3 AI Stocks to Buy Immediately

The article highlights three AI-adjacent stocks the author считает still undervalued: Fluence Energy (FLNC), Atkore (ATKR), and Hesai (HSAI). Fluence is said to have doubled order intake to $2.0 billion, Atkore is expected to reaccelerate on data center and solar demand, and Hesai is forecast to grow revenue 43% this year and 40% next while nearly doubling EPS to $6.35 by 2027. The piece is largely a bullish stock-picking commentary rather than a company-specific catalyst, so the likely market impact is modest.

Analysis

The important read-through is that AI spend is broadening from chips into the unglamorous bottlenecks where budget authority is still weak and competitive intensity is lower. That usually creates a better setup for asymmetric rerating because these names can grow into estimates for several quarters before the market fully revises the terminal multiple. The common thread is not “AI exposure” but exposure to capacity-constrained infrastructure that must be added regardless of model winners, which makes demand less discretionary than most investors assume.

The second-order effect is that hyperscaler capex is increasingly forcing adjacent customers to compete for the same electrical, cabling, and sensing supply chains. That can lift pricing power in pockets that were previously cyclical, but it also creates a barbell: suppliers with differentiated specs and installation complexity can reprice faster, while commoditized industrials remain vulnerable to stop-start ordering. The market is still underestimating how quickly backlog can translate into margin expansion once utilization moves up, especially where fixed-cost absorption is still recovering from the last downturn.

The main risk is timing mismatch, not thesis failure. These setups can stay range-bound for 1-3 quarters if investors remain fixated on near-term macro or if capex announcements do not convert into shipped revenue quickly enough. For the higher-beta names, the catalyst sequence matters: order growth first, then backlog visibility, then margin inflection; missing any one of those can cause a sharp de-rating even if end-demand is intact.

Contrarian view: the crowd is likely right on direction but wrong on magnitude. A lot of these beneficiaries may already have “AI premium” in the narrative, yet not in the multiple, which means the best risk/reward is where earnings revisions are still underappreciated and the market is anchoring to prior-cycle cyclicality. The opportunity is to own the businesses where AI is driving a structural step-up in utilization, not just incremental headline revenue.

Key Decisions for Investors

• Long FLNC on a 3-6 month horizon; use post-earnings momentum as entry confirmation. Risk/reward is attractive if data-center-linked order intake continues to convert, but cut exposure if backlog growth stalls for two consecutive quarters.
• Long ATKR as a rerating trade over 6-12 months; the market is still pricing it like a housing cyclicals name. Best entry is on any pullback toward the low-$70s, with upside from backlog normalization and AI/data-center mix improvement.
• Long HSAI on a 6-12 month horizon; this is a consolidation winner with higher multiple potential if robotics/autonomy shipments continue to scale. Position size should be smaller due to geopolitical/listing risk, but upside is asymmetric if U.S. investors re-rate the category.
• Pair trade: long FLNC or ATKR vs short a basket of traditional industrial cyclicals with no AI linkage. The goal is to isolate AI-infrastructure demand from macro housing/PMI noise and capture multiple expansion in the better-placed names.
• For higher-conviction expression, buy 6-9 month call spreads in FLNC or ATKR rather than outright stock. That limits drawdown if the market delays the re-rating while preserving upside if earnings revisions accelerate.