The AI trade is changing.
In his January 2026 investor letter1, Sparkline Capital’s Founder & CIO Kai Wu argues that markets are shifting from the AI buildout phase to the AI adoption phase. As Wu writes, “We believe the AI boom has reached a pivotal moment, with the focus now shifting from buildout to adoption.”
For the past several years, capital has poured into AI infrastructure — chips, data centers, model builders. But history suggests the long-term winners of technology revolutions are rarely the builders. Instead, they are the early adopters who use the infrastructure to drive productivity, revenue growth, and cost efficiencies.
Wu frames it clearly: “AI early adopters offer a path forward. Currently overlooked, they are starting to generate real AI-driven ROI while offloading capital spending risk.”
The Proof Is Emerging
Sparkline analyzed global earnings calls and categorized AI mentions into three tiers: simple usage, economic gains, and true ROI. The data show a meaningful shift.
Since 2017, firms reporting AI-driven ROI have risen from near zero to 155 companies. Those reporting economic gains have grown to 675. More importantly, companies that can point to quantified AI returns have outperformed.
Wu notes: “Companies deploying AI have also delivered superior stock returns.” Firms citing AI-driven ROI have earned the highest excess returns.
In other words, AI is no longer just narrative — it is increasingly measurable.
The Valuation Gap
Here’s where it gets interesting.
Despite accelerating adoption, the market is not pricing early adopters at a premium. According to Sparkline’s analysis, “there is virtually no correlation between AI adoption and valuation.”
Meanwhile, AI infrastructure stocks trade at nearly twice the broader market multiple.
Wu warns that infrastructure valuations assume rapid near-term demand. Adoption, however, tends to follow a slower S-curve. If demand lags expectations, multiple compression risk rises.
That creates a relative opportunity.
“With valuations missing this for now, we see an opportunity to invest in early adopters before the market catches on.”
The Index Problem
Most investors may be more exposed to infrastructure risk than they realize.
Wu points out that “a whopping 46% of the S&P 500 is currently in AI infrastructure stocks.” Market-cap weighting naturally tilts portfolios toward what has already worked.
Alternative indexes often overcorrect — concentrating instead in AI laggards.
Sparkline positions its ETFs as a middle path: maintaining AI intensity while reducing infrastructure concentration.
Bottom Line
The AI revolution isn’t ending. It’s evolving.
The buildout phase rewarded capital spending. The adoption phase will reward execution, efficiency, and measurable ROI.
As Wu concludes, “As the AI boom shifts from buildout to adoption, investors need a new approach.”
The key question for 2026 isn’t whether AI wins.
It’s who captures the value next.
Footnote:
1 Wu, Kai. AI Adopters: Beneficiaries of the Boom. Sparkline Capital LP, 30 Jan. 2026. Sparkline 2025 ETF Investor Letter.
