Confusion to Conviction

Three months ago we wrote about the market's capitulation on software. This quarter is about what came after it, the first signs of a comeback confirmed, and the initial proof that our process works exactly as designed, even when it takes real discomfort to get there.

The Goodhart Global Future Leaders Fund returned 22% in Q2 2026, against 14% for the benchmark, an active return of over 8%. Since the launch on the 7th November 2025, the fund is up 3% against 12% for the benchmark. Just to reiterate, we are not managing this fund to beat an index; the benchmark exists to calculate the performance fee, as set out in our launch letter.  Our long-term investment objective remains to seek annualised returns of approximately 15% over the long term and Q2 shows this is possible. The path has not been smooth. March was the trough, with the fund and the benchmark falling together as AI disruption fears and geopolitical noise peaked. Capitulation in March was followed by a sharp recovery, a pattern I have invested through in previous roles. We would rather not repeat it too often, though we anticipated higher volatility, shorter cycles and the need for more agility as the four forces made the backdrop for markets harder.

Nowhere is this clearer than with Soitec. It was one of our largest detractors in our first reporting period. This quarter it has been, without close competition, our best contributor, at one point up more than 700% year-to-date from its January low, and up 125% in the quarter alone. We wrote last quarter that we had increased our position following meetings with management, sensing the fundamentals were closer to turning than the share price implied. That decision has been the single best call we have made since launch. The narrative shifted from oversupply in the smartphone supply chain to Soitec's silicon photonics business becoming a genuine AI data centre beneficiary, and the market has re-rated the stock sharply in response. Soitec is now our largest single contributor to returns since inception and sales estimates have only just turned in what could be a multi-year cycle.

ASML has benefited from a related but distinct dynamic. For some time, leading edge EUV tools looked like a two-buyer market: only TSMC and Samsung could afford to purchase at the scale required. Improving fundamentals across both logic and memory now mean Micron, SK Hynix, Intel, Tesla and others can all afford to pay up for the tools they need to stay competitive. That is a broader, healthier customer base than the market has priced in, and it will be interesting to see how ASML chooses to price its next generation of tools into that demand. We would not be surprised if pricing power surprises to the upside.

Alongside Soitec and ASML, we have built new positions in ACM Research and ams-OSRAM this quarter, adding to our existing holding in PDF Solutions since launch. All three mid-caps have been strong contributors and, more importantly, sit in parts of the value chain we believe can do well largely independent of the overall semiconductor capex cycle. These are not simply leveraged bets on AI infrastructure spending; they are businesses with their own idiosyncratic drivers, which is exactly the kind of diversification within conviction we look for.

The software portfolio has told a related story. Since the January and February sell off, an early but consistent pattern has emerged: this has been a multiple event, not an earnings event. Veeva's FY2027 consensus earnings estimates have risen from around $8.00 to $9.06 over the past year, while the shares are still down around 40% from their November peak. Kinaxis tells the same story, estimates up from roughly C$4.70 to C$5.17 even as the stock fell from the 150s into the 100s. The clearest fault line has been infrastructure versus application software. Infrastructure names, the picks and shovels of AI workloads, have broadly recovered their January drawdown, while application software remains roughly 30% lower over twelve months, a gap that has had little to do with what the actual numbers are doing. We think that gap is a source of future return rather than a reason for concern, and both Veeva and Kinaxis have accelerated buybacks into the weakness, exactly the capital discipline we want to see from management at these valuations.

Healthcare has been the more difficult side of the portfolio this quarter. Some of that weakness reflects investors selling healthcare positions to fund larger, or new, positions in semiconductors and AI infrastructure elsewhere in the market, a rotation we understand even where we think it has gone too far. Valuations across several of our healthcare holdings have rarely looked cheaper, and the demographic tailwinds we set out in our recent note on future leaders in healthcare have not gone anywhere. Ageing populations and stretched healthcare budgets remain some of the most durable structural trends we can identify. Bonesupport and Kuros Biosciences both had a weak quarter in share price terms, but we have not seen a single downgrade to consensus estimates or company guidance for either business. The weakness has been a market de-rating, not a fundamentals problem, hence we remain comfortable with our positions.

If there is a single lesson from the past nine months, it is that our process rewards patience, but only when that patience is backed by continuous, active reassessment. We do not hold names because we bought them; we hold them because of their economic potential and how they fit into the overall portfolio construction, revisited every time new information arrives, telling us the thesis is intact. The decision to add to Soitec in February, into a falling share price and a narrative that had turned hostile with multiple listed short sellers, was not a hunch. It followed direct engagement with management and a genuine change in our own numbers. That is agile investing in practice: not moving quickly for its own sake but being willing to move decisively when the facts change, even before the market has caught up.

We want to end this letter with something other than numbers. Launching a fund into the kind of volatility we experienced in our first quarter is not for the faint hearted, and we are acutely aware that our investors sat through a genuinely uncomfortable period with very little to show for it beyond our conviction that we were right. That patience is now being rewarded, and we do not take it for granted. Thank you.

I also want to thank my co-PM, Andrew Heap. The period since launch has, at various points, tested every part of our process and our partnership, and Andrew has been an outstanding partner throughout it, rigorous when the market wanted us to panic, and just as rigorous now that the market has started to agree with us. We built this fund on the belief that sticking to a proven philosophy, even when it is deeply unfashionable to do so, is how genuine long-term returns are generated. Q2 does not prove that belief was correct on its own, but it is the clearest evidence yet that it might be. I look forward to using this experience to give us the springboard to meet our longer-term targets.

Outlook

We remain convinced that the AI infrastructure buildout has further to run, and that the businesses enabling it, rather than the hyperscalers spending the capital, represent the more attractive long-term opportunity. That view is reinforced by the growing competition facing those hyperscalers, and a genuine question mark over their own cash generation, historically their single greatest attraction to investors. Healthcare remains a source of frustration in the short term, but conviction remains in the long term, and we will continue to look for the catalysts that close the gap between fundamentals and share price. We remain focused on identifying tomorrow's future leaders today, and we remain grateful for the trust our investors have placed in us to do so.