As of the end of September, the top 10 holdings included Robinhood Markets (5.3%), Nvidia (5.2%), Snowflake (4.1%), Oracle (4.0%), Take-Two, Netflix, Tesla, Meta Platforms, TSMC, and Broadcom. Sector weights were led by software (36.6%), internet (23.0%) and semiconductors (21.1%). Performance for A (dist) – USD to Sept 30 shows a performance of +32.1% over one year and annualised +33.8% over three years, outpacing the Russell 1000 Equal Weight Technology Index on those horizons.
Wilson frames the current cycle as the “dawn of AI” rather than its peak, arguing that the demand for AI compute is still climbing as both users and usage intensify. “What we don’t have [enough of] are graphics processing units or GPUs … Every single one of them is red hot”. He believes that supply could eventually overtake demand, but notes that today’s build-outs are multi-year projects, bottlenecked by land, power and infrastructure timelines.
The way Wilson sees it, a hallmark of the strategy is its equal-weight benchmark rather than a market-cap index, which he views as an “advantage”, especially as the largest US stocks take up greater index share. “If you have a market-cap-weighted benchmark, and you don’t like a company, you would still potentially own it,” he says. “We don’t like a name, we don’t own it, and we take that capital and we invest it into those companies that we think are interesting and have alpha-generating characteristics”.
The fund invests at least 67% in the equities of tech-related companies linked to the US economy and may also invest in small caps and select Canadian names.
See also: A monthly stock take of value-up efforts
He adds that the team is structurally underweight the Magnificent 7 (Mag 7) relative to the S&P 500 and prefers to find “disruption from below”. “Our opportunity is … investing in smaller-cap companies or mid-sized companies that we see an opportunity for them to become very large,” he says, adding that the fund generates more alpha from smaller companies than the larger ones.
An example is Robinhood, which he classifies as a technology company despite its “financials” label. “We bought [Robinhood] at close to $12 per share… and that was just two years ago,” he says. The position is now among the fund’s largest, reflecting management’s post-pandemic reset and a platform that keeps users engaged across market cycles.
Wilson also highlights a networking company initially valued at a US$1 billion ($1.3 billion) market cap for its “high bandwidth, low cost” expertise and leverage of far denser connections in AI data centres. A customer deferral triggered a 50% one-day plunge early in their ownership. Still, the team added the stock on conviction, and the company now has a market capitalisation of nearly US$24 billion. “That was one of our best investments that we’ve made in the last few years,” he says, citing a similar early call on Nvidia back in 2016.
See also: Reaping dividends along your investment journey
Where AI goes next
Wilson expects AI’s impact to spread well beyond consumer chat interfaces. “We’re going to see this more in healthcare, biotech, financials, industrials and manufacturing,” he says. He points to healthcare administration and billing, where “experiences that are still really terrible” are ripe for process improvement, and notes that drugmakers are applying AI to small-molecule discovery and targeted therapies. On factory floors, he sees scope for “physical AI”, including humanoid robotics to support costs in labour-intensive processes.
On timing, Wilson says it is “still too early” to see broad profit capture on the software side, with companies hiring engineers and sales staff to build usage. Monetisation models for frontier models, such as subscription, enterprise licensing, ads or commerce take-rates, are still being tested. He notes the near-term earnings payback is most obvious “for one company that’s selling the chips”, while downstream returns become clearer with scale and slower capex growth.
When asked about a “tech bubble”, Wilson argues that today’s conditions differ from the dot-com era. Usage per user and “test-time compute” are both increasing, with the latter following breakthroughs that let large language models internally “think” through candidate answers. While supply could eventually catch up, he sees multi-year capex constrained by real-world execution. In the meantime, consumer adoption is notable — he refers to hundreds of millions of monthly active users across major platforms — and he expects enterprise adoption to follow the familiar path, with employees using tools in the workplace before official IT rollouts.
Within the AI stack, Wilson sees notable activity outside the incumbent GPU leader. He cites hyperscalers’ in-house silicon and announcements around alternative infrastructure as drivers for networking, storage and other components. He also notes a pendulum swing in talent and venture capital.
For the past 20 years, Wilson notes that capital has been flowing into and funding software, while the semiconductor and hardware sectors have seen less investor attention. Even on the talent side, the semiconductor industry was not highly sought after. Today, this has changed and the market is more appreciative of the semiconductor space. “Maybe we have overbuilt our software capabilities,” he says.
In the pipeline
For more stories about where money flows, click here for Capital Section
Looking ahead, Wilson believes that the tech industry has considerable room to grow without being a bubble. He expects more tech companies to go public, such as private AI companies that have scaled from near-zero revenue a year ago to nine-figure run-rates today, which, to him, is an indicator that enterprise demand is taking shape.
For investors fixated on valuation and timing, he offers a pragmatic view: corrections will occur, but technology’s share of daily life tends to rise through cycles. The fund’s track record over the past three and five years has benefited from this reality, and the equal-weight framework provides room to express high-conviction ideas outside the largest caps. As of the end of September, the portfolio’s beta over one year was 1.19, with a Sharpe ratio of 1.06 — consistent with a higher-octane, high-alpha approach that accepts volatility in pursuit of outperformance.
As AI infrastructure scales and adoption moves from consumer novelty to enterprise workflow, Wilson aims to stay early and flexible. His “3Cs” — curiosity, creativity and collaborative effort — underpin the fund’s search for the next wave of giants that are gradually emerging from beneath today’s index heavyweights.
