Forget Intel. Buy This Tech Stock in September Instead.

Intel shares rose 10% this week after reports it will increase prices on certain CPUs by 10% in October, but the article cautions that the stock trades at 88x trailing earnings and 57x forward earnings. Intel's adjusted Q3 EPS is expected to decline to $0.38 from $0.41 in Q2 as capex rises, including a $20B 2026 capex outlook. Sandisk, up 615% year-to-date and 2,195% over 12 months, is presented as the more attractive AI-storage exposure at 23x earnings and 8x forward earnings, supported by $93.9B of contracted revenue over the next four to five years.

Analysis

The reported CPU pricing action is not yet company-confirmed and should not be capitalized into INTC estimates without evidence of realized ASPs, unit retention, and mix. At an elevated earnings multiple, even a 10% price increase has limited valuation support if it reflects component-cost pass-through or customer pre-buys rather than durable pricing power; the more important variable is whether gross margin can expand while foundry losses and capex intensity remain contained. Near term, an October price-list implementation can support sentiment, but the 1-3 month catalyst is Q3/Q4 guidance quality and any revision to 2026 capex or subsidy-linked conditions.

SNDK's apparent forward valuation discount is more likely the market pricing memory-cycle normalization than overlooking contracted revenue. Long-duration commitments reduce volume uncertainty but do not eliminate risks around contract pricing, customer concentration, qualification delays, or the eventual return of NAND supply; a falling spot-price environment can compress the multiple before reported earnings weaken. The cleaner second-order beneficiary of sustained storage tightness may be NAND equipment and controller suppliers, while hyperscalers face higher AI infrastructure cost per deployed GPU and may moderate storage-heavy workloads.

Contrarian view: INTC's price move could be more vulnerable than SNDK's if it is treated as proof of an AI-demand inflection despite weak evidence on incremental CPU share. Conversely, SNDK after an extreme run is not a low-risk value trade merely because forward EPS is high: cyclical memory names often screen cheapest at peak earnings. The key spread is durable margin expansion versus a cyclical earnings peak, not headline P/E.

Key Decisions for Investors

• Do not add directional INTC exposure on the reported price increase alone; set an alert for confirmed October pricing plus Q4 gross-margin guidance. Consider a 1-3 month tactical short only if management fails to validate ASP traction or raises 2026 capex again; invalidate on sustained gross-margin upside and credible customer/volume evidence.
• Prefer a risk-defined relative-value expression: long SNDK / short INTC in equal beta-adjusted dollars over the next 1-3 months, contingent on verification of SNDK contract backlog and pricing terms. Thesis is that INTC's multiple embeds execution while SNDK's estimates retain earnings leverage; exit if NAND spot pricing turns decisively lower or INTC guides material margin improvement.
• For existing SNDK longs, reduce gross exposure into earnings and retain upside through call spreads rather than unhedged stock. The 6-18 month risk/reward depends on supply additions and contract renewals; monitor NAND spot prices, utilization announcements from major memory producers, and hyperscaler capex revisions.
• Watch QCOM as a read-through rather than a direct substitute: broader handset/edge-AI price realization would validate industry pricing power, while isolated CPU increases would strengthen the case that INTC is using pricing to offset costs rather than capturing demand-led share gains.