January 2025

The fund returned +1.0% in January, outperforming the MSCI All World Index (AUD) which fell 2.2%.

Markets sold off late in the month as several factors hit at once: Kevin Warsh being floated as the next Fed Chair, volatility in precious metals, and renewed nerves around AI infrastructure spending following the release of Kimi k2.5.

Since October, global markets have been choppy, range bound, and highly rotational. Returns have been driven by short, sharp moves rather than sustained trends, with elevated volatility across regions.

Against that backdrop, the Fund’s returns have been modest, but importantly we have experienced smaller drawdowns than the major indices and outperformed after fees. This has been driven by a higher cash balance, disciplined position sizing, and a consistent hedging program.

Volatility has created opportunities to add selectively where we see clear mispricing, but we have remained patient while trends remain unclear. Our approach can lag in sideways markets, but it is designed to be well positioned to compound when conditions improve.

Markets sold off following Kevin Warsh’s nomination on January 30, as investors focused on his push to shrink the Federal Reserve’s balance sheet.

Trump has been clear that he wants lower rates, and Warsh would not have been nominated unless he agreed. The tension is liquidity. Cutting rates loosens financial conditions, but shrinking the Fed’s balance sheet drains liquidity and pushes up long term yields. Markets appear to be pricing this risk, with curve steepening and higher long dated yields in early February.

Warsh places less emphasis on labour markets and more on fiscal excess and monetary expansion as drivers of inflation, with AI driven productivity acting as a disinflationary force. If that view is right, sequencing matters. Cut rates first, then reduce the balance sheet gradually. The risk is execution. Until Powell’s term ends in May, messaging is likely to remain mixed.

The US economy remains resilient. Atlanta Fed GDPNow continues to point to expansion. AI related capex is now a meaningful contributor to growth.

Trueflation is trending lower, giving the Fed room to ease without reigniting inflation. This environment supports our positioning in quality businesses accelerating earnings, especially those with exposure to the AI buildout.

AI is already changing how work gets done. One person can now do what previously required a team, and that shift is accelerating.

Productivity and earnings growth are likely to exceed current expectations for years. Forecasts built on historical assumptions are likely to prove too conservative. We are positioning across the infrastructure enabling AI, the platforms deploying it, and the companies that benefit directly from the productivity gains.

Over the past six weeks, AI has taken a clear step forward. What was once a question and answer tool has evolved into autonomous agents that can build software, control a computer, and complete multi step tasks with minimal intervention. Tools such as Claude Cowork and OpenClaw are now good enough to delegate real work, and adoption is accelerating.

We are using these coding tools to build systems which improve the investment process. We run a real time portfolio dashboard with live visibility across valuations, exposures, fundamental changes, risk, and performance. AI systems scan markets, monitor portfolio positions, and surface actionable ideas.

Alongside this, I have built a custom AI stock analysis system refined through hundreds of iterations. It can quickly screen ideas or run a full deep dive across fundamentals, technicals, and cross market context, incorporating real time signals such as Google Trends, job postings, social sentiment, and analyst research.

The AI first pass process on our internal dashboard starts with selecting one of many LLMs linked via API.

The AI Analyst then works through predefined categories.

It then produces a structured output covering the key investment considerations.

This is just one of the tools we are using. It does not replace fundamental research or company and expert meetings. The data needs to be verified through our normal process, but the speed materially improves decision making and allows for a faster first pass.

OpenClaw pushes this further and, importantly, shows where software is heading. It is an open source autonomous agent that runs locally and connects directly to data providers and everyday tools. Once connected, the agent becomes the interface. There is no workflow to learn and no application to open. You interact with the agent, and it pulls data directly from the source.

Portfolio holding DigitalOcean rallied as a result. DigitalOcean provides simple hosting and deployment for tools like OpenClaw, positioning itself as core plumbing for the agent era.

Claude Cowork and OpenClaw have clarified what the market is now grappling with. If work is done by an agent connected directly to data feeds, the value of traditional middleman software declines. Interfaces and features matter less. Subscription based workflows are easier to bypass.

AI driven productivity is reshaping software economics. Building products is now far cheaper, and markets are reacting. The S&P 500 Software ETF is down around 25% since October as investors reassess software valuations in an AI first world.

Some of this reset is warranted. The new AI tools show how quickly features can be replicated. That does not mean software is finished, but it does mean moats matter more and multiples should reflect that reality.

Many high quality software businesses are now under real valuation pressure, with several large cap names deep into bear market territory. For example:

There are opportunities here, particularly among businesses with genuine network effects, deeply embedded workflows, or control over critical infrastructure and data that AI applications depend on.

AI requires physical infrastructure at scale.

Supply cannot respond quickly. New chip fabs and copper mines take years to build, while hyperscalers are already spending more than $200bn a year and increasing. The forces driving commodity demand appear to be strengthening.

At month end we held 18 long positions, a Nasdaq futures hedge, and 16% cash. Net exposure was 76%. Offshore equity exposure is fully hedged back to AUD, reflecting our view that the AUD will continue to appreciate.

Exxon Mobil is our largest position at 6.4% of the portfolio. Power demand is rising and Exxon is positioned to benefit through strong cash flow generation and ongoing capital returns, including buybacks. The stock contributed strongly in January.

Alphabet remains our largest technology holding. We re entered the position during the month. Alphabet has a full stack position across custom silicon, data centres, cloud, and distribution through Search, YouTube, and Android, with Gemini as its flagship model layer.

Amazon is a new position. It is infrastructure at scale. AWS remains the largest cloud infrastructure provider by market share and is core plumbing for the AI buildout.

We also hold a number of value oriented positions alongside selective commodity exposures.

We are entering an important reporting period. Several key holdings release earnings over the coming days, which will provide fresh insight into AI infrastructure spending and cloud demand.

The AI infrastructure cycle remains our core opportunity. With 16% cash, we retain flexibility to add on weakness or move decisively when risk reward skews in our favour.

Our approach is unchanged. Concentrate capital in the best ideas, cut losers quickly, let winners run, and deploy capital only when the setup is compelling.

Please reach out if you would like to chat.

Regards