders; Q3 normally the weakest sequential quarter (Q1 FY27 call).
4 What it is worth
The explicit FY27–36 cash flows alone are worth $946 a share, 17% above the price, before any terminal value. Cost per GB is worth $79 a share: at a stalled cost-down the base case is $1,594.
5 Signposts
- Late Jan 2027: FQ2 gross margin of 85.0% (guide 84.5–85.5%); confirmed above 83%.
- Late Apr 2027: Q3 revenue −3.4% q/q in our model against a flat consensus.
- Counter-evidence: buybacks at $1,562 in Q1 cost more than today’s $806; our model buys at value, not at the price.
Earnings quality is high: GAAP and non-GAAP EPS differ by 4% or less and free cash flow matched net income in FY26A
Source: Company filings; Penmark Research model (gross and operating margin, %; GAAP).
Source: Company filings; Penmark Research model (diluted EPS, $, pre-split). FY27E GAAP exceeds non-GAAP because $513m of Nanya gains are excluded from non-GAAP.
| Cash conversion | FY26A | FY27E | FY28E | FY29E |
|---|---|---|---|---|
| Free cash flow, $m | 11,494 | 31,570 | 34,805 | 27,123 |
| Net income, $m | 11,433 | 34,205 | 35,803 | 26,195 |
| Free cash flow / net income | 101% | 92% | 97% | 104% |
| Stock compensation, % of revenue | 1.1% | 0.8% | 0.8% | 1.1% |
| Receivable days (derived) | 48 (Q4) | 54 (Q1 FY27) | – | – |
Source: Company filings (A); Penmark Research model (E). Receivable days: receivables over quarterly revenue x 91 (Q4 FY26: $4,708m / $8,965m; Q1 FY27: $6,585m / $11,031m).
| Choice | What it does | Effect on our numbers |
|---|---|---|
| Stock compensation | Treated as a cost in operating income and free cash flow; non-GAAP EPS excludes it | $382m in FY27E, 0.8% of revenue; no flattering |
| Flash Ventures capex | Joint-venture fab spending is mostly lease-funded and sits outside cash capex | Gross capex $2.4bn vs cash capex $1.1bn in FY27E; we deduct cash only, so FY28E free cash flow margin (66%) exceeds the company’s 50% frame |
| Equity gains | Q1 FY27 other income includes a $513m gain on the Nanya stake | Excluded from non-GAAP EPS; not forecast again |
| NBM prepayments | Customer deposits are contract liabilities, not debt | $1.5bn deducted as debt-like in our equity bridge, held flat so not counted twice |
| Refund liabilities | Price-protection and rebate accruals of $1.6bn at Q1 FY27 | Up from $1.5bn; watch for a reversal if prices fall |
Source: Company 10-K FY26, 10-Q Q1 FY27; Penmark Research model (Methodology tab).
We forecast revenue of $48.8bn in FY27E and $52.4bn in FY28E: 1.5% and 13.5% below consensus
Revenue driver tree: see Exhibit 30 (page 8); FY28E revenue is bits up 17% times a blended price down 8%.
Source: Q1 FY27 call (guidance, 29 Oct 2026); Yahoo Finance consensus, 6 Nov 2026 (16–17 estimates); Penmark Research model. Revenue US$bn; non-GAAP EPS $ pre-split.
| Line | Penmark | Consensus | Gap | Reason |
|---|---|---|---|---|
| Revenue Q2 FY27E | $12.39bn | $12.51bn | −0.9% | Q2 at guide midpoint; price +6% q/q, bits +6% |
| Revenue FY27E | $48.83bn | $49.56bn | −1.5% | Q3 bits −8% on seasonality; prices +5% / +2% in Q3 / Q4 |
| Revenue FY28E | $52.39bn | $60.59bn | −13.5% | Non-NBM price −15% from 2H CY27 supply relief; consensus assumes no fall |
| Non-GAAP EPS FY27E | $230.14 | $227.80 | +1.0% | Higher margin; our buyback share count (148m average) |
| Non-GAAP EPS FY28E | $277.52 | $287.30 | −3.4% | Lower revenue offset by a 130m share count |
| EPS FY28E, 151m sh. | $239.50 | $287.30 | −16.6% | Share-count-neutral comparison |
Source: Penmark Research model; Yahoo Finance consensus (6 Nov 2026). Consensus EBITDA is not published and is not shown.
