September Effect: 2 Memory Stocks to Buy on a Potential Market Dip

The article flags a potential September-October equity pullback in a midterm-election year, noting the S&P 500 declined at least 5% in 15 of the past 24 such cycles. It identifies Micron and Sandisk as potential dip-buy candidates, each trading at forward P/E multiples below roughly 6, supported by AI-driven memory demand and constrained DRAM/NAND supply. Micron benefits from tight DRAM supply as HBM consumes roughly 3x the wafer capacity of standard DRAM, while Sandisk targets mid- to high-teens revenue growth and about 80% adjusted gross margin between fiscal 2028 and 2030.

Analysis

The investable distinction is not simply “AI memory”: MU has meaningful operating leverage to broad DRAM pricing, while SNDK is a higher-beta expression of enterprise SSD and NAND contract pricing. HBM capacity diversion can tighten conventional DRAM even if MU’s HBM share lags SK Hynix, creating a favorable mix of price realization and utilization; that benefit should show first in MU’s next one-to-two quarterly gross-margin guide. Conversely, SNDK’s long-dated earnings multiple embeds execution assumptions that are not yet independently validated by contracted volumes, pricing floors, or customer concentration disclosures.

A seasonal equity drawdown would be a weak standalone entry signal. The more relevant 1-3 month catalyst is whether hyperscaler capex and NVIDIA platform shipments continue to translate into memory-content demand without a sequential inventory build at server OEMs and SSD distributors. If AI spending pauses, memory equities can de-rate before spot pricing weakens because the market will discount the next supply-addition cycle; MU is better capitalized for that outcome than SNDK, whose pure-play exposure has less business diversification.

The consensus may be underestimating that supply discipline is self-reinforcing: incremental advanced-memory capacity competes for tools, packaging, engineering labor, and wafers, slowing a rapid commodity response. But forecasts extending tightness several years should not receive full valuation credit: memory producers have historically raised capex once cash flows inflect. Watch quarterly capex plans from Samsung, SK Hynix, and Micron, plus DRAM/NAND contract-price direction; a return to aggressive capacity additions would compress both earnings estimates and multiples within months.

Key Decisions for Investors

• Prefer MU over SNDK on any broad semiconductor-led pullback of 10%+ over the next 1-3 months: initiate a 2:1 long MU / short SNDK relative-value position. MU offers broader DRAM exposure and lower reliance on FY2028 assumptions; exit if MU guides gross margin or DRAM bit shipments materially below consensus, or if the pair underperforms by 15%.
• Do not buy SNDK solely on a low forward multiple tied to fiscal-2028 estimates. Upgrade to a directional long only after evidence of multiyear NAND contract volumes, realized pricing, and near-term margin guidance supports the long-range model; otherwise treat it as a high-beta watchlist name rather than core exposure.
• Use NVDA earnings and hyperscaler capex commentary as the near-term demand read-through rather than election-season timing. A reduction in accelerator supply-chain visibility or signs of server/SSD inventory accumulation would be a trigger to reduce memory exposure before memory spot prices respond.
• Monitor ASML and leading-edge memory capex orders over the next two quarters. A material acceleration in equipment bookings from Samsung or SK Hynix is a thesis-falsifier for sustained scarcity and warrants taking profits on MU/SNDK longs even if current contract pricing remains firm.