Energy: out of favor, with tailwinds = opportunity

The portfolio is deliberately concentrated in energy names, with just over half now invested in best-in-class businesses: strong assets (high resource certainty, long life, low cost), materially undervalued on earnings power at normalized commodity prices, and run by capable, aligned capital allocators. The macro is a tailwind, not the reason we own them. The lessons of recent years are proving very valuable — I am better positioned to pick the best assets and the best teams than in 2022.

Energy makes up roughly 4% of the MSCI World, versus a long-run weight closer to 10% and more than 15% in 2008 — one of the lowest weightings in modern history. Current prices are set in large part by financial markets rather than physical barrels, with positioning unusually short and government intervention massively distorting signals. Yet this underweight comes at a time when physical markets are increasingly tight. Global demand for oil and natural gas is at record levels, while inventories are exceptionally low by historical standards.

Four supports underpin the energy thesis.

1. Global population growth and rising prosperity remain the floor under the entire complex. The world continues to add people (roughly 70–75 million per year) and income. The first things rising incomes buy are energy-intensive: consumption, protein, mobility. Efficiency gains and electrification are real, but not enough to offset that underlying growth. Even bearish forecasts point to a fossil-fuel demand plateau in advanced economies, not a collapse.

2. Energy is the hidden AI trade: the binding constraint on AI is shifting from chips to power. A handful of companies are set to spend on the order of US$1 trillion a year on data centers — almost a quarter of Germany's GDP, every year — forecast to add 750 TWh of power demand by 2035, equal to Germany's entire annual electricity consumption plus half of France's. Whether that spending will ever earn an adequate return is unknowable; that it will happen is not. It means round-the-clock demand that renewables cannot cover, requiring nuclear, gas, coal, and heavy grid investment. As it turns out, access to abundant and affordable energy is as decisive in the digital age as it was during industrialization.

3. The world has already lost roughly 1.5 billion barrels of oil supply during the Strait of Hormuz closure, along with around a fifth of global LNG supply for months. Russia has suffered material damage to its energy infrastructure from increasingly powerful Ukrainian drone attacks, with 1/3 of its refining capacity currently offline. The strait has partially reopened, but the physical market is far from normalized: OECD inventories are at their lowest since records began in 2003, and the U.S. Strategic Petroleum Reserve sits at its lowest since 1983. The buffers that cushioned the shock are largely spent, at the cost of heightened geopolitical vulnerability. Moreover, the U.S. releases were structured as loans, not sales: borrowers must return the barrels plus a premium of over 20%. Today's relief is, quite literally, tomorrow's demand.

4. Higher-for-longer interest rates are choking capital-intensive, short-life projects — the surest setup for tomorrow's supply shortage. Overstretched fiscal balance sheets and excessive government spending will likely keep inflation, and therefore rates, higher for longer; that pressure only grows if fiscal intervention is needed to cushion AI-driven white-collar job losses. The beneficiaries are long-life, low-cost assets already in production and already paid for; that is precisely what we own. U.S. shale, the growth engine of global supply over the past decade, is short-cycle and structurally high-cost, and is now showing signs of depletion.

None of this is a prediction. A genuine reopening of Hormuz, or the return of OPEC spare capacity, could cap prices for a time, and a concentrated portfolio like ours will be volatile. But these are different risks from owning long-duration promises at speculative valuations. We own cash-generative assets bought cheaply. If the market takes longer to recognize that, the cost is patience, not thesis failure.