Source: Company filings; Penmark Research model. US$m; opex = gross profit less operating income.
| Costs and cash | FY27E | FY28E | FY29E | Reasoning |
|---|---|---|---|---|
| Gross margin | 85.4% | 85.3% | 80.9% | Cost per GB −8% to −11%; price falls outweigh from FY29E |
| Operating expenses, % of revenue | 5.2% | 5.3% | 7.1% | Mostly fixed: R&D $1.3bn; rises as revenue falls |
| Operating margin | 80.2% | 80.0% | 73.8% | Guided about 75%; base is above in FY28E, below in FY30E |
| Cash capex, $m | 1,099 | 1,297 | 1,027 | Gross capex 5% of revenue; 45% cash-funded |
| Net working capital, $m | 5,127 | 5,501 | 4,358 | Receivables grow with NBM terms; release in the reset |
Source: Penmark Research model (Financial statements and Revenue build tabs), base case.
Two methods value Sandisk at $1,681 and $1,427 a share; the price implies a 25.6% discount rate
| Primary: mid-cycle P/E plus cash | Bear 35% | Base 45% | Bull 20% | Weighted |
|---|---|---|---|---|
| Mid-cycle NOPAT, FY29E–33E average, $m | 8,076 | 22,759 | 50,244 | 23,117 |
| Value of business at end-FY28E (11.0x), $m | 88,832 | 250,352 | 552,680 | – |
| PV of mid-cycle value, $m | 70,645 | 199,097 | 439,529 | – |
| PV of Q2 FY27–FY28E cash, $m | 44,559 | 52,988 | 58,729 | 51,186 |
| Operating equity value, $m | 115,205 | 252,085 | 498,258 | – |
| Equity value incl. $5,465m non-operating assets, $m | 120,670 | 257,550 | 503,722 | 258,876 |
| Value per share, $ (154m shares) | 784 | 1,672 | 3,271 | 1,681 |
| Cross-check: full-cycle DCF per share, $ | 726 | 1,445 | 2,610 | 1,427 |
Source: Penmark Research model (Valuation tab), valued at 2 Oct 2026. Per-share values pre-split; DCF at 14.0% cost of equity, 3.0% terminal growth. PV of cash is stub FCF plus FY28E FCF.
Source: Company filings; Penmark Research model (Market Check tab). Mid-cycle NOPAT, FY29E–FY33E average, US$bn; price-implied figure holds the 11.0x multiple.
| Reverse DCF: what $806 implies | Price | Penmark |
|---|---|---|
| Price-implied mid-cycle P/E (NOPAT $23.1bn) | 3.7x | 11.0x |
| Price-implied discount rate on weighted cash flows | 25.6% | 14.0% |
| Price-implied FY29E–33E operating margin | 23.6% | 69.6% |
| Explicit FY27E–36E cash flows per share (no terminal value) | $946 | vs price $806 |
Source: Penmark Research model (Market Check tab); reverse DCF solved at the model terminal growth of 3.0%.
Why these methods
Value depends on how far prices fall after supply catches up and how much peak cash is banked first. The mid-cycle P/E averages trough and recovery earnings (FY29E–33E) at 11.0x, below the 13.6x median of Micron’s FY18–FY26 year-end P/E (Exhibit 15). Rejected: forward P/E on peak earnings, EV/EBITDA (JV depreciation sits in wafer cost), P/B and sum of the parts.
