Inside Active: T. Rowe’s Rizzo on Finding Technology Winners

The article argues that AI may broaden the set of technology winners beyond the current mega-cap cohort, highlighting a shift in how future tech leadership could be identified. It features Dominic Rizzo of T. Rowe Price discussing his investment framework for the Global Technology Equity Fund and Technology ETF. The piece is largely thematic and interview-based, with no specific earnings, guidance, or price-moving data.

Analysis

The market is still treating AI as a narrow semiconductor-and-hyperscaler trade, but the bigger opportunity is likely a second-order re-rating in the less crowded layers of the stack: power, networking, storage, software tooling, and industrial automation. If AI capex remains durable, the beneficiaries are not just the obvious compute suppliers; the real compounding can come from companies that convert incremental AI demand into recurring installed-base expansion and higher utilization of existing assets.

The competitive dynamic is also changing from “who has the best model” to “who can absorb capital fastest without destroying ROIC.” That tends to favor incumbents with scale and distribution, but it can also create openings for niche enablers that are ignored in mega-cap concentration trades. The underappreciated loser set is anyone whose valuation depends on stable legacy IT budgets; AI spend can cannibalize other software and infrastructure line items before it becomes fully additive, creating a period where winners take share faster than end demand grows.

Near term, the key risk is that enthusiasm runs ahead of monetization: the market can stay rewarded for capex announcements for months, but earnings will eventually have to show throughput, margin expansion, or usage-based revenue. If enterprise adoption slows or power/interconnect constraints bottleneck deployment, the trade can de-rate quickly because expectations are now embedded in multiple layers of the supply chain. The most likely reversal catalyst is not “AI disappointment” broadly, but a specific sign that incremental capex is yielding lower-than-expected revenue per dollar invested.

The contrarian takeaway is that consensus may still be underpricing duration: AI infrastructure spending could persist longer than feared because companies are buying optionality, not just immediate ROI. That supports a barbell of quality compounders plus select laggards with asymmetrically high operating leverage, while fading crowded names that have already been repriced for perfection.

Key Decisions for Investors

• Go long an AI infrastructure basket (e.g., NVDA/AVGO/ANET/CRDO) on 1-3 month pullbacks; use weakness to add because the setup still supports multiple expansion if capex revisions keep drifting up. Risk/reward is favorable until evidence of capex digestion emerges.
• Pair long MSFT or AMZN vs. short a broad legacy software basket over 3-6 months; AI should reallocate spend from maintenance software to platforms with embedded distribution. This captures the winner-takes-more dynamic while reducing exposure to pure AI sentiment swings.
• Initiate a long position in power-grid beneficiaries and data-center enablement names where available; the second-order bottleneck is electricity, not model quality. Hold 6-12 months and expect outperformance if AI buildout remains capacity constrained.
• Avoid chasing the most crowded mega-cap AI proxies after sharp rallies; instead, use call spreads rather than outright longs to express upside with defined downside over 2-4 months. The market can re-rate faster than fundamentals, but a monetization miss can unwind multiples abruptly.
• Watch for any earnings season sign that AI capex is rising faster than revenue contribution; if that appears, rotate from beta-heavy AI hardware into balance-sheet-strong compounding software names. That trade should work over the next 1-2 quarters if sentiment becomes more selective.