Lessons from Silicon Valley

The companies referenced in this insight are shown for illustrative purposes only. Their inclusion should not be interpreted as a recommendation to buy or sell. The information provided is to illustrate our current investment activities and approach only, and should not be construed as offer, research or investment advice. Past performance and other indicators or metrics do not predict future outcomes. Please read important information at the end of this insight.

1. AI & digital infrastructure: still early, still capacity‑constrained
 

On the infrastructure side, the message was remarkably consistent: we are still in an early, multi‑year buildout of AI and data infrastructure and the key constraint is capacity, not demand. The companies we met are all levered to this build‑out in different but complementary ways. Here are some examples:

• Super Micro Computer, a key supplier of AI servers and datacentre solutions, with real non‑speculative order books.
• Lumentum and Cohu, sitting in the optical and test bottlenecks that need to be solved if data centres are to keep scaling.
• CleanSpark, building high‑density compute and power infrastructure at the intersection of AI and energy.
• Stem, optimising power systems that combine solar panels with batteries, which are critical if the AI cycle is not to hit a wall on the power side.

The main takeaway?

These companies are investing aggressively but with far more discipline than in past hype cycles. When we dig into their plans, we see payback periods and product mix being managed carefully, not just “growth at any price”. That’s exactly the nuance you only pick up when you meet management teams in person. For us, this reinforces a clear portfolio stance: we want exposure to the infrastructure, but only where we are convinced the capital discipline matches the growth opportunity.

2. The real profit pools: data and software above the hardware

Above all that hardware, a data and software layer is rapidly emerging as the most defensible profit pool. The companies we met turned that abstract idea into something very concrete. For example:

• Planet Labs is transforming dense satellite imagery into a subscription data and AI‑ready time series business. This isn’t about pretty pictures from space; it’s about geospatial intelligence on tap.
• Guardant Health is doing something similar in oncology, a deep clinical data asset paired with a fast-ramping colorectal-cancer blood test, with guidelines and reimbursement likely to catch up.
• Coinbase is positioning itself as core infrastructure between crypto and traditional finance with regulation increasingly a tailwind, not a headwind.
• Upstart and Affirm are the financial analogues, using alternative data and AI to underwrite or structure consumer credit. After a volatile period, both are clearly prioritising risk discipline and funding robustness over unchecked growth.

As active managers, we don’t take the AI label at face value. We want to know: where is the data moat? How sticky is the revenue model? How does regulation shape the economics? On this trip, we came back convinced that data‑rich, software‑driven models are where long‑term compounding is most likely to happen. That’s exactly where we strive to take a high-conviction view in the convertible space.

3. Healthcare & life sciences: optionality on better treatment paradigms

In healthcare and life sciences, the companies we met looked less like “stories” and more like structured optionality on better care pathways. Here are some observations:

• Travere Therapeutics offers multiple billion‑dollar opportunities in rare kidney and metabolic diseases, with a near‑term regulatory catalyst in FSGS. This is not blue‑sky R&D; it’s a pipeline tied to very specific and addressable needs.
• Tandem Diabetes Care is attacking the diabetes technology side, lifting margins via the pharmacy channel and going after the under‑penetrated insulin‑intensive Type 2 segment.
• Halozyme Therapeutics is a royalty and platform business, riding the structural shift from hospital-based intravenous infusions to more convenient subcutaneous injections at home, a powerful tailwind for both patients and payers.

What matters for us is not just the science, but the economic architecture: reimbursement, regulatory risk, and capital allocation. Those are details you only fully understand after challenging management face‑to‑face. Our takeaway in healthcare is that we want focused, high‑value niches where regulation actually protects the moat. That’s where convertibles can offer attractive equity upside participation while still providing downside protection, as long as the issuer has strong credit quality.

4. Credit & consumer finance: discipline is the new growth

On the more credit‑cyclical side of the spectrum, the backdrop is very different, but the
conclusion is similar, discipline wins. For example:

• Encore Capital is operating in a very favourable environment for overdue consumer debt in the US. Elevated revolving balances, plus tighter regulation, are reinforcing the advantage of scale and compliance.
• Across Affirm, Upstart, and Encore Capital, the common theme is that credit risk remains elevated versus pre‑COVID. The winners are those who can price, segment, and collect better without sliding down the risk curve.

As convertible investors, this is exactly where our style matters: we are not buying “credit beta”, we are picking specific platforms where the combination of data, process, and regulation stacks the odds in favour of resilient economics. We then express that view with vehicles that give us optionality on the upside and a solid bond floor if the credit cycle gets rough.

What this means for our investors

To our investors, the message is simple: we are an active and high‑conviction manager. We are not buying “the market”; we are building a carefully selected portfolio of names we know well, often supported by direct conversations with management.
 

Our information edge comes from being on the ground, as we were on this Silicon Valley trip, and bringing those insights straight into security selection, position sizing, and structure analysis. We believe that the convertible bond asset class can turn these powerful secular themes into a more asymmetric, risk‑managed return profile.