Why the methods differ by 17.8%
The gap exceeds 15% because the DCF capitalises less. Our base-case terminal value is $250bn on FY36E NOPAT of $30.4bn, about 8.2x (derived), against the 11.0x applied at end-FY28E. Terminal value is 32% of DCF enterprise value; mid-cycle value is 79% of the base-case operating equity. Both methods rate the stock Buy.
| Equity bridge at 2 Oct 2026 | US$m |
|---|---|
| Cash and Nanya stake at fair value | 5,136 + 2,289 |
| Less NBM prepayments (debt-like) | −1,529 |
| Less non-current tax payable; no debt | −431 |
| Net non-operating assets, added to operating equity value | 5,465 |
Source: Company 10-Q Q1 FY27; Penmark Research model (Valuation tab). $5,465m / 154m shares (derived).
Value stays between $1,258 and $2,157 across 63 multiple and discount-rate cases; the DCF never falls below $1,096
Source: Company filings; Penmark Research model (Valuation tab). Probability-weighted value per share, $ pre-split; cost of equity 10–18% by row, mid-cycle P/E 8–14x by column.
Source: Company filings; Penmark Research model (Valuation tab). Probability-weighted DCF value per share, $ pre-split; cost of equity by row, terminal growth by column.
| Case | Prob. | Value | Key assumptions | What it needs |
|---|---|---|---|---|
| Bear | 35% | $784 | Floors renegotiated (−40/−35/−15%); non-NBM −60/−40%; bits +12% a year; cost-down stalls | Debates 1–4 all go the market’s way |
| Base | 45% | $1,672 | Floors −25/−25/−10%; non-NBM −50/−35%; bits +14–19%; cost −8% to −11% | Supply relief from 2H CY27 |
| Bull | 20% | $3,271 | Company FY28–30 targets met: mid-to-high-teens growth, ~80% gross margin | Prices hold to FY30 |
| Weighted | 100% | $1,681 | Target $1,940; $242.48 post-split | +140.5% expected return |
Source: Penmark Research model (Valuation and Scenarios tabs). Probabilities are analyst judgement, not statistical estimates; bear and base values tie to the four debates on page 6.
Source: Company filings; Penmark Research model (Valuation tab). Weighted value of $1,681 at 2 Oct 2026 rolled 1.09 years at 14.0%; no dividend. Pre-split $.
Six dated events decide the call: the first is late January 2027, and failed contract floors are the biggest risk to the thesis
| Date | Event | Penmark expects | Consensus | Value per share (probability) |
|---|---|---|---|---|
| Late Jan 2027 | FQ2 results at the guide midpoint, Q3 guide holds price gains | $12.39bn; $56.95 EPS | $12.51bn; $57.12 | +$57 (60%) |
| Jan–Feb 2027 | Hyperscaler CY27 capex guides (data-center bits) | FY28E bits +17% | Not published | −$117 if cut (25%) |
| CQ1’27–CQ2’27 | Contract price prints: October NAND wafer prices flat | Non-NBM −15% in FY28E | FY28E revenue +22% | +$74 / −$143 |
| Next 2–3 quarters | More NBMs; second tranches for existing customers | NBM share of bits 67% in FY28E | Not modelled | +$55 (50%) |
| Late Apr 2027 | FQ3 results: seasonally weakest quarter; FY28 price signal | Revenue −3.4% q/q | $12.48bn (−0.2% q/q) | Rating test |
| Aug 2027 | FQ4 results and NBM renewals or extensions | Floors drift −25% / −25% / −10% | Not modelled | −$163 if floors fail (25%) |
Source: Company calendar; Penmark Research model (Thesis tab); Yahoo Finance consensus, 6 Nov 2026. Probabilities and values are model estimates, pre-split.
