EOD Wrap; Sandisk (SNDK): what does mgmt's MT framework and buybacks mean for EPS?

Good afternoon. Apologies for the delay on this one — was running around after the close. QQQs +1.1% as yields dropped 4-7bps across the curve and market now pricing in only 23bps worth of rate hikes. Oil -2%. Overall, solid breadth continues in Tech and the factor whiplash days seem to be behind us now. Despite the WDAY +18%/Silverlake buyout rumors which had app sw/”AI losers” rallying, still plenty of semis up on the day including SNDK +14% after putting out some better mid-term growth numbers at their analyst day. Apparently the best AI monetization strategy is getting bought by PE. After eight months of existential debate, software is now back to flat on the year! Overall, the market continues to feel very strong as we make our way through August.

---

Post-close, AMAT -4.5% looks fine with solid numbers and strong Oct guide, but below a very elevated buyside bar.

Modest F3Q beat and guided F4Q revenue ~8% above the Street, with EPS also comfortably ahead. More importantly, mgmt said customer visibility has improved and expects another strong growth year in 2027, supporting the longer-term WFE/semicap bull case. Buyside was closer to $10.5B for the Oct guide after LRCX’s much bigger upside, so $10.25B probably doesn’t push buyside #s higher. Net/net, nothing thesis-changing negative.

Key KPIs vs Street

• F3Q Revenue: $9.12B, +25% y/y vs. Street ~$9.0–9.02B, +23.5% y/y; guide was $8.95B. Buyside was looking for more like ~$9.5B.

• Gross Margin: 50.4% vs. Street ~50.0% and 50.0% last qtr.

• EPS: $3.50 vs. Street ~$3.40–3.42.

• F4Q Revenue Guide: $10.25B, +12.5% q/q / ~+41% y/y vs. Street ~$9.52–9.60B; buyside bogey ~$10.5B.

• F4Q EPS Guide: $4.02 vs. Street ~$3.69–3.72.

• China: 28% of revenue vs. ~27% last qtr.

---

We’ll focus most of our time on SNDK recap then cover the rest. Let’s get to it…

AI/SEMIS

• SNDK +14% the big winner today after mgmt came through at the analyst day (full slide deck here), helping put the focus back on “stronger for longer" & buyback NAND thesis which bulls are holding on to. Through FY30, management is targeting mid/high-teens revenue growth, ~80% GMs, ~75% operating margins and ~50% adjusted FCF margins, with capex staying around a mid-single-digit % of revenue. The ~80% GM was broadly in the range investors expected, but still important given some bears were looking for something below that; the revenue outlook is probably incremental as street was only expecting growth of +20%/-17%/-28% y/y for FY28E/29E/30E. Mgmt also sounded more constructive on NAND pricing, seeing 2027 flat-to-up versus the 15–20% decline some investors have been modeling. SNDK announced they have eight NBMs, three with U.S. hyperscalers, with the longest contracts running 4–5 years and roughly 2/3 of FY28 capacity is already spoken for through NBMs, with FY29 potentially looking similar, while customers have recently been coming back for additional volume. Importantly: Mgmt formally committed to returning 100% of excess FCF after reinvestment and maintaining a debt-free balance sheet, primarily through buybacks. On HBF, a little light on detail but said HBF has now taped out, with samples expected in 2027 and a potential 2028 volume ramp, and management is talking about HBM-class bandwidth at roughly 1/8 the cost with 8–16x the capacity

  

  

  

  • CY27 buyside numbers not really changing here (still at ~$300+). Just taking their model at face value mechanically produces something close to $700 in CY30 off of a CY27 street EPS number (mid-to-high-teens revenue growth, ~80% gross margin, ~75% operating margin, ~50% adjusted-FCF margin) and deploy 100% of the cash into the stock a@$1,500. It’s not like anyone is going out that far to value SNDK given the open question right now is how long does the cycle continue past 2028, but it provides a good framework to think about valuation.