← SNDK forward model

SNDK · Forward model · Consumer · NBM Floor case

What has to happen in Consumer

Model as of

This page changes Consumer inside the complete SNDK model. Every other vertical, corporate programme and valuation assumption stays in the selected portfolio case.

SNDK forward model
Horizon
Consolidated fair value $726.22 all other verticals held in this portfolio case
Final-quarter revenue $117M 2% of company revenue
Explicit segment contribution $3.53B EBITDA less segment capex, before corporate items

The one case with a contract under it. Sandisk states a minimum of $93.9bn of contracted NBM revenue computed at floor pricing across eight customers, weighted average duration above four years, and says the economics carry attractive margins even at floor pricing. This case runs the spot market to no premium at all and lets only the contracted book carry the business: revenue converges toward the low twenties of billions a year against a $48.96bn consensus for FY2027 alone. What it does not reach: it says nothing about the roughly half of FY2027 bits and third of FY2028 bits that are not under NBM, and the floor applies to the variable pricing component only, so the realised floor depends on a fixed/variable split Sandisk has not disclosed.

Consumer

Basis quarter$556M
Final quarter$117M
Implied CAGR−27%
Final revenue mix2%

Retail cards, USB drives and consumer SSD - 6.2% of basis-quarter revenue and shrinking, because it is the residual claimant on the bit pool. FY2026 is the tell: Consumer exabytes fell mid-teens percent while revenue rose 29% on a low-fifties percent price increase, then the line fell 32% sequentially in a quarter the company grew 51%, as bits were pulled into Datacenter. Management calls it mix flexibility, and this is the line that gets sacrificed.

Last four quarters
2026 Q1 $652M Reported
2026 Q2 $907M Reported
2026 Q3 $820M Reported
2026 Q4 $556M Reported
Memory cardsUSB flash drivesConsumer and portable SSD
Bit index (2026 Q4 = 100) 100/qtr growing −1.5% per quarter Index, not exabytes: 2026 Q4 = 100. Sandisk publishes Consumer bit growth, never a bit level.
Revenue per index point $6M drifting +14.0% per quarter Basis revenue $556M / 100 index points. Derived; the basis quarter was itself a 32% sequential fall.
Consumer

Latest: $117M (2031Q4E)

Period Value
2024Q3 $573M
2024Q4 $523M
2025Q1 $514M
2025Q2 $598M
2025Q3 $571M
2025Q4 $585M
2026Q1 $652M
2026Q2 $907M
2026Q3 $820M
2026Q4 $556M
2027Q1E $585M
2027Q2E $585M
2027Q3E $566M
2027Q4E $535M
2028Q1E $498M
2028Q2E $459M
2028Q3E $420M
2028Q4E $383M
2029Q1E $348M
2029Q2E $316M
2029Q3E $286M
2029Q4E $259M
2030Q1E $235M
2030Q2E $213M
2030Q3E $192M
2030Q4E $174M
2031Q1E $157M
2031Q2E $142M
2031Q3E $129M
2031Q4E $117M

Assumptions & reasoning

  • Units are an INDEX, not exabytes: the basis quarter is 100 and price per index point is that quarter's revenue divided by 100. Sandisk discloses bit GROWTH by end market and has never published an exabyte LEVEL, so a physical unit count here would be an invented disclosure.
  • This is the vertical where a holiday season is visible, and it is still modelled aseasonal: the research brief measured a window-to-window spread of 0.274 against an FY2025 within-year amplitude of 0.15, with only two fiscal-year windows and one of them a 53-week year. An aseasonal path is modestly wrong in a repeating way; four factors fitted to two disagreeing windows would be wrong permanently.
  • At 6.2% of basis-quarter revenue a 10% error in this line moves consolidated revenue by 0.6%, so the model spends no further complexity here.
SNDK model map

Explore another vertical