1. Google cloud re-accelerates.

Cloud revenue jumped 82% to $24.77B making the capex spending easier to defend, but Alphabet must convert contracted demand into cash before AI infrastructure becomes the new money pit.
With CapEx now expected to reach ~$200 billion in FY26 (up from $185 billion previously), most of Alphabet's operating cash flow will be plowed back into AI spending.

2. AI’s next bottleneck.

US data center demand will reach 194 gigawatts by 2035.
For context, a typical large-scale nuclear reactor has a power output of about 1 GW.

New York City requires about 2 GW to power its lights, air conditioners and 665 miles of subway track.

3. Electrifying Europe.

The EU electrification plan aims to increase the share of electricity in final energy consumption from the current 23% to 46% by 2040.
Utilities is a key beneficiary via required investment in electricity networks and clean power generation to facilitate this increase.
The EU27 target implies 2040 total power demand of ~5,000TWh, representing a doubling over 2025 demand of 2,529 TWh.
The majority of generation capacity will come from Renewables plus further flexgen (Batteries and CCGT) to balance the system. Such a material expansion in electricity generation will most likely require continued growth in electricity network capex investments to enable generation delivery to demand.
There is of course an ETF European utilities.

Below: EU target versus current forecasts

4. The structural bull case for uranium is still in its early stages.

Global nuclear reactor capacity is projected to grow 44% over the next decade. Every new 1 GW reactor requires: ~400 tonnes of uranium for its initial core load. ~160 tonnes per year thereafter to keep operating.

5. Diversification away from the crowded AI trade?

Investing in a global equities index is currently not as diversified a strategy as you might expect. Based on our analysis of the past 12 months of global stock market returns, the MSCI All Country World Index had a correlation of 0.79 with AI stocks, meaning the returns were closely connected.

Healthcare stocks, on the other hand, had a correlation of minus 0.06 with AI. In other words, there was almost no relationship between the movements in AI stocks and those of healthcare stocks over the past year.
This suggests that healthcare has served well as a diversifier. Looking forward, we expect it to maintain its traditional role as a “defensive” sector that can protect portfolios during downturns due to its history of persistently strong earnings growth. Profits continue to be boosted by long-term shifts such as demographic change as well as innovation in pharmaceuticals and medical technology.
The strength of healthcare earnings in the past three decades has usually translated into a premium for healthcare valuations versus the broader market. Yet AI's dominance in the past few years means healthcare is now trading at a 15% discount.

We are buyers of Astrazeneca, Abbott and Sanofi.

Not a subscriber yet?