Buy 3 AI-Led Stocks Amid Solid Estimate Revisions and Upside Potential

Zacks highlights Quanta Services, Keysight Technologies and Amphenol as AI-infrastructure beneficiaries with improving earnings estimates and implied brokerage-target upside of 25.3%, 17.7% and 21.4%, respectively. Quanta's backlog reached a record $53.4 billion, up 49% year over year; Keysight's fiscal Q3 orders increased 56% to $2.091 billion; and Amphenol's IT datacom sales rose 63% organically, driven by AI demand. Consensus current-year EPS estimates increased 1.4% for PWR, 2.7% for KEYS and 1.1% for APH over the past 30 days.

Analysis

The investable read-through is a rotation within AI capex from compute suppliers toward the physical bottlenecks: grid interconnection, power delivery, high-speed connectivity and validation. PWR has the longest-duration earnings visibility but also the greatest exposure to permitting, utility capital-budget timing and labor execution; APH has the cleanest content-per-rack leverage and greater operating-margin scalability; KEYS is the higher-beta beneficiary if silicon-photonics, advanced packaging and 1.6T networking ramps accelerate. This favors APH over PWR on a 6-12 month basis if hyperscaler buildouts remain capital-intensive but construction schedules slip.

Near term, the cited estimate revisions are not independently sufficient catalysts: all three require upcoming results to demonstrate that AI orders are converting into revenue without rising expedited-freight, labor or component costs. The key 1-3 month confirmation points are APH datacom organic growth and incremental margin, KEYS order mix/book-to-bill and software attach, and PWR backlog conversion plus margin guidance. A higher-for-longer rate regime matters more for PWR because delayed utility approvals and customer financing can defer large projects, whereas APH's shorter-cycle component demand can remain resilient until actual server-rack shipments slow.

Consensus may be underpricing a second-order constraint: power availability can shift spend from new greenfield campuses toward retrofits, networking upgrades and power-density improvements. That is relatively favorable for APH and Eaton (ETN), while PWR's opportunity is larger but more lumpy and subject to project sequencing. Conversely, a hyperscaler capex pause would hit KEYS and APH earlier; PWR's contracted work should lag the downturn but would face multiple compression if backlog growth decelerates. QBTS has no fundamental linkage to this setup; its presence appears promotional rather than a supported AI-infrastructure read-through.

Key Decisions for Investors

• Initiate a 6-12 month long APH / short PWR pair, sized beta-neutral: favor APH's higher recurring content and faster conversion against PWR's execution- and rates-sensitive project exposure. Target 10-15% relative outperformance; exit if APH datacom growth falls below 25% year-over-year or PWR raises margin/backlog-conversion guidance materially.
• Buy KEYS only on a post-results confirmation of sustained order momentum and improving software/services mix; use a 3-6 month position with a 12-15% downside stop. Upside requires evidence that advanced-node and optical test demand is broadening beyond a small hyperscaler/customer cohort; a book-to-bill below 1.0x would falsify the thesis.
• Maintain PWR as a watchlist long rather than chase momentum ahead of evidence on labor availability, project starts and cash conversion. Add on a 10%+ pullback or after confirmation that backlog conversion and segment margins are intact; reduce if utility/developer project deferrals emerge or free-cash-flow conversion misses guidance.
• Use ETN as a complementary, lower-duration power-density beneficiary rather than adding broad AI beta. A long ETN / short an equal-beta basket of data-center construction-sensitive industrials can monetize retrofit and electrical-content demand if grid delays constrain greenfield capacity over the next 6-18 months.