Source: Company filings; Penmark Research model (Thesis tab). Event-specific value change per share, $ pre-split, one driver at a time. “As an event” values the risk alone against the base case; “on the full path” values the market’s whole assumption set for that driver.
| Risk to the Buy | Likelihood | Value / share | Evidence |
|---|---|---|---|
| NBM floors fail | 25% | −$163 | Guarantees cover 18% of $93.9bn minimum value |
| Deeper non-NBM price reset | 45% | −$143 | 2023-style bust: −60% / −40% |
| Data-center builds slip | 25% | −$117 | Policy shock; memory names fell 15–22% on 5 Nov |
| DRAM cost inflation | 40% | −$39 | Cost per GB +5.5% q/q in Q1 FY27 |
Source: Penmark Research model (Thesis tab). Likelihoods are analyst judgement. “As an event” values the risk alone against the base case; “on the full path” values the market’s whole assumption set for that driver.
What would end the thesis
Contract floors cut, or non-contract prices falling faster than −15% for FY28. Either would show in FQ3 results (late Apr 2027). On the bear path the base-case value falls to $1,262 and the price target to $1,456; the rating stays Buy, but conviction falls.
Regulatory, legal and key-person
Sandisk guarantees part of Flash Ventures’ equipment leases and depends on Kioxia for all wafers; the amount is not in our equity bridge. US-China export policy and data-center permitting rules are macro exposures. David Goeckeler is chairman and CEO. ESG is not a value driver here.
Institutions hold most of the register, shorts have halved, and the Street’s median target is 2.5x the price
Source: Signals 13F data, quarter to 30 Jun 2026 (reported 109.4m shares, 75% coverage); shares outstanding at 17 Aug 2026. Shares, m.
Source: Signals, FINRA short interest (settlement dates 15 Jun to 15 Oct 2026); days to cover 1.0 on each date. Shares, m.
Source: StockAnalysis (S&P Global), 25 analysts, 3 Nov 2026 (mostly set before the 5 Nov fall); Penmark Research. $ pre-split.
Source: Ticker Nerd, 25 analysts, 16 Oct 2026; one Buy-to-Hold change on 2 Nov 2026 (Fox Advisors) is not included.
| Holder | Shares, m | % of 146.4m | Value at 30 Jun 2026 |
|---|---|---|---|
| Vanguard | 14.76 | 10.1% | $33.6bn |
| BlackRock | 10.30 | 7.0% | $23.4bn |
| FMR LLC | 7.86 | 5.4% | – |
| State Street | 5.93 | 4.0% | – |
| Invesco | 5.61 | 3.8% | – |
Source: Signals 13F data, quarter to 30 Jun 2026; % of shares outstanding at 17 Aug 2026 (derived). Value shown where reported.
| Liquidity and insiders | Latest |
|---|---|
| Average daily value traded, 63 sessions | $16.0bn |
| Insider sales, Sep–Nov 2026 | All under Rule 10b5-1 plans (CEO, 17 Sep; EVP Shek, 1 Oct and 2 Nov) |
| Largest recent sale | 600 shares at $1,158.26, 2 Nov 2026 |
| Listing | Nasdaq Global Select Market since 24 Feb 2025 |
| Nanya stake held by Sandisk | 139m shares, ~$2.3bn at 2 Oct 2026; three-year lock-up from 2026 |
Source: Signals (insider transactions, short interest); yfinance (63-session average daily value); company filings.
What it means: low short interest and 10b5-1 sales give no signal; the fall since 2 Oct fits driver C.
