Our top and bottom performing stocks in the market's record run over the past 6 weeks

The article highlights a strong six-week stretch for the market, with the S&P 500 up 6.7% and the Nasdaq up 10.6% since the April 16 meeting. Arm was the top performer, surging 97.9% on AI-driven CPU demand, while CrowdStrike (+60.6%) and Palo Alto Networks (+53.8%) rallied on continued cybersecurity strength; Qnity added 25.3% after blockbuster earnings and raised guidance. Losers included Meta (-9.5%), Home Depot (-7.9%), and Capital One (-7.1%), reflecting higher AI spending, stubbornly high mortgage rates, and softer credit conditions.

Analysis

The key second-order read is that this tape is rewarding “picks-and-shovels” AI exposure more consistently than the headline compute beneficiaries. Arm and Qnity are telling us the market is paying up for infrastructure that scales with model deployment, not just model training; that matters because the next leg of AI spend likely shifts from capex narratives to throughput, power efficiency, and packaging bottlenecks. In that setup, suppliers with pricing power can keep rerating even if hyperscaler multiples compress.

Cyber is more nuanced: the sector is still being treated as a secular winner, but ZScaler’s miss creates a near-term dispersion event rather than a blanket warning. CrowdStrike and Palo Alto look best positioned to absorb the shock because they sit deeper in the platform stack and can bundle more aggressively, while weaker peers lose share faster when budgets get scrutinized. The takeaway is that “AI hurts cyber” is not the right trade; the right trade is long dominant platforms, short smaller single-product names with guidance sensitivity.

Meta’s reaction suggests the market is drawing a hard line between AI investment that is visibly monetized and AI spend that must be justified ex ante. Without a cloud annuity to cushion higher capex, every incremental dollar spent on AI is being discounted more heavily than the same dollar at the other hyperscalers. That creates a relative-value opportunity in the mega-cap complex: the market may be underestimating how much of Meta’s drawdown is sentiment-driven rather than a true deterioration in operating momentum.

Housing and rate-sensitive financials remain the laggards because the market still does not believe in a clean disinflation / easier policy path over the next quarter. Home Depot is effectively a duration trade disguised as a retailer, so if yields stay sticky for another 1-2 months, earnings revisions can keep drifting lower. Capital One’s softer move looks more idiosyncratic and may offer a better recovery setup once consumers stabilize and Discover synergies become more visible.

Key Decisions for Investors

• Stay long CRWD and PANW into earnings on a 1-2 month horizon; use ZS weakness as evidence of share shift rather than sector contagion. Risk/reward favors the leaders if guidance confirms platform consolidation.
• Buy ARM on pullbacks but trim strength aggressively; the stock is benefiting from a multiple expansion phase that can reverse quickly if AI capex sentiment cools. Better as a tactical long than a core hold at these levels.
• Pair long Q / short ZS for 1-3 months to express winner/loser dispersion in cybersecurity and semicap materials exposure. Q has operating leverage and guidance momentum; ZS faces execution risk and sentiment fragility.
• Consider a relative-value long META / short a basket of the highest-quality hyperscalers only if you expect capex scrutiny to overshoot. The trade works if the market re-rates monetization discipline over spend intensity.
• Avoid adding HD until 10-year yields and mortgage rates break lower for a sustained period; if rates stay elevated for another quarter, the stock likely remains a value trap rather than a mean reversion story.