The 4 Best AI Stocks to Buy in May

The article argues AI demand remains exceptionally strong, highlighting Nvidia’s expected 72% revenue growth, Broadcom’s AI semiconductor revenue of $8.4 billion in Q1 FY2026 (+106% YoY), and TSMC’s raised 2026 revenue growth guidance to 30%. It also spotlights Nebius as a high-growth AI cloud beneficiary, with analysts projecting 523% revenue growth in 2026 and 206% in 2027. Overall, the piece is a bullish stock-picking article centered on AI infrastructure beneficiaries rather than a single new catalyst.

Analysis

The core setup is not “AI enthusiasm” but a sustained capacity shortage across the stack. When hyperscalers remain constrained, pricing power migrates from software narratives to the physical bottlenecks: accelerators, packaging, foundry capacity, and custom silicon design. That favors NVDA and AVGO near term, but the real second-order winner is TSM because it monetizes every incremental architecture shift regardless of whose chip wins the socket.

The market is still underappreciating duration risk: current AI capex is being committed against multi-year data-center builds, so the revenue stream is more durable than a typical chip cycle, but also more exposed to execution slippage. Any delay in power delivery, grid interconnects, or server rack integration can push revenue recognition out by quarters, which matters more for names trading on elevated expectations like NBIS. That makes NBIS the highest-beta expression of the theme, but also the most vulnerable to a growth-vs-capacity mismatch if customer conversion or financing tightens.

The consensus is likely too anchored on GPUs as the only trade. Broadcom’s custom ASIC mix is a subtle threat to GPU share at the margin, not because it kills NVIDIA, but because it expands the total AI silicon market while redistributing wallet share toward lower-cost, workload-specific silicon. That supports a barbell: own the infrastructure toll collectors, but hedge the most crowded multiple expansion names because the first leg of this rally can be correct on earnings and still overstate long-run normalized margins.

The key risk is not demand collapse; it is a sequence of smaller disappointments: capex pauses, customer digestion after the next quarter, or guidance that stays strong but stops accelerating. In that scenario, the stocks with the highest expectations-to-delivery gap should de-rate first, while TSM and AVGO should hold up better due to their broader customer bases and less binary end-demand exposure.

Key Decisions for Investors

• Overweight TSM vs NVDA for the next 3-6 months: TSM has the best risk-adjusted exposure to the AI buildout because it monetizes both GPU and ASIC growth without bearing the same valuation risk as a single-platform winner.
• Pair trade long AVGO / short a basket of AI-adjacent high-multiple infrastructure names for 1-2 quarters: AVGO has direct custom-silicon leverage with less execution risk than the pure neocloud layer.
• Treat NBIS as a tactical long only via call spreads into strength, not stock: use 3-6 month upside structures to express the growth story while capping downside if customer ramp timing slips.
• Maintain a core long NVDA but finance it by trimming into post-earnings spikes: the trade is still constructive over 12 months, but the multiple is now more sensitive to any sign of decelerating sequential growth.
• If AI capex headlines continue accelerating, add on pullbacks in TSM and AVGO rather than chasing momentum in NBIS; these are the names most likely to outperform once the market shifts from narrative to free-cash-flow quality.