NVIDIA & 2 Profitable Stocks to Buy Now for Explosive Upside

A Zacks screen identified NVIDIA, Micron, and Sandisk among 16 stocks meeting its profitability and growth criteria. Their trailing 12-month net profit margins are 63.7%, 63.8%, and 56.5%, respectively; expected earnings growth for the current year is nearly 95% for NVIDIA, 135.3% for Micron, and almost 201% for Sandisk. The article presents these figures as evidence of potential upside, without reporting company-specific news or market reactions.

Analysis

The screen risks confusing peak profitability with durable earnings power. Memory pricing is cyclical: Micron and Sandisk margins can reverse faster than annual growth estimates imply if supply expands or AI-related demand disappoints. Their reported earnings growth is especially vulnerable to a low comparison base and revisions to commodity-like pricing assumptions. NVIDIA’s economics are less directly exposed to memory spot pricing, but its results still depend on AI infrastructure budgets converting into sustained accelerator demand; strong margins do not insulate it from customer capex pauses or competition.

The second-order read-through is supply discipline. If Micron and other memory producers keep capacity additions measured, tight supply could support margins across the memory complex; aggressive additions by competitors such as SK Hynix, Samsung or Kioxia would undermine that setup. For NVIDIA, any moderation in accelerator deployments could also pressure adjacent AI infrastructure suppliers, while substitution toward alternative accelerators could weaken its pricing power.

Near term, this article is a screening signal, not a catalyst: broker ratings and forward growth estimates can already reflect optimism, and no valuation or estimate-revision data are provided. Over 1–3 months, monitor earnings revisions, memory contract pricing, inventory and company capex commentary. Over 6–18 months, the key question is whether AI demand sustains elevated investment while memory supply remains disciplined. The contrarian risk is that backward-looking margins make cyclical memory names look safer than they are; the counterpoint is that persistent AI demand may extend the cycle. Falsify the constructive memory thesis on deteriorating pricing, rising inventories or capacity-led guidance cuts; falsify the NVIDIA demand thesis on customer capex reductions or weakening accelerator orders.

Key Decisions for Investors

• Do not initiate positions solely from this profitability screen. Before adding MU or SNDK, verify current memory contract-price trends, inventory, supply plans and the direction of consensus earnings revisions.
• For the next 1–3 months, keep NVDA as the cleaner AI-demand exposure only if customer capex and order commentary remain firm; reassess on evidence of budget pauses, order softness or estimate downgrades.
• Treat MU and SNDK as higher-cycle-risk exposures rather than interchangeable AI beneficiaries. Consider reducing or hedging them if memory pricing weakens or producer capacity plans turn more aggressive.
• No immediate pair trade is justified by the supplied information: compare valuation, estimate revisions and price action first. A relative-value position would be falsified if the expected divergence in AI demand or memory pricing does not appear.