Investor Joe Terranova buys these AI plays with 'strong momentum, strong fundamentals’

The common thread is not “AI” in the abstract, but utility-like beneficiaries of AI capex and latency requirements. TWLO and GNRC are both being treated as picks-and-shovels names with improving narrative convexity: software that sits in the customer interaction layer, and power backup that sits behind the data center. That creates a second-order read-through for the rest of the basket: suppliers of grid equipment, power generation, and mission-critical communications should continue to attract incremental capital even if the model names consolidate.

The more interesting near-term risk is that these trades have become crowded factor expressions of momentum plus AI. When that happens, fundamentals matter less than the next data point that challenges the story—guide-downs, capex pauses from hyperscalers, or any sign that backup-power orders are being pulled forward rather than structurally reaccelerating. TWLO is especially vulnerable to a multiple reset if agentic-AI demand proves more incremental than durable, while GNRC is more exposed to cycle timing because power-theme winners can overshoot well before earnings catch up.

On the consumer side, SBUX, DKS, and ASO look more like selective technical/earnings normalization trades than durable secular longs. The market is rewarding names where management can create a path to margin or traffic inflection, but the second-order effect is that any disappointment gets punished harder because positioning is already leaning constructive. Travel looks similar: H and AAL can work tactically, but both remain hostage to fuel, pricing, and consumer elasticity, so the cleaner expression is relative value rather than outright beta.

The contrarian takeaway is that the market may be underestimating how fast capital rotates from the obvious mega-cap AI winners into the “infrastructure of AI” laggards once the narrative broadens. That argues for owning the enablers while they are still being repriced, but fading the most crowded momentum names if breadth narrows. In practice, the best risk/reward is in pairing AI-infrastructure winners against weaker discretionary or travel exposure, rather than chasing the headline theme outright.