The model shows revenue peaking at $52.4bn in FY28E, free cash flow of $34.8bn, and a share count down 36% by FY30E
| US$m, FY ends Jun/Jul; pre-split | FY25A | FY26A | FY27E | FY28E | FY29E | FY30E | FY31E |
|---|---|---|---|---|---|---|---|
| Income statement | |||||||
| Revenue | 7,355 | 20,248 | 48,832 | 52,392 | 41,505 | 35,129 | 35,281 |
| Growth y/y | – | 175% | 141% | 7.3% | −20.8% | −15.4% | 0.4% |
| Gross profit | 2,212 | 14,472 | 41,700 | 44,715 | 33,579 | 26,946 | 26,813 |
| EBITDA | −1,214 | 12,538 | 39,309 | 42,175 | 30,986 | 24,308 | 24,098 |
| Operating income | −1,377 | 12,389 | 39,161 | 41,921 | 30,618 | 23,867 | 23,610 |
| Net income | −1,641 | 11,433 | 34,205 | 35,803 | 26,195 | 20,457 | 20,239 |
| Diluted EPS, GAAP ($) | −11.32 | 73.76 | 230.89 | 274.79 | 226.81 | 194.14 | 206.91 |
| Non-GAAP EPS ($) | – | 70.88 | 230.14 | 277.52 | 230.08 | 197.86 | 211.09 |
| Cash flow | |||||||
| Cash flow from operations | 84 | 11,671 | 32,668 | 36,102 | 28,151 | 22,029 | 21,191 |
| Cash capex (incl. JV investment) | −204 | −177 | −1,099 | −1,297 | −1,027 | −869 | −953 |
| Free cash flow | −120 | 11,494 | 31,570 | 34,805 | 27,123 | 21,159 | 20,238 |
| Share repurchases | 0 | 4,524 | 31,332 | 34,805 | 27,123 | 21,159 | 20,238 |
| Balance sheet and shares | |||||||
| Cash | 1,481 | 4,762 | 5,000 | 5,000 | 5,000 | 5,000 | 5,000 |
| Total equity | 9,216 | 15,736 | 18,991 | 20,409 | 19,925 | 19,684 | 20,165 |
| Debt | 1,849 | 0 | 0 | 0 | 0 | 0 | 0 |
| Diluted shares, period end (m) | 145.0 | 157.2 | 139.1 | 121.5 | 109.5 | 101.3 | 94.4 |
Source: Company filings (A); Penmark Research model (E, shaded), base case. FY25A non-GAAP EPS not in model; FY25A debt of $1,849m repaid in FY26.
| Ratios at the 5 Nov 2026 price | FY25A | FY26A | FY27E | FY28E | FY29E | FY30E | FY31E |
|---|---|---|---|---|---|---|---|
| Ratios | |||||||
| Gross margin | 30% | 71% | 85% | 85% | 81% | 77% | 76% |
| Operating margin | −19% | 61% | 80% | 80% | 74% | 68% | 67% |
| FCF margin | −2% | 57% | 65% | 66% | 65% | 60% | 57% |
| ROE (average equity) | – | 92% | 197% | 182% | 130% | 103% | 102% |
| Net debt / EBITDA | n.m. | net cash | net cash | net cash | net cash | net cash | net cash |
| P/E at $806.43, non-GAAP | – | 11.4x | 3.5x | 2.9x | 3.5x | 4.1x | 3.8x |
Source: Penmark Research model; ROE on average equity; P/E on non-GAAP EPS at $806.43. Net cash from FY26A (debt $0); FY25A net debt of $368m, ratio not meaningful on negative EBITDA.
| US$m; pre-split | FQ2’26A | FQ3’26A | FQ4’26A | FQ1’27A | FQ2’27E | FQ3’27E | FQ4’27E |
|---|---|---|---|---|---|---|---|
| Revenue | 3,025 | 5,950 | 8,965 | 11,031 | 12,394 | 11,973 | 13,434 |
| Data center revenue | 440 | 1,467 | 2,977 | 4,802 | – | – | – |
| Gross margin | 50.9% | 78.4% | 84.6% | 85.0% | 85.0% | 85.4% | 86.0% |
| Operating income | 1,065 | 4,111 | 7,037 | 8,780 | 9,915 | 9,583 | 10,882 |
| Net income | 803 | 3,615 | 6,903 | 8,101 | – | – | – |
| Non-GAAP EPS ($) | 6.20 | 23.41 | 39.25 | 49.51 | 56.95 | – | – |
| Free cash flow | 980 | 2,993 | 7,083 | 6,309 | – | – | – |
Source: Company filings (A); Penmark Research model (E, shaded). FQ2’27E EPS is the model non-GAAP figure; EPS, net income and cash flow shown for reported quarters only. FQ4’26A derived as full year less Q1–Q3.
Adjusted EPS of $49.51 reconciles to GAAP $52.60 in Q1 FY27 through a $513m equity gain and $91m of stock compensation
| Q1 FY27A, US$m | GAAP | Adjustment | Non-GAAP / adjusted |
|---|---|---|---|
| Operating income | 8,780 | +91 (stock compensation) | 8,871 |
| Operating margin | 79.6% | +0.8pp | 80.4% |
| EPS, diluted ($) | 52.60 | −3.09 (excludes $513m Nanya gain, adds back stock compensation) | 49.51 |
| Free cash flow | 6,309 | −613 (excludes deposits and Flash Ventures activity) | 5,696 |
Source: Company earnings release and 10-Q, Q1 FY27 (29 Oct 2026); Penmark Research model (Historicals tab).
Glossary
NBM: the company’s multi-year supply agreement model, with committed bits, floor pricing and customer deposits. Non-NBM: bits sold at prevailing contract prices. Bit / exabyte: units of NAND capacity shipped; no volumes are disclosed, so our bit index sets FY27E = 1.00. Flash Ventures: Kioxia joint ventures (Sandisk 49.9%) that make all Sandisk wafers. BiCS8 / BiCS9: NAND node generations. HBF: high-bandwidth flash, samples planned CY27, excluded from our model. TLC / QLC / SLC: 3-, 4- and 1-bit-per-cell NAND. Mid-cycle NOPAT: FY29E–FY33E average after-tax operating profit.
Ratio definitions
P/E: price over non-GAAP EPS. EV/EBITDA: market value less net non-operating assets over EBITDA (operating income plus own D&A). ROE: net income over average equity. FCF: cash from operations less cash capex. Net debt / EBITDA: debt less cash; Sandisk has no debt. Price per share is pre-split unless stated; divide by 8 for the 20 Nov 2026 post-split basis. Total return: target over price less one, no dividend.
Peer set
Samsung Electronics, SK hynix, Micron and Kioxia are the four listed NAND producers; YMTC is private. HDD makers Western Digital and Seagate are shown for margin and valuation context only. CY27E P/E: Samsung 2.8x, SK hynix 2.6x, Micron 4.8x, Kioxia 3.3x (median 3.05x).
Methodology note
Revenue equals bits shipped times price per GB, split by NBM and non-NBM contract. Primary valuation is 11.0x mid-cycle P/E on FY29E–FY33E NOPAT plus the present value of Q2 FY27–FY28E cash; the cross-check is a full-cycle DCF to FY36E at 3.0% terminal growth and a 14.0% cost of equity. Notes: no own-multiple history (listed Feb 2025), so Micron’s through-cycle P/E stands in; consensus EBITDA is not published, so the estimates comparison uses revenue and EPS; the NBM floor level is undisclosed and implied by the $93.9bn minimum contract value.
Rating basis. Buy applies when probability-weighted total return to the target exceeds +10%; here it is +140.5%. The rating holds at cost of equity ±1 point, at the lowest NBM floor grid point and on base-only valuation.
Data sources and as-of dates
Company 10-K FY26 (filed 17 Aug 2026), 10-Q and earnings call for Q1 FY27 (29 Oct 2026), Investor Day (13 Aug 2026); Signals (ownership, short interest and insider data to 15 Oct–3 Nov 2026); yfinance (prices to 6 Nov 2026; peer ratios as of that date); TrendForce (Aug–Oct 2026); Yahoo Finance consensus (6 Nov 2026); Penmark Research valuation model (price date 5 Nov 2026).
Analyst certification and important disclosures
Ratings
Ratings reflect the expected total return to our 12-month price target, including dividends, weighted across bull, base and bear scenarios.
| Buy | Expected total return above +10% |
| Hold | Expected total return between −10% and +10% |
| Sell | Expected total return below −10% |
We review ratings at each set of results and whenever the facts change materially. A share price can move outside its band before the rating changes, and targets may not be reached.
Analyst certification
The Penmark Team certifies that (1) the views in this report accurately reflect its views about the subject company and its securities, and (2) no part of its compensation was, is or will be directly or indirectly related to the specific ratings, price targets or views in this report.
Contact
Penmark Research
Questions about our research:
research@penmarkresearch.com
All our research and coverage:
penmarkresearch.com
Basis of preparation
This report uses information publicly available up to its publication date. Share prices are closing prices on the trading day before the publication date, unless stated otherwise. Forecasts, valuations and price targets are Penmark Research’s own, and forecast periods are marked E. Sources are shown beneath each exhibit.
Valuation and price targets
Price targets are based on the valuation methods set out in each report and are weighted across bull, base and bear scenarios. The main risks to each rating and price target are described in the report. Penmark Research updates its coverage after each set of results but is not obliged to update a report at any other time.
Independence and conflicts of interest
Penmark Research is independent. It does not provide investment banking, brokerage, market-making or asset management services, does not trade securities, and does not accept payment from any company for research coverage. Companies do not see or review Penmark Research reports before publication.
No member of the Penmark Team holds a long or short position, directly or through derivatives, in the securities of any company Penmark Research covers.
Regulatory status
Penmark Research publishes general, impersonal investment research on a regular basis. It is not registered as an investment adviser with the US Securities and Exchange Commission or any state securities authority, is not a registered broker-dealer and is not a member of FINRA. Penmark Research does not give advice tailored to any person’s circumstances.
Not investment advice or an offer
This report is for information only. It is not personalised investment advice and is not an offer or solicitation to buy or sell any security or other financial instrument. It does not take into account any reader’s objectives, financial situation or needs. Readers should make their own investment decisions and seek independent professional advice where appropriate.
Forward-looking statements and risk
Forecasts, valuations and price targets are estimates based on assumptions that may prove wrong, and they may not be achieved. Opinions are current as of the publication date and may change without notice. Investing in equities involves risk, including the loss of capital. Share prices can be volatile, and past performance is not a reliable indicator of future results.
Data and sources
Company financials, operating data and filings come from company filings, earnings releases and earnings calls. Market data and any consensus estimates come from third-party sources named beneath the relevant exhibit; consensus figures from aggregators are unofficial. Penmark Research does not use other brokers’ research. Third-party data belong to their providers and may not be extracted from this report.
Accuracy and liability
Information is taken from sources believed to be reliable, but Penmark Research makes no representation or warranty as to its accuracy, completeness or timeliness. To the fullest extent permitted by law, Penmark Research accepts no liability for any loss arising from use of this report. Descriptions of litigation and regulatory matters reflect allegations and public records, not findings, unless stated. If a published report is corrected, the corrected version carries an “Updated” date at penmarkresearch.com.
Distribution and access
Penmark Research reports are available to registered users of penmarkresearch.com. Downloads and programmatic access are governed by the Penmark Research subscription terms. This report is not directed at, or intended for, any person in Singapore or India, and access from those countries is blocked. It is not intended for distribution to, or use by, any person in any jurisdiction where that would be contrary to law or regulation or would subject Penmark Research to registration or licensing requirements there.
Copyright and permitted use
© Penmark Research. All rights reserved. Penmark Research is a trade name of Brightriver Labs. Readers may share a link to this report but may not copy, reproduce, redistribute, sell or use its content, including for text and data mining, without written consent. Contact: research@penmarkresearch.com.