Not Nvidia. Not AMD. 2 Chip Stocks I'd Buy Before the End of the Year.

The article argues Taiwan Semiconductor and Micron could benefit from continued AI infrastructure demand, citing TSMC’s more than 70% share of global foundry revenue in Q2 and management’s view that memory supply-demand conditions will be tighter in 2027 and 2028 than in 2026. As of Nov. 4, Micron traded below 5 times forward earnings after falling about 30% over the past month, while TSMC traded at 17 times 2027 estimated earnings after a roughly 19% decline. The author sees potential upside of up to 50% for TSMC next year and greater, more volatile upside for Micron; these are investment opinions, not reported forecasts from the companies.

Analysis

The apparent valuation gap is partly a cycle-comparison trap: TSM’s foundry economics are structurally differentiated, while MU’s earnings multiple can look exceptionally low near peak memory pricing and earnings. A re-rating in MU therefore needs sustained pricing and favorable product mix—not merely continued AI spending. Micron management’s tight-supply outlook is a useful signal, not independent proof; watch realized memory pricing, customer inventories, and capacity additions by Samsung Electronics and SK Hynix. A supply response or digestion in data-center orders could reverse MU’s earnings faster than the headline AI build-out slows.

TSM’s moat supports more durable earnings, but it does not eliminate risks from customer concentration, capital intensity, advanced-packaging constraints, or Taiwan geopolitics. Samsung and Intel Foundry are potential competitive alternatives, though substitution may be limited by qualification and execution requirements. The recent drawdown may reflect positioning and risk reduction as much as changed fundamentals; without price or flow data, do not treat it as confirmation of mispricing.

Near term, volatility can persist. Over 1–3 months, the key catalysts are company guidance and evidence that memory pricing and AI-related orders remain firm. Over 6–18 months, capacity discipline and actual returns on AI infrastructure spending determine whether current earnings are sustainable. The contrarian risk is that investors extrapolate AI demand while underweighting the cyclical supply response.

Key Decisions for Investors

• Prefer TSM as the core exposure; consider a smaller, higher-volatility MU position rather than an equal-weight allocation. The rationale is more durable foundry economics versus greater earnings sensitivity to commodity memory pricing.
• Do not chase the selloff solely on the article’s forward multiples. Stage any entry around the next guidance and results cycle; require confirmation from memory pricing, inventory trends, and AI-related order commentary before adding MU.
• Falsify the MU thesis if pricing or management’s supply-demand outlook weakens, customer inventories rise, or competitors signal materially faster capacity growth. Reassess TSM if guidance weakens on utilization, advanced-packaging execution, or customer demand; geopolitical escalation is a separate downside catalyst.
• No price target is supportable from the supplied data. Track the next reported results and guidance, and avoid treating the article’s potential-return claims as independently verified.