These 2 Potential Stock Splits Look Like Screaming Deals Right Now

Micron (MU) and SanDisk (SNDK) are highlighted as beneficiaries of the data-center memory build-out, with SandDisk up to ~$1,600 (+>~$1,300 vs < $300 at start of year) and Micron at about $970 (starting < $300, but ~20% below all-time highs). The article attributes growth mainly to pricing (SandDisk: two-thirds of growth from higher prices, one-third from increased output) and notes supply capacity won’t normalize until late 2027–2028. While it discusses a potential stock split (Micron last split in 2000) and upcoming Micron Q4 results as a catalyst, it frames the underlying fundamentals as strong enough to buy regardless of splitting.

Analysis

The fundamental trade is not the split narrative; it is the duration of the memory pricing squeeze. In the next 1-2 quarters, the key mechanism is operating leverage: every incremental bit of output is being sold into a tight market, so gross margin expansion can outpace revenue growth. That makes MU the cleaner expression because DRAM and NAND exposure gives it broader pricing participation, while SNDK is a purer but more fragile NAND bet.

Second-order winners are the AI infrastructure stack that needs less memory intensity per dollar of compute than the market is assuming: if memory stays expensive, hyperscalers and server OEMs will have to optimize configs, which can shift mix toward higher-margin systems and delay some deployments. Losers are memory-heavy device assemblers and storage-adjacent buyers whose bill of materials gets marked up; if the cycle persists into 2027-2028, the competitive advantage compounds for the few vendors with scale and balance-sheet capacity to keep capex elevated.

The consensus risk is that investors are extrapolating spot pricing and current tightness too far out. Memory supply is notoriously self-correcting once new capacity and process-node improvements arrive, and demand can crack quickly if data-center capex pauses or AI server build plans normalize. What would falsify the bullish thesis: a sequential pricing deceleration in the next earnings print, management commentary implying capex pullback by hyperscalers, or signs that bit growth/supply additions are arriving earlier than the market expects.

Near term, the stock-split angle can add retail flow, but that is a sentiment catalyst, not a valuation driver. Over 1-3 months, the tradable setup is still the earnings/guidance path; over 6-18 months, the real question is whether this is a multi-quarter supercycle or just a delayed normalization with a sharper-than-expected downcycle after 2028.

Key Decisions for Investors

• Long MU on any post-earnings pullback into guidance, using a 1-3 month horizon; prefer stock or call spreads over chasing SNDK after its larger run-up. Risk/reward is best if management reiterates tight supply and price discipline.
• Pair trade: long MU / short SOXX or SMH into the next memory-price update. This isolates the memory cycle while hedging broader semis beta; thesis breaks if semiconductor leadership broadens beyond memory.
• If already long SNDK, trim into strength and rotate part of the exposure into MU. SNDK has more pure NAND torque but less diversification if NAND pricing peaks before DRAM.
• Set an alert on MU earnings for any sequential pricing commentary or capex normalization language. If management signals supply relief before late-2027, de-risk immediately; that would compress the multiple fast.
• Do not buy the stock-split story outright; treat it as a flow catalyst only. If the market bids the names another 10-15% on split speculation without a guidance raise, fade the move with tight risk.