After strong outperformance in 2024 and 2025, investors began 2026 by digesting their AI-fueled tech gains while reassessing the risk backdrop. Higher-for-longer rate expectations and persistent geopolitical conflicts weighed on the sector, as did the potential for AI to reshape business models. Notably, software’s plummet in March put it in the crosshairs of a broader debate over whether AI becomes a durable growth accelerant or a catalyst for intensified competition. In their examination, our research analysts are focused on management teams building for a world of both humans and agents, where product strategy, distribution, and embedded margin potential are rethought rather than defended with broad “insulated” narratives.

More recently, that existential debate took a back seat as mega caps bounced back in a big way. Top-heavy concentration has made cap-weighted outperformance more common, but even by the trends of the past few years, this run-up is outsized. Technology sentiment has a long history of swinging between extremes; in that context, maintaining a longer-term perspective makes the difference.

An active growth story

Tech ETF flows have crossed $20 billion year-to-date, the most of any sector and a reflection of renewed confidence in growth opportunities tied to AI adoption and infrastructure. The rise of active management remains a trend. Approximately 75% of tech ETF flows this year have been into active strategies, and 18 of the 30 new tech ETFs launched in the past 12 months are actively managed. As the AI story matures, competitive moats are tested, and winners and losers diverge, we believe active management will be even more important. Periodic tech pullbacks can be painful, but for a skilled manager, they’re also when emerging winners are most likely to be identified and captured.

Memory mania

The latest craze within the ETF market is the memory subtheme. Developing a large language model requires significantly more data, requiring higher bandwidth memory. Meanwhile, the largest providers of these critical memory components are unable to meet demand. This is driving up price and, by extension, earnings potential for these companies. However, it’s important to keep in mind that this bottleneck is a function of supply/demand imbalance which may prove to be cyclical, as historically has been the case, as opposed to a major architectural shift much like we are seeing elsewhere in areas like networking or compute.

Spotlight on JTEK: capturing AI’s future

It’s important to remember that early category leaders don’t always define the eventual outcome, think Netscape in the early days of the web or MySpace in early days of social Media. With a portfolio management team averaging more than 25 years of industry experience, the JPMorgan US Tech Leaders ETF (JTEK) is uniquely positioned to capture the AI winners of tomorrow.

• Currently, the portfolio is positioned to benefit from bottlenecks across compute, memory and networking.
• Joe Wilson and team expect hyperscaler capex to drive uneven demand shocks and embed cyclical risk in even high-growth beneficiaries.
• JTEK also allocates to tech-enabled leaders outside of the tech sector that benefit from cheaper inference without sizable upfront capex. Many non-tech companies, particularly within the communication services and consumer sectors, are benefitting from AI through margin expansion and execution.