Joseph Wilson
Portfolio Manager, U.S. Equity
Published:
Tech has started 2026 on a more mixed but still constructive note after several years of outperformance. Investors are digesting strong returns from 2024 and 2025 while weighing evolving themes, including how AI could reshape software business models, higher for longer rate expectations, and geopolitical uncertainty. Even so, investor interest in tech remains solid, while industrials have benefited from defense spending and some other cyclical sectors have faced outflows. Tech remains essential, but the focus has shifted to efficiency, and investors are increasingly favoring active strategies as dispersion rises and AI narratives evolve.
Against this backdrop, Jon Maier, Chief ETF Strategist, sat down with Joe Wilson, Portfolio Manager of the JPMorgan U.S. Tech Leaders ETF (JTEK).
We’re off to a more volatile start to the year in 2026. As a growth-oriented technology investor, what keeps you grounded in times like this?
The market’s sentiment around tech tends to oscillate between extremes. Tech is either a “bubble” or “dead,” and right now we’re in a downturn and we’re working through this volatility. As a long-term investor, you want to go through these periods where you’re underestimated, where tech’s influence on the broader economy is underappreciated. One signal I’ve focused on is that if you chart the NASDAQ Composite vs. the S&P 500 Index, it tends to trend upwards (below). That march higher is a function of tech permeating through the economy. Every few years we see a relative pullback of about 7-8%, and those are painful, but that’s also when you want to be searching for emerging winners.
Software has been a major driver of this volatility – what are your thoughts on the space?
Since November 2025, the sector is down more than -30%.¹ During that period, JTEK has held a 20-percentage point underweight to software. The spotlight is on software right now and investors are asking tough questions: can you co-opt LLMs and accelerate topline, or should we expect elevated competition? My view is that it will pay to be selective. This isn’t the time to superficially say that “infrastructure software is insulated” and hope to be defensive – you want to be selective and aligned with management teams that want to aggressively create a software market for humans and agents. That means thinking very differently about topline and the embedded margin potential in their business.
If your view is that software companies need to find efficiencies, does that mean developers lose their jobs? How does this broader AI adoption story impact the job market?
Software companies that leverage LLMs effectively will run leaner. They’re going to get more done with less, and that means cuts. That said, I’m bullish on developers diffusing through the economy and for the quality of software to step higher. Think about all the software that you interact with, and how it could be better. That’s the opportunity.
As the supply of intelligence grows, I don’t want to underestimate the potential displacement in the broader job market, but I’m optimistic about how much ingenuity there is in the economy. When you make coding easier, you unlock domain experts that in turn become more active participants in shaping their industries.
Where in the AI stack do we want more exposure (infrastructure, hardware, software etc) over the next 6–12 months, and why?
Whenever there’s this much change, it’s important to remind everyone that humility is built into the process. If you think about the internet era, AOL and Netscape were heralds of what people imagined the future to look like. Then, when Internet Explorer “won,” it won in a narrow sense, where the breakaway innovation of Google and Meta would come out of left field. I think you can draw parallels to today, where OpenAI and Anthropic are these incredible companies, but the world will continue to change and surprise investors. At this stage, humility to what’s around the corner is not optional.
When we think about the AI stack, our fundamental work points to bottlenecks across compute, memory, and networking. When a handful of hyperscalers are expected to spend >$600b on capex, you should expect uneven demand shocks. ² Companies like Lumentum in optical networking are expected to grow >80% this calendar year, while a memory hardware company like SanDisk is expected to grow topline >160%.³ We’re overweight both and have benefited from this transformative growth, but are focused on the cyclical risk embedded in these positions.
Outside of core tech, which “tech-first” leaders are we using to diversify AI risk?
Going back to where we started, I have conviction that technology will diffuse through the economy. Right now, every company is fielding questions about their AI strategy – similar to how they were fielding questions about their “internet strategy” 25-years ago, or their “mobile strategy” 15-years ago. No one is asking these questions now, because these once disruptive technologies have become table stakes. The same will happen with AI, so having the flexibility to invest in tech-enabled leaders is upside that also diversifies the portfolio.
Concretely, we see a lot of differentiation in terms of how companies are leveraging better and cheaper AI inference, where they can benefit from AI without the uncertainties that come with sizable, upfront capex. That spans core holdings in Shopify and Robinhood that have improving topline and margins, and also in businesses that are creating new markets like Tempus AI in algorithmic testing, or reinventing existing industries like Lemonade in